Citation
An Assessment of Alternatives for A National Computerized Criminal History System

Material Information

Title:
An Assessment of Alternatives for A National Computerized Criminal History System
Creator:
United States. Congress. Office of Technology Assessment.
Publisher:
U.S. Congress. Office of Technology Assessment
Publication Date:
Language:
English
Physical Description:
xi, 208 p. ; 28 cm.

Subjects

Subjects / Keywords:
computerized records ( kwd )
criminal history records ( kwd )
fingerprint card repositories ( kwd )
Due process of law ( lcsh )
Criminal justice, Administration of ( lcsh )
Genre:
federal government publication ( marcgt )

Notes

General Note:
A report by the Office of Technology Assessment (OTA) that addresses "1) the status of criminal history record systems in the United States; 2) the alternatives for a national computerized criminal history (CCH) system; 3) the possible impacts of such a system; and 4) the relevant policy issues that warrant congressional attention to ensure that the beneficial impacts of a national CCH system are maximized and the possible adverse impacts controlled or minimized" (p. iii).
General Note:
Original is missing pages 86-88, 106-108, 122-124, 135-136, 150-152, 189-190, 200-202.

Record Information

Source Institution:
University of North Texas
Holding Location:
University of North Texas
Rights Management:
This item is a work of the U.S. federal government and not subject to copyright pursuant to 17 U.S.C. §105.

Aggregation Information

IUF:
University of Florida
OTA:
Office of Technology Assessment

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

An Assessment of Alternatives for a National Computerized Criminal History System October 1982 NTIS order #PB83-166678

PAGE 2

Library of Congress Catalog Card Number 82-600643 For sale by the Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20402

PAGE 3

-.. Foreword This report addresses four major areas: 1) the status of criminal history record systems in the United States; 2) the alternatives for a national computerized criminal history (CCH) system; 3) the possible impacts of such a system; and 4) the relevant policy issues that warrant congressional attention to ensure that the beneficial impacts of a national CCH system are maximized and the possible adverse impacts controlled or minimized. Conducted at the request of the House and Senate Committees on the Judiciary, this study is the last of four components of the OTA assessment of Societal Impacts of National Information Systems. The other components include a September 1981 OTA report on Computer-Based National Information Systems: Technology and Public Policy Issues; a March 1982 background paper on Selected Electronic Funds Transfer Issues: Privacy, Security, and Equity; and an August 1982 OTA report on Implications of Electronic Mail and Message Systems for the U.S. Postal Service. In preparing this CCH report, OTA has drawn on working papers developed by OTA staff and contractors, extensive related research on criminal history record systems carried out by SEARCH Group, Inc., and others, and operating data and descriptive information provided by the Federal Bureau of Investigation and various States. The final draft of this report was reviewed by the OTA project advisory panel and by a broad spectrum of interested individuals and organizations from the criminal justice community. OTA appreciates the participation of the advisory panelists, external reviewers, and others who helped bring this study to fruition. It is, however, solely the responsibility of OTA, not of those who so ably advised and assisted us in its preparation. Director ///

PAGE 4

Computerized Criminal History Systems Advisory Panel Arthur S. Miller, Chairman Professor of Law Emeritus The George Washington University Joe Armstrong Armstrong Associates Jerry Berman Legislative Counsel American Civil Liberties Union Robert Gallati Professor of Criminal Justice Northeastern University Lance Hoffman Professor of Electrical Engineering The George Washington University Gerald Caplan* Professor of Law The George Washington University John Kennedy Director, Rodino Institute of Criminal Justice Jersey City State College Steve Kolodney Executive Director SEARCH Group, Inc. Barry Mahoney Director of Research Institute for Court Management Jeffrey A. Meldman Professor of Management Science Massachusetts Institute of Technology Arthur R. Miller Professor of Law Harvard University J. Francis Pohlhaus** Former Counsel National Association for the Advancement of Colored People Christopher Pyle Professor of Political Science Mount Holyoke College James Rule Professor of Sociology State University of New York at Stony Brook Richard M. Schmidt General Counsel American Society of Newspaper Editors Carl Vorlander Executive Director National Association of State Information Systems Iram Weinstein System Planning Corp. Elyce Zenoff Professor of Law The George Washington University *Resigned in 1980. *~ eCea ~ e d, iv

PAGE 5

OTA Computerized Criminal History Systems Project Staff John Andelin, Assistant Director, OTA Science, Information, and Natural Resources Division Sam Hale, Interim Program Manager Communication and Information Technologies* Fred B. Wood, Project Director** Jean E. Smith, Study Coordinator*** Zalman A. Shaven, Senior Analyst Fred W. Weingarten, Senior Analyst* Elizabeth A. Emanuel, Administrative Assistant Shirley A. Gayheart, Secretary Jeanette Contee, Wordprocessor Contractors Kenneth C. Laudon, John Jay School of Criminal Justice, City University of New York Donald A. Marchand, Bureau of Governmental Research and Service, University of South Carolina Myron Uretsky, Business Experience International, Inc. Renee G. Ford, Tifford Producers, Inc. OTA Publishing Staff John C. Holmes, Publishing Officer John Bergling Kathie S. Boss Debra M. Datcher Joe Henson *March 1981 through February 1982; Stephen E. Doyle served as Program Manager through February 1981; Fred W. Weingarten was appointed Program Manager effective March 1982. l Ruann E p en g ov S er} ,d a 9 ~oject Directo r fro m C)ctober 1978 to June 1979; Marcia J. Mc,Naughton served as Study Director from October 1978 to July 1980. l **Since March 1981.

PAGE 6

Acknowledgments The following individuals served as subcontractors or reviewers. However, the report is solely the responsibility of OTA. Subcontractors Under the direction of Donald A. Marchand: Eva G. Bogan, Steven W. Hays, Sena H. Black, Lynne Eickholt Cooper, and Mark E. Tompkins of the Bureau of Governmental Research and Services, University of South Carolina. Under the direction of Kenneth C. Laudon: Abbe Mowshowitz, Michael B. Brink, Karl Kindig, Alan Mitchell, Marcy Block, Celeste Norris, Michael Cheek, Richard Rahm, George Wright, Marian Stevens, Steven Ruth, Thomas Carrillo, Donald Betzold, Mark Robin, Sharon Weber (City University of New York), and Richard Faust (Calculogic Corp.). Reviewers Hon. Sidney P. Abramson, Minnesota State Court H. A. Albert, Texas Department of Public Safety S. S. Ashton, Jr., U.S. Department of Justice Conrad Banner, FBI Identification Division Louise Becker, Congressional Research Service Dwight E. Bee, Illinois Department of Law Enforcement Howard G. Bjorklund, Wisconsin Division of Law Enforcement Services Robert J. Bradley, Missouri State Highway Patrol W. Gray Buckley, Colorado Bureau of Investigation William G. Connelie, New York State Police William J. Corbett, Ann Arbor (Michigan) Police Department William C. Corley, North Carolina Police Information Network Adam F. DAlessandro, New York Division of Criminal Justice Services Thomas Dalton, Seattle University Norman Darwick, International Association of Chiefs of Police Richard L. DeHahn, U.S. Department of Justice Richard W. Delaplain, Alaska Court System Robert L. Edwards, Florida Department of Law Enforcement Stuart A. Elliot, Kansas Highway Patrol L. G. Finn, Louisiana State Police Harry Garmat, Connecticut State Police Larry Goldman, International Association of Chiefs of Police Carl Gray, Jr., Kansas Highway Patrol David G. Hall, Wyoming Division of Criminal Identification Margaret J. Horvath, Florida Department of Law Enforcement Gerald L. Hough, Michigan State Police Donald G. Ingraham, Alameda County (California) District Attorneys Office Benjamin R. Jones, Pennsylvania State Police Julie Jones, U.S. Department of Justice Paul H. Kusuda, Wisconsin Division of Corrections E. W. Manseau, Georgia Bureau of Investigation David F. Nemecek, FBI National Crime Information Center Ronald L. Nolfe, American University A. Lawrence Peterson, Montana Board of Crime Control Thomas Poch, Ramsey County (Minnesota) District Attorneys Office Larry P. Polansky, District of Columbia Courts James R. Rebo, Minnesota Supreme Court Information Systems Office J. Randolph Riley, Tenth Prosecutorial District (Raleigh, N. C.) District Attorney Louis H. Sakin, Massachusetts Criminal History Systems Board Duane A. Sciford, Alameda County (California) Public Defenders Office Jack J. Schmerling, National Legal Aid and Defenders Association Jan M. Schultz, Alabama Administrative Office of Courts P. H. Schultz, Washington State Patrol Charles J. Shuster, District of Columbia Police Department Terril Steichen, National Telecommunications and Information Administration, U.S. Department of Commerce Walter E. Stone, Rhode Island State Police Timothy P. Sweeney, National Law Enforcement Telecommunications System Ray L. Turner, Memphis (Tennessee) Police Department W. R. Wagner, Jr., Virginia State Police Frederic Withington, Arthur D. Little, Inc. Fred Wynbrandt, California Department of Justice Raymond Young, FBI Consultant vi

PAGE 7

Content s Chapter Page Overview ix 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. Summary . . 3 Nature and Origins of Criminal History Records 21 Evolution to Computerized Criminal History Records 31 Description of the National Crime Information Center in Context 39 NCIC Technology and Costs ... 51 Legal/Regulatory Framework for NCIC, Ident, and State CCH Systems 61 Use and Users of NCIC/CCH, Ident, and State CCH Systems 77 Record Quality in Federal and State Criminal History Information Systems. 89 State and Local Management of Criminal History Information Systems 99 Major Structural Alternatives for a National Computerized Criminal History System . . . 109 Possible Impacts of a National CCH System on the Criminal Justice Process . 125 Other Impact Areas Relevant to a National CCH System: Employment and Licensure, Minority Groups, Federalism, Monitoring or Surveillance Potential, and Constitutional Rights . . . . 137 Congressional Policy Considerations, Part I . 153 Congressional Policy Considerations, Part II . . . . 167 Appendixes Page A. Status of NCIC Hot Files . . . . . 191 B. State Officials Responding to OTA 50-State Survey Conducted in 1979-80 193 C. State Repository Officials Responding to OTA User Survey Conducted in 1979 . . . . . . . 195 D. State Officials Contacted in OTA Followup Survey Conducted in August 1982 ..... . . . 197 Index . . . . . . 203 vii

PAGE 8

Overvie w The United States already has a national criminal history system. It is partly manual and partly computerized, and includes criminal record and fingerprint card repositories maintained by the Federal Bureau of Investigations (FBI) criminal identification file (known as Ident) and 49 State identification bureaus. The national system also includes the computerized criminal history (CCH) files in the FBIs National Crime Information Center (NCIC) and in 27 States. Seven other States have a computerized name index to their manual files, and 10 more States are in the process of implementing a computerized index. As of October 1981, Ident held about 6 million automated criminal history records, NCIC/ CCH held about 1.9 million, and the 27 State CCH files collectively held about 11.5 million records. For the interstate exchange of criminal history records, the national system uses the U.S. Mail, the NCIC communication network, and the National Law Enforcement Telecommunications System (NLETS). The many local and metropolitan criminal history record systems, either manual or automated, are also part of this national system. Thus, most of the building blocks for a national computerized criminal history record system are already in place. Technically, there are many ways that a national CCH system could be designed. At one end of the spectrum, criminal history records for all offenders could be stored in a central national repository. At the other end, a national repository could be limited to records of Federal offenders, with records of State offenders stored only in the respective State repositories. The emerging consensus among Federal and State criminal record repository and law enforcement officials favors the latter, with only Federal offender records and an index to State offenders (known as the Interstate Identification Index or III) maintained at the national level along with a national fingerprint file on serious criminal offenders. Criminal history records are used at all levels of government, by all sectors of the criminal justice community, and increasingly by the noncriminal justice community as well. To the extent that a national CCH system provides information that is more complete, timely, and verifiable (based on positive identification) than is presently available, the system would improve the functioning of the criminal justice process. The most significant improvements are likely to be in the areas of criminal investigations, police booking and intake, pretrial release and bail decisions, and presentence investigation reports. For example, the impact could be particularly significant in pretrial release and bail decisions, which typically must be made within 36 to 72 hours after arrest. If accurate and complete, CCH records could help prosecutors and judges better balance the need to protect the public from harm by defendants out on bail versus the need to minimize the detention of defendants on charges for which they have not been tried under due process of law. The potential contribution of a national CCH system becomes even more important in view of State bail and sentencing reforms that place greater reliance on criminal history information, and the many recommendations of the U.S. Attorney Generals Task Force on Violent Crime that involve criminal history records. Depending on how a national CCH system is controlled and used, the quality of the CCH records exchanged, and the standards set for access and operation, the system could have important implications for employment and licensure, Federal-State relationships, and civil and constitutional rights, as well as for public safety and the administration of justice. Full implementation of III (or any other national CCH system) raises a number of issues that warrant congressional attention to ensure that beneficial impacts are maximized and potentially adverse impacts are controlled or minimized. Policy Control. Considerable debate has focused on which agency or organization(s) should have direct policy control over a national CCH system. Suggestions include a consortium of States, a broadened and strengthened NCIC Advisory Policy Board (APB), an independent board, and/or the FBI. For example, a strengthened APB could include greater repix

PAGE 9

resentation from the prosecutorial, judicial, correctional, and public defender sectors than at present, and could include an advise and consent role, at least with respect to State and local participation in a national system. There are many other possibilities, but the key issue is how to devise a mechanism that will effectively represent the interests of the diverse users of a national system, and afford them a strong and possibly controlling role. Record Quality .-Since 1970, Congress has expressed its concern about the completeness and accuracy of criminal history records. Based on the results of record quality research conducted by OTA and others, the quality of criminal history records at the State level has improved; however, significant problems remain, especially with respect to court disposition reporting. The average nationwide disposition reporting level increased from 52 percent in 1970 to 65 percent in 1979, but has changed little since then (to 66 percent in 1982). Fourteen of 41 States in 1979 and 13 of 47 States in 1982 indicated that disposition reporting to the State repositories was less than 50 percent. In both 1979 and 1982, eight States indicated a reporting level of less than 25 percent. With a national index, the FBI would no longer maintain non-Federal records, and the problems of record quality in Ident and NCIC/ CCH would be reduced. However, the quality of records maintained by the States, as well as the quality of any index based on those records, would still be a matter of concern. The progress made by many States in recent years indicates that continued improvement in disposition reporting is possible but would require a significant further commitment measured in manpower, dollars, and system improvements at the State and local levels. As of August 1982, 49 of 50 States maintain transaction logs of criminal history records disseminated, 35 of 46 routinely employ quality control checks on the accuracy of input data, 30 of 49 have automated or manual procedures for the regular review of court disposition reporting, and only 13 of 49 States have conducted a record quality audit. File Size and Content.Under the III concept, the national index would include only names and identifying information. Index size would depend on what limitations are placed on entries (e.g., with respect to types of offenses and the handling of juvenile offender records), how long entries are kept in the index (e.g., limited retention period for some types of entries), and how the index is initially established and then maintained and updated. The index could have as many as 21 million entries if all persons with arrests for serious or significant offenses were included. Noncriminal Justice Access. Significant noncriminal justice use of Federal and State criminal history record systems, coupled with widely varying State statutes defining authorized users and State policies on sealing and purging, has generated concern about control of access to criminal history records. As of fiscal year 1981, about 53 percent of requests to Ident were from noncriminal justice users (30 percent Federal and 23 percent State/local). As of 1982, roughly 15 percent of requests to State CCH systems were for noncriminal justice purposes. Noncriminal justice access to a national index could be prohibited entirely, or could be permitted only under stronger Federal guidelines than presently exist. A dual index could be established, one for criminal justice use and a second for noncriminal justice use, perhaps with the latter based on disposition or conviction information only. Even under the status quo, access to a national index would require complicated safeguards (which are technically feasible with a computer-based system) to be consistent with a wide variety of existing State laws and regulations, and would require some means to resolve conflicts among State laws, and between Federal statutes and Executive orders and State laws. Oversight and Audit.Oversight mechanisms would help assure Congress, the public, and others that a national index (or any other national CCH system) is operating within the boundaries of law and regulation, and to help identify any problems that might emerge. Congress could require an annual report and periodic audits of Federal and State CCH files to help ensure compliance with whatever system standards may be established. x

PAGE 10

Federal Funding. Throughout the 1970s, it was Federal Government policy to support the development of State CCH systems and the implementation of Federal privacy and security regulations. However, Federal support has been phased out. The following three areas are possible priorities for further Federal funding: 1) improving court disposition reporting on a nationwide basis; 2) upgrading criminal history record systems in States that are operating manually, or assisting those in the process of automating their name index and/or file; and 3) improving procedures in all States where necessary to assure the accuracy and completeness of criminal history information, to conduct audits of local users, to maintain and periodically review transaction logs, and to train employees and users. Message Switching.Unless all criminal history records were stored in one place (e.g., a national CCH repository), a national CCH system would require some electronic means to transfer criminal history records (and inquiries for such records) among and between the various State and Federal repositories and participating agencies. The transfer or switching of messages from one State to another through the NCIC computer has been a point of controversy with respect to the impact on Federal-State relations and the potential for monitoring and surveillance use. There are several message switching alternatives for III. First, inquiries could be switched via NCIC, with records returned via the NLETS message switching system. This approach has been used in pilot tests of III. Second, both inquiries and records could be switched via NCIC. Third, both inquiries and records could be switched via NLETS. Fourth, records could be switched via NLETS and inquiries via NCIC or NLETS. Fifth, the use of NCIC or NLETS could be optional for switching of both inquiries and records. Any Department of Justice or FBI message switching role in a fully operational III (or other national CCH system) would probably require congressional approval. Federal Direction and Legislation.In the 7 years since Congress last considered legislation on criminal history record systems, both the States and the FBI have made significant progress in improving the interstate exchange of criminal history information and in implementing State and Federal privacy and security regulations. Substantial consensus has developed around III, and pilot tests indicate that III is technically feasible. Nonetheless, absent Federal direction and probably some modest Federal funding, full implementation of III is likely to take many years. Also, further improvement in nationwide record quality and some kind of national standards on record access and use are needed. Resolution of conflicts between and among State and Federal laws is a necessity. Legislation represents one of the strongest measures to provide Federal direction and ensure accountability and control. Legislation could provide explicit authority for a national index or other national CCH system, and include statutory guidelines for its operation and use. In addition to the areas listed above, legislation could address access, review, and challenge procedures; criminal penalties; privacy standards; and possibly a prohibition on unauthorized intelligence or surveillance use of a national CCH system. In sum, legislation appears to be the most appropriate vehicle for guiding the full implementation of a national CCH system in a way that will enhance the efficiency and effectiveness of the criminal justice process, protect civil and constitutional rights, and properly balance the roles and responsibilities of the Federal and State Governments. xi

PAGE 11

Chapter 1 Summary

PAGE 12

. .-. Contents Page Introduction . . . . . . . . . . . . . 3 Current Status of Criminal History Record Systems.. . . . . 3 Criminal History Record Repositories . . . . . . . 3 Interstate Exchange of Criminal History Records . . . . 4 Use of Criminal History Records . . . . . . . . 4 Multi-State Offenders.. . . . . . . . . . . 5 Fingerprint Identification. . . . . . . . . . 5 Record Quality . . . . . . . . . . . . 6 Significance of Record Quality Problems . . . . . . 7 Privacy and Security Protection . . . . . . . . 8 Alternatives for a National CCH System . . . . . . 8 Improving Response Time . . . . . . . . . 9 Improving Record Quality . . . . . . . . . 10 Shifting Preferences on System Structure . . . . . . 10 Possible Impacts of a National CCH System . . . . . . 11 Criminal Justice Process. . . . . . . . . . 11 Employment and Licensure . . . . . . . . . 12 Minority Groups . . . . . . . . . . . 13 Federalism . . . . . . . . . . . . 14 cost . . . . . . . . . . . . . . 14 Surveillance Potential .. $ . . . . . . . . 15 Message Switching . . . . . . . . . . . 16 Congressional Policy Considerations . . . . . . . 17 Policy Control . . . . . . . . . . . . 17 File Size and Content . . . . . . . . . . 17 Record Quality . . . . . . . . . . . 17 Nonc riminal Justice Access + . . . . . . . . 17 Oversight and Audit.... . . . . . . . . . 18 Public Participation. . . . . . . . . . . 18 Comprehensive Legislation . . . . . . . . . 18 III Development Plan... . . . . . . . . . 18 AIDS/CCH Consolidation. . . . . . . . . . 18

PAGE 13

Chapter 1 Summar y Introductio n This report addresses four major areas: l l l l the status of criminal history record systems in the United States; the alternatives for a national computerized criminal history (CCH) system; the possible impacts of any such system; and relevant policy issues that warrant congressional attention to ensure that the beneficial impacts of a national CCH system are maximized and the possible adverse impacts controlled or minimized. These areas are of concern because: criminal history records are a vital part of the criminal justice process; advances in computer and communication technologies can help to improve the completeness, accuracy, and timeliness of such records; and the use of criminal history records, particularly when exchanged-via a national system, can have important implications for public safety and the administration of justice, employment and licensure, Federal-State relationships, and civil and constitutional rights. Until the 1850s, criminal history records in the United States were largely informal notes maintained by local police officers. Since that time, however, such records have become more formalized, centralized, widely used, and technology-based. Originally, criminal history records were known as rap sheets. They contained information ranging from personal identifiers (e.g., height, weight, eye color, fingerprints, and/or identification numbers) to arrests (date, jurisdiction, and charges). The information contained in criminal history records also may include court disposition of charges, sentencing, incarceration, and the like. While police were the earliest users of criminal history records, such records are now used to varying degrees at all stages of the criminal justice process by police, prosecutors, public defenders, judges, and probation officials, among others. They are also used for noncriminal justice purposes such as employment and licensing decisions and security checks. Current Status o f Criminal Histor y Record System s Criminal History Recor d The use of computers is already widespread. Repositorie s Ident has made progress in automating its own operations through the Automated IdenCriminal history records are stored at the local, State, and Federal levels. Since 1924, the Federal Bureau of Investigation (FBI) has maintained a national repository of fingerprint cards and rap sheets in its Identification Division (known as Ident). Forty-nine of the fifty States now have their own criminal history record repositories. tification System (AIDS). As of October 1981, almost 6 million of Idents criminal records had been automated (representing more than one-fourth of the individuals in the criminal file) and fingerprints for about 70 percent of the individuals in the file had been converted to a machine-readable (automated) format. Since 1971, the FBI has also maintained a 3

PAGE 14

4 An Assessment of Alternatives for a National Computerized Criminal History System CCH file in its National Crime Information Center (NCIC), although only eight States currently keep records in this file. As of October 1981, it contained about 1.9 million records, including approximately 0.5 million Federal offender records. At the State level, as of August 1982, 27 States had CCH files, 7 had an automated name index, and 16 had a completely manual system. Ten of the sixteen manual States are in the process of implementing an automated index, and two are implementing a CCH file. Also, the 27 States with CCH files accounted for about 85 percent of all criminal fingerprint card activity, and collectively maintained about 11.5 million CCH records as of September 1981. At the local level, most major metropolitan police departments use computerbased systems (19 have direct lines to NCIC). For those 12 States in the process of implementing an automated name index and/or CCH file, the estimated time to completion ranged from 1 month, to 1 year, to an indefinite time period, due largely to variations and/ or uncertainties in staffing and funding. With full implementation, all but four States would have at least an automated name index, and two of the four remaining manual States do have plans to automate. Interstate Exchange o f Criminal History Record s The exchange of criminal history records among the States and between the States and Federal Government can be accomplished in several ways. The exchange of records with Ident is almost entirely by mail, since Ident does not have direct communication lines to the States. Exchange with NCIC/CCH is almost entirely electronic, since NCIC has direct communication lines to all 50 States (49 of which are authorized to access the NCIC/CCH file) and to several Federal agencies. Use of the CCH file involves about 4.4 million transactions annually, but only about 3.5 percent of total NCIC traffic. Only eight States keep records in the CCH file. Of the 10 files maintained in NCIC, the bulk of traffic involves the eight so-called hot files, which furnish an electronic bulletin board capability used by law enforcement agencies to list wanted or missing persons or stolen properties (e.g., vehicles, guns, and securities). NCIC is currently testing the concept of an Interstate Identification Index (III) in which the NCIC/CCH file includes only records for Federal offenders plus a national index of State offenders, and the participating States maintain both single and multi-State offender records. The exchange of crimina l history records can also be accomplished via the National Law Enforcement Telecommunications System (NLETS), a computerized message switching network linking local, State, and Federal agencies. Operated by a nonprofit corporation controlled by the States, NLETS does not hold or manage record files, but provides the capability to switch Crimin al history records among 49 of the 50 States. Some Federal agencies, notably the Treasury and Justice Departments, can exchange criminal history records over their own nationwide communication networks that interface with NCIC/CCH. Use of Crimina l History Record s Criminal history records are used at all levels of government, by all sectors of the criminal justice community, and increasingly by the noncriminal justice community as well. During fiscal year 1981, about 18 percent of Ident use was by law enforcement agencies, 29 percent by other criminal justice agencies (e.g., prosecutors, courts, and corrections), and 53 percent by noncriminal justice agencies (primarily for employment and licensing and security checks). About 33 percent of Ident use was by Federal agencies and 67 percent by State/local agencies. Based on 1979 and 1982 OTA surveys, the use of State CCH repositories was roughly 56 percent by law enforcement agencies, 29 percent by other criminal justice, and 15 percent by nonc riminal justice. Data from the 1981 III

PAGE 15

. -. Ch. lSummary l 5 pilot test suggest that NCIC/CCH is used almost entirely by criminal justice agencies about 86 percent by law enforcement and 14 percent by other criminal justice (about 12 percent by Federal agencies and 88 percent by State/local agencies). The widespread use of criminal history information throughout the criminal justice process has been confirmed by other user surveys conducted by OTA, the Department of Justice, the Florida Department of Law Enforcement, and dozens of independent researchers. The use of criminal history records becomes even more important in view of State bail and sentencing reforms that place greater reliance on criminal history information, and the many recommendations of the U.S. Attorney Generals Task Force on Violent Crime that involve criminal history records. For example, in April 1982, the U.S. Senate Committee on the Judiciary approved legislation, recommended by the Task Force, requiring handgun purchasers to wait 14 days to pick up their weapons so that police departments will have time to conduct a criminal record check. The picture is a little less clear with respect to noncriminal justice use. As noted above, the use of Ident is already greater for noncriminal justice than for criminal justice purposes, and as of August 1982, 7 of 45 States reported that noncriminal justice use of criminal history records accounted for more than 40 percent of total use. At least 14 States have recently enacted (since 1979) or have pending State legislation or regulations that further broaden noncriminal justice access. Delays resulting from the noncriminal justice workload reached the point where Ident suspended most State and local applicant services (for licensing and employment checks) for fiscal year 1982. These will be reinstated on October 1, 1982, but on a fee-for-service basis. Multi-State Offender s Based on 1979 research, OTA found that about 30.4 percent of individuals in the FBI Ident criminal file had arrests in more than one State, which closely approximated a 1974 FBI estimate of 30 percent and a 1981 FBI estimate of 33 percent. Based on 1981 data available to OTA for eight States, multi-State offenders ranged from a low of about 3 percent to a high of 36 percent, with Federal offenders excluded. The average was about 12 percent, and only one State was above 16 percent. Nonetheless, the percentage of multiState offenders appears to be significant. Whether the crimes co remitted by multi-State offenders tend to be more or less serious than those of single-State offenders could not be positively determined from information available to OTA. This is an area of possible further study. Fingerprint Identification Criminal justice practitioners believe that, at present, fingerprints are the only reliable and consistent basis for positive identification. The exchange of records based on names alone results in a high percentage of errors due to the frequent use of aliases and similarities among many common surnames. In a 1982 III pilot test, the FBI found that almost one-third of the matches between individuals and records were in error when based on name-searching techniques alone. Both Ident and State identification bureaus process fingerprint cards received from criminal justice agencies, but manual fingerprint processing is extremely time-consuming and labor-intensive, and therefore costly, especially at the high volumes presently experienced. A 1981 FBI survey estimated that 4.16 million criminal fingerprint cards were received annually by State identification bureaus, and 2.91 million criminal fingerprint cards by Ident. Idents experience exemplifies the enormity of the problem. As of October 1981, there were million crimin al fingerprint cards representing 21 million individuals in the Ident criminal file. During fiscal year 1981, Ident received an average of 12,684 criminal fingerprint cards daily. Surveys conducted for the FBI in 1979 and 1980 indicated that the average Ident response time for processing fingerprint cards was about 36 workdays. As of July and Octo-

PAGE 16

6 l An Assessment of Alternatives for a National Computerized Criminal History System ber 1981, the FBI estimated that Ident internal processing time (excluding mailing time) was averaging 27 and 25 workdays, respectively, for all categories of inquiries (both fingerprint checks and name checks). As of July 1982, processing time had improved, at least temporarily, to about 13 days, due to Idents l-year suspension of record checks for federally chartered or insured banking institutions and State and local employment and licensing authorities. There is general agreement that improvement in fingerprint processing time is necessary, particularly to meet needs that arise early in the crimin al justice process where decisions must be made very quickly, for example, in bringing charges and setting bail. OTA did not assess specific alternatives for improvement, but major studies have recently been completed by the Jet Propulsion Laboratory (JPL) and the International Association for Identification. However, it seems clear that fingerprint identification is properly viewed as an integral part of any national CCH system and that automated fingerprint classification and search technology offers substantial promise for improvement. Record Qualit y Since 1970, Congress has expressed its concern about the completeness and accuracy of criminal history records. Section 524(b) of the Crime Control Act of 1973 required the Law Enforcement Assistance Administration (LEAA) to promulgate regulations to provide safeguards for the privacy and security of criminal history records, including their completeness and accuracy. The 1975 regulations (known as title 28, Code of Federal Regulations, pt. 20) apply to the Federal Government and to all States whose criminal history record systems were federally funded in whole or in part. Federal courts have also ruled on record quality issues. For example, in Tarlton v. Saxbe (1974) the U.S. Court of Appeals for the District of Columbia ruled that the FBI had a duty to prevent dissemination of inaccurate arrest and conviction records, and had to take reasonable precautions to prevent inaccuracy and incompleteness. Most States now have statutes or regulations requiring agencies to ensure reasonably complete and accurate criminal history information, including timely reporting of court dispositions. The number of States with statutes or regulations on record quality increased from 14 in 1974 to 45 in 1979, and to 49 in 1981. Based on the results of record quality research conducted by OTA and others, the quality of criminal history records has improved since 1970; however, significant problems remain. For Ident, OTA record quality research found that, based on a 1979 sample of arrest events, about 30 percent of the Ident records that could be verified lacked a court disposition that had occurred and was confirmed by the district attorney in the local area responsible for prosecution. A 1980 study by JPL found that Ident receives dispositions for about 45 percent of the arrests reported. OTA also found that about one-fifth of the Ident arrest events sampled were inaccurate when compared with charging, disposition, and/or sentencing information in local records. With respect to NCIC/CCH, OTA record quality research found that, based on a 1979 sample of arrest events, about 27 percent of the CCH records that could be verified lacked a court disposition that had occurred. About one-fifth of the arrest events sampled were inaccurate with respect to charging, disposition, and/or sentencing information. While it is possible that NCIC/CCH and Ident record quality has improved since 1979, OTA is not aware of any comparable research conducted by the FBI or others to document such improvements. At the State level, a comparison between a 1979 OTA 50-State survey and a 1973 General Accounting Office (GAO) study (based on a 1970 50-State survey conducted by LEAA) shows some improvement in the average level of disposition reporting. The GAO study found the average level to be about 52 percent for the 49 States responding; the OTA study showed an average level of about 65 percent for the 41 States responding. However, the

PAGE 17

Ch. lSummary l 7 - 1979 average for computerized States (with a CCH file and/or automated name index) as opposed to manual States was even higher (about 71 percent compared to 50 percent for manual States). Given that in 1970 only one State (New York) had a CCH system, the results indicated that most of the improvement in disposition reporting over the 1970-79 period was in States with CCH systems. OTA also sampled State records in one major urban jurisdiction in each of three States. For the three urban jurisdictions, disposition reporting was 58, 60, and 85 percent. Several States contacted by OTA have achieved further improvement in disposition reporting since 1979. However, between 1979 and 1982, average disposition reporting levels for all States responding improved only marginally, to about 66 percent. In the OTA 50-State survey, 14 of 41 States responding in 1979 and 13 of 47 States in 1982 indicated that disposition reporting to the State repository was less than 50 percent. In both 1979 and 1982, eight States indicated a reporting level of less than 25 percent. Significance of Record Quality Problem s On the one hand, Federal and State law emphasizes the importance of complete and accurate criminal history records, but on the other, the law authorizes the dissemination of records, whether or not they are accurate and complete, for a variety of purposes. For example, Federal regulations and FBI operating procedures assign agencies that enter records into Ident or NCIC the responsibility to assure that information on individuals is kept complete, accurate, and current. The FBI helps to maintain the integrity of the NCIC files through automatic computer edits and purges, quality control checks, and periodic record validations by originating agencies. Similar procedures are possible in Ident through the use of AIDS. Yet, with few exceptions, Federal and State law authorizes the dissemination of criminal history records with or without dispositionsto the criminal justice community. Law enforcement and prosecuting agencies, in particular, find that an incomplete and/or inaccurate record can be useful as a pointer to the location of complete and accurate information, even though an arrest-only record is not admissible in criminal trial proceedings under the laws of criminal evidence in most jurisdictions. With respect to noncriminal justice use, Federal regulations permit dissemination of Ident and NCIC/CCH records without dispositions to Federal noncriminal justice agencies if authorized by Federal statute or Executive order. Dissemination is also permitted to State and local noncriminal justice agencies if authorized by Federal or State statutes and approved by the U.S. Attorney General, except for records without dispositions where the arrest charge is more than 1 year old and is not under active prosecution. At the State level, as of mid-1981, 37 States authorized dissemination of arrest-only records to a variety of State and local noncriminal justice agencies (primarily for employment and licensing purposes), and 27 States authorized such dissemination to private sector organizations and individuals. The disclosure of such records to private parties frequently depends upon factors other than State law, such as local law, local agency policy, or the impact of the States public record or freedom of information law. In most court cases where the completeness or accuracy of criminal records has been challenged, the balancing of individual rights of privacy and due process versus the maintenance of public safety and welfare has proven a difficult challenge to the courts. Yet the Federal courts have found violations of civil and constitutional rights, particularly when arrest-only information is used in minority employment decisions (see Gregory v. Litton Systems, 1970) and when arrest information without otherwise available disposition information is used in criminal justice decisions such as setting bail (see Tatum v. Rogers, 1979). ISEARCli Group, Inc., 7Fends in State Securit~r and Pri\acI Z~gislation, Sacramento, Calif., November 1981, p. 10.

PAGE 18

8 l An Assessment of Alternatives for a National Computerized Criminal History System Privacy and Security Protectio n While very important, record quality (accuracy and completeness) is only one aspect of privacy and security protection. In enacting section 524(b) of the 1973 Crime Control Act, Congress also stressed the importance of protecting individual privacy by limiting record dissemination to lawful purposes, by permitting individuals to access, review, and challenge their records, and by ensuring the security of criminal history record systems. Title 28 of the Federal regulations required States accepting Federal funding to develop specific policies and procedures in these and other areas. Since 1974, when statistics were first compiled, the States have made substantial progress. For example, as of mid-1981, over twothirds of the States had statutes ardor regulations that: l l l establish a State regulatory authority for privacy and security of criminal justice information systems (46 States in 1981 compared with 7 in 1974); place some restrictions on the dissemination of criminal history information (all States and the District of Columbia in 1981 compared with 12 in 1974); establish the rights of individuals to inspect their criminal history records (43 States compared with 12); provide criminal sanctions for violation of privacy and security laws (39 States compared with 12); and establish the rights of individuals to challenge the accuracy and completeness of record information pertaining to them (35 States in 1981 compared with 10 in 1974). Nonetheless, even where States have enacted laws or regulations, wide diversity remains in the specific provisionsfor example, in sealing and purging procedures, in statutory limitations on criminal history file content, and in the definition of authorized users. Also, States vary widely in their implementation of privacy and security measures such as record quality audits, court disposition monitoring, quality control checks, and routine review of transaction logs. Based on a 1982 50-State survey, OTA found that only 13 of 49 States responding had ever conducted a record quality audit. Thirty of 49 had automated or manual procedures for the regular review of court disposition reporting, and 35 of 46 routinely employ quality control checks on the accuracy of input data. Forty-nine of fifty States maintained transaction logs of criminal history records disseminated, although most indicated that the logs were reviewed only when a specific abuse was indicated. Alternatives for a National CCH Syste m The United States already has a national itan criminal history record systems, either criminal history system. It is partly manual and partly computerized, and includes criminal record and fingerprint card repositories maintained by Ident and 49 State identification bureaus. The national system also includes the CCH files in NCIC and 27 States. For the interstate exchange of criminal history records, the national system uses the U.S. Mail, the NCIC and NLETS communication networks, and, to a lesser extent, the communication networks of the Justice and Treasury Departments. The many local and metropolmanual or automated, are also a part of this national system. Thus, many but not all of the building blocks for a national computerized criminal history record system are already in place. Technically, there are several ways that a national CCH system could be designed. At one end of the spectrum, criminal history records for all offenders could be stored in a central national repository such as Ident. The full development of AIDS or the NCIC/CCH file

PAGE 19

Ch, lSummary l 9 -. could constitute a national CCH repository when hooked up to the NCIC (or other) communication lines to permit nationwide electronic access. The repository would include records on roughly 21 million persons with arrests for serious or significant offenses. At the other end of the spectrum, a central national repository could be limited to records of Federal offenders (approximately 0.5 million), and records of State offenders would be stored only in the respective State repositories. An intermediate alternative (known as the single-State/ multi-State approach) would be for a national repository to maintain records of all multiState as well as Federal offenders, with singleState offender records stored by the States. Given the constitutional prerogatives of the States with respect to crimin al justice, and the fact that 49 of the 50 States now maintain their own State repositories, records on State offenders will continue to be maintained by the States whether or not a national CCH system is implemented. Therefore, any State records maintained in a national repository will incur extra costs (to the Federal Government for storing the records and to the States for updating the records). Cost control has thus been one of the driving forces behind efforts to keep the recordkeeping function decentralized so that duplication between the Federal and State Governments is minimal. For any alternative where all records are not maintained in a central repository, two other capabilities are necessary-an index to records not stored centrally, and a means to exchange or transfer records stored in 50 or more locations. There are several technical options here. For example, a national index could be maintained centrally at one location, such as Ident or NCIC in Washington, D.C. or NLETS in Phoenix, Ariz. Records could be exchanged via the NLETS or NCIC communication networks or both. Regional systems have also been proposed. However, OTA found little evidence to support the feasibility of regional systems. On the contrary, NLETS traffic logs indicate that criminal history traffic between the States does not conform to regional patterns. During the 1981 III pilot test, almost three-quarters of the hits on Florida records (matches between an inquiry and a record) originated from the Midwest and West. In addition, the 1979 OTA record quality research found that a high percentage (about 75 percent for Ident) of multi-State offenders had arrests in at least one noncontiguous State. A so-called ask-the-network system is also a technical possibility. In the ask-the-network approach, there would be no central index. Instead, each State would, in effect, poll any or all of the other 49 States plus the FBI when seeking CCH information. OTA found that a significant percentage of multi-State offenders (about 43 percent for Ident, again based on 1979 data) had arrests in three or more States. Considered together with the high percentage of multi-State arrests in noncontiguous States, it appears that all States and the FBI would have to be polled every time in order to make sure arrests were not missed, but the inquiryto-hit ratio would then be very low. Under similar circumstances, NLETS found that many States began to ignore the inquiries. Also, the FBI and various State criminal justice officials believe that an ask-the-network approach would not be cost effective, and would be harder to secure against unauthorized access. Nonetheless, ask-the-network systems are used successfully in the defense intelligence community and in the private sector, and their potential use in a national CCH system is an area of possible further research. Improving Response Tim e The operating experience of the Ident AIDS program and several State identification bureaus has documented that a much shorter turnaround time is possible with automated systems than with manual. The JPL study of AIDS concluded that full automation could reduce the overall Ident processing time for fingerprint checks from about 36 workdays to about 3 hours. Further improvements could result from the use of high quality facsimile

PAGE 20

10 l An Assessment of Alternatives for a National Computerized Criminal History Sy. stf?m electronic transmission. For example, New York State already makes relatively extensive use of this technology. New York responds to fingerprint inquiries submitted via facsimile within an average of 1 hour and 50 minutes, and within 3 hours 90 percent of the time. The response times for computerized criminal history record checks could be even faster. In theory, the response time for a national CCH repository would be measured in seconds. Indeed, as of April 1982, NCIC/CCH processing time was averaging less than onehalf second per inquiry, with very few inquiries taking more than 5 seconds. The III pilot test has demonstrated that even for a national index alternative, response times of less than an hour are possible. During a February-March 1982 test, response time was less than 1 hour 96 percent of the time, less than 5 minutes 76 percent of the time, and less than 1 minute 48 percent of the time. Thus, it appears that the III response time could approach the response time achieved by individual States with online CCH files, which is frequently in the range of 5 to 20 seconds. Response times for States with manual files would be considerably longer. Improving Record Qualit y While computerization can improve the response time of fingerprint and criminal record checks, improvements in record quality are more difficult to achieve. This is because high record quality depends on timely and accurate submissions from a large number of criminal justice agencies. Court disposition reporting appears to be a significant problem in many States. Available evidence indicates that strengthening State and local criminal history systems and court disposition reporting systems is a prerequisite to further improving CCH record quality, regardless of the national CCH system structure. Particularly important are efforts to upgrade court administration, establish standardized (and perhaps even codified) court reporting procedures, improve the coordination between judicial and other criminal justice agencies (especially law enforcement) responsible for timely record update actions, strengthen field audits of reporting procedures and record quality, and increase funding and technical assistance to implement computerbased systems where appropriate. Shifting Preferences on System Structur e An OTA survey of State repository officials found that, as of 1979, officials from 24 States out of 42 responding preferred the national index alternative, known as III. Officials from 11 States preferred the single-State/multiState alternative. Since that time, many other Federal and State officials have shifted their support to III. The NCICs Advisory Policy Board, NLETS Board of Directors, and SEARCH Group, Inc., have all endorsed III which, if fully implemented, would mean that all State records would be maintained by the States themselves. Only Federal offender records and an identification index would be maintained at the national level. In a 1982 OTA followup survey, officials from about two-thirds of the States indicated a clear preference for the III concept, with officials from most of the other States either actively considering III or seeking further information on which to base a decision. However, many States, even some of those strongly supporting III, noted a variety of implementation problems which might preclude their participation, in some cases for years. Many of these officials also support the concept of a National Fingerprint File (NFF), considered to be an integral part of III. The NFF would be limited to fingerprint cards and related personal descriptors on each criminal offender. The NFF would contain no arrest or disposition data. It would perform the technical fingerprint search to establish positive identification or nonidentification based on fingerprint cards received from State identification bureaus or Federal agencies. It would also assign FBI identification numbers, and could enter identification data into III. The NFF concept is predicated on single-source

PAGE 21

.-. Ch. 1Summary submission policies. That is, only one agency per State would be authorized to submit fingerprint cards. Submission of only one fingerprint card per subject per State would be permitted. OTA surveyed the States with respect to single-source fingerprint card submission and found that, as of August 1982, 18 States had implemented single-source submission (compared with 17 in a September 1981 FBI survey) and four more had scheduled a late 1982 implementation, for a total of 22 States. Officials from about one-third of the other States indicated that implementing single-source submission could be difficult due to a potential work overload, staff and funding shortages, local agency resistance, and/or privacy concerns. Possible Impacts of a National CCH Syste m Criminal Justice Proces s To the extent that a national CCH system provides information that is more complete, timely, and verifiable (based on positive identification) than is presently available, the system would improve the functioning of the criminal justice process. The most significant improvements are likely to be in the areas of criminal investigations, police booking and intake, pretrial release and bail decisions, and presentence investigation reports. For example, after an arrest, police make or participate in decisions concerning whether to release or how long to hold the suspect, whether to fingerprint, and the level of charges to be placed. Each of these decisions clearly affects the creation of a criminal history record, and conversely, criminal history records (and thus a national CCH system) may potentially influence these decisions. Since postarrest police decisions often must be made quickly, a national CCH system could make criminal history records more readily available, thus increasing their use. The impact of a national CCH system could be particularly significant in pretrial release and bail decisions, which typically must be made within 36 to 72 hours after arrest. If accurate and complete, CCH records could help prosecutors and judges to better balance the need to protect the public from harm by defendants out on bail, versus the need to protect the constitutional rights of defendants. Many States have laws or rules requiring judges to consider prior convictions in determining pretrial release conditions. It is important, however, that CCH records be complete and accurate. In Tatum v. Rogers (1979), a U.S. district court found a violation of constitutional (sixth, eighth, and 14th amendment) rights when arrest information without otherwise available disposition information was used in setting bail. Criminal history information is also used in the preparation of presentence investigation reports. These are used by judges in arriving at a sentence suited to offenders, and are subsequently used by the courts and corrections departments in assigning offenders to appropriate institutions. Problems that arise in the preparation of presentencing reports include incomplete disposition data and insufficient resources (time and money) for verification. It would appear that a national CCH system would be advantageous if based on accurate and complete records that could be obtained quickly and easily. A national CCH system could also affect other aspects of the criminal justice process. For example, criminal history records are very important to specialized programs (e. g., prior felon, career crime, and violent felon programs) that assign police investigators and special prosecutors to individuals who have prior felony convictions. Also, an arrestees criminal history record can affect the prosecutors decisions concerning whether to bring or drop charges, the level and number of charges, and

PAGE 22

12 l An Assessment of Alternatives for a National Computerized Criminal History System whether to negotiate at trial for lower charges through plea bargaining. An offenders criminal history is also an important factor in determining initial correctional custody rating (level of supervision needed) and institutional placement (e.g., maximum, medium, or minimum security), and is one of many factors considered in parole decisions. Employment and Licensur e To the extent that a national CCH system is accessible for noncriminal justice purposes, the system would be very likely to have a significant impact on a large number of employment and licensing decisions. Criminal history information is used in employment and licensing decisions to protect the public or the employer from harm. Criminal records may be used to screen individuals out of positions where they might easily cause harm to other citizens or coworkers or present an excessive risk to the protection of valuable assets (e.g., money, securities, precious jewelry, and other property). However, limiting job opportunities on the basis of a criminal record in effect involves an additional punishment for crime, that is, a civil disability, in addition to the punishment administered by the court. This civil disability may in turn hinder the rehabilitation of offenders and prevent them from becoming useful and productive members of society, even if they want to do so and are otherwise capable. Former offenders who cannot find suitable employment may become dependent on public welfare or return to crime. Federal and State legislatures must balance these considerations when requiring criminal history checks or character evaluations (which frequently include record checks) for literally millions of public sector jobs or publicly licensed private sector jobs. The private sector also frequently seeks criminal history information in making employment decisions. The impact of a national CCH system for noncriminal justice use is complicated by several factors. First, States (as well as the Federal Government) vary widely in their noncriminal justice access and dissemination policies. As noted earlier, a significant portion of State and Federal criminal history record repository use is for noncriminal justice purposes. Thus, without some kind of national standards on access, a national CCH system is likely to be heavily used for nonc riminal justice purposes and in ways that heighten the already existing conflicts-of-law among and between various States and the Federal Government. Second, noncrimin al justice use is even more sensitive to record quality than is criminal justice use. There is no doubt that the use of criminal history information affects employment and licensing decisions. The results of research, case studies of employers, surveys of employer attitudes, as well as the experience of Federal and State parole officers, all suggest that any formal contact between an individual and the criminal justice process is likely to influence an employers decisions on job applicants. A record of arrest and conviction will have the greatest influence, but even a record of arrest and acquittal will frequently work to the disadvantage of the applicant. This problem is aggravated because criminal history records are designed for use by those who are familiar with the criminal justice process and who understand the limitations of a record. At best, a criminal history record provides a snapshot or series of snapshots of a persons contact with the crimin al justice process at various points in time. Much of the contextual and background information necessary to properly interpret the record is not included. A record is more likely to be misinterpreted when used by someone outside the criminal justice system, particularly when a significant percentage of criminal history records contain inaccurate, incomplete, or ambiguous information. Third, there is considerable disagreement over the extent to which criminal history records can predict future employment behavior, except in particular cases such as repeat violent offenders. Other factors such as education, prior work experience, length of time in

PAGE 23

Ch. 1Summary l 1 3 the community, and personal references may be more predictive. On the other hand, the high recidivism rates suggest that once a person is arrested or convicted, he or she is much more likely to be convicted of a subsequent crime within a few years than those without a prior criminal record. Whether or not this is relevant to or predictive of employment behavior is a matter of debate. States such as New York have required by statute that any agency seeking criminal history information establish a strong relationship between the nature of the job and specific kinds of criminal offenses. Florida, with its open records policy, is at the other extreme. Fourth, criminal history records involve a sizable proportion of all persons in the labor force. After a careful review of existing research, OTA estimated that as of 1979 about 36 million living U.S. citizens had criminal history records held by Federal, State, and/or local repositories. Of these, OTA estimated that about 26 million persons were in the labor force (representing, conservatively, 28 to 30 percent of the total labor force), and thus were potentially exposed to employment disqualifications because of an arrest record. Of the 36 million, OTA estimated that about 35 percent had no arrests for serious crime and one arrest for a minor crime, and about 24 percent had more than one minor arrest but no major arrests. The remaining 41 percent (roughly 15 million persons) had at least one arrest for a serious crime. These aspects of noncriminal justice use warrant congressional consideration in formulating policy on any national CCH system. Minority Group s Some minority groups account for a disproportionate percentage of arrest records. For example, various studies have estimated the percentage of blacks with arrest records as ranging from 30 percent nationwide to over 50 percent in certain cities such as Philadelphia. As of February 21, 1980, blacks accounted for about 29 percent of all records in the NCIC/CCH file, which is almost triple the percentage of blacks in the total U.S. population. Statistics of 1981 on felony arrests in California indicated that, at least in that State, blacks also accounted for a disproportionately high percentage of law enforcement releases and complaints denied (37.7 and 38.7 percent, respectively, compared with 30.6 percent of felony arrests).* Releases and complaints denied may occur for a variety of reasons, such as insufficient evidence, refusal of the victim to prosecute, lack of probable cause, unavailable witness, or illegal search. As discussed earlier, a criminal arrest record, even without convictions, can have an adverse effect on employment and licensing applicants. Indeed, the courts have found that a policy of refusing employment to blacks with an arrest record without convictions had a racially discriminatory impact because blacks are arrested substantially more frequently than whites in proportion to their numbers (see Gregory v. Litton Systems, 1970). Similar judicial reasoning has been extended to black applicants refused employment due to criminal convictions where the offense does not significantly bear upon the particular job requirements (see Green v. Missouri Pacific RR, 1975). In this context, any discriminatory impacts from the use of national CCH information would depend on whether and under what conditions nonc rimina1 justice access is permitted. The potential for disc riminatory impacts could be minimized if records or index entries based on arrest-only information, as well as information on arrests not leading to conviction, were actively sealed or otherwise effectively removed from the file, at least for noncriminal justice purposes. Some States, such as New York, do this for their own files, but many States do not. California has struck a middle ground. Felony arrests that result in detention only are retained in the California State criminal history record repository for 5 years, and *A law enforcement release occurs when police detain and arrest a person, obtain fingerprints, and report the arrest to the State record system, but subsequently release the person and do not present the case to the district attorney. A complaint is denied when the police arrest and present a person to the district attorney, but the district attorney decides not to prosecute the case.

PAGE 24

74 An Assessment of Alternatives for a National Computerized Criminal History System felony arrests that otherwise do not result in a conviction are retained for 7 years. Federalis m The balance of authority and power between Federal, State, and local governments has been a central issue in the debate over a national CCH system. Because of the decentralized nature of the U.S. criminal justice process and because the generation and use of criminal history information occurs mostly at the State and local levels of government, most States seek a primary role in any national CCH system. State governments have basic jurisdiction over law enforcement and criminal justice within their borders under their constitutionally reserved powers, and many have been reluctant to share this jurisdiction with the Federal Government, except with respect to Federal offenders. Most States have appreciated other kinds of support from the Federal Government, such as FBI fingerprint identification services and LEAA funding for State CCH system development, as long as this support was provided on a voluntary basis and the States retained control over the operation and use of their own criminal history record systems. The Federal Government has a legitimate interest in: 1) the enforcement of Federal criminal law, 2) the prosecution of Federal offenders, whether intrastate or interstate, and 3) assisting with the apprehension of interstate and international criminal offenders who cross State and/or national borders. To the extent that crime is perceived as a national problem deserving national attention, the Federal Government also has a defined role in the provision of voluntary support to State and local law enforcement and criminal justice activities. Many of the proposed alternatives for a national CCH system encounter difficulties resulting from the historic constitutional division of powers and duties in the U.S. Federal system. Since the standards of the States vary so widely (e.g., with respect to dissemination, record quality, and sealing and purging of criminal history records), any national standards for a CCH system could easily conflict with those of at least some States. From the perspective of many States, a national CCH system like III would have a minimal impact on Federal-State relationships assuming that it retained State policy control over the CCH records, avoided any significant conflict with State laws and practices on the collection and use of criminal history information, and kept State costs at an affordable level. Nevertheless, III (or any other national CCH system) would have interstate and national as well as intrastate impacts. A strong argument can be made that, regardless of the specific system structure, the Federal Government has the responsibility and authority to establish some kind of system standards. From a legal standpoint, Federal action could be based on: 1) the criminal record information needs of Federal agencies as established by various Federal statutes and Executive orders (e.g., Executive Order Nos. 10450, 12065, and 10865); 2) the implementation of Federal regulations for State and local criminal justice information systems that have used Federal funding (title 28, Code of Federal Regulations, pt. 20); 3) the interstate commerce clause of the U.S. Constitution; and/or 4) the constitutional provisions (including the first, fourth, fifth, sixth, ninth, and 14th amendments) guaranteeing individual rights of privacy and due process. cos t Throughout the 1970s, it was Federal Government policy to support the development of State CCH systems and the implementation of the Federal regulations. From 1970 to 1981, LEAA provided a cumulative total of about $207 million in categorical grants to the States for comprehensive data systems and statistical programs. About $39 million was for 145 CCH-related grants awarded to 35 different States. These grants peaked in 1976 and ended in 1981. In addition, some portion of LEAA block grants to the States was used for criminal justice information systems. This avenue of Federal support has also been phased out.

PAGE 25

Ch. 1Summary l 1 5 Thus, at present the States and localities would have to bear most of the cost of any national CCH system. The difficulty of finding new money or cutting back other expenses could discourage State participation. Financing could be particularly difficult for States whose criminal history record systems are not yet well developed, whose need for a national CCH system is not perceived to be great, and whose ability to pay is limited. OTA did not independently estimate the cost of a national CCH system. The Federal share would presumably include some portion of the cost of Ident (which totaled about $58.7 million in fiscal year 1980 and whose full automation has been estimated at $50 million by JPL) and NCIC ($6.1 million in fiscal year 1981), plus the costs of Federal agencies participating in the system. The actual Federal share would depend on the specific alternative implemented, and whether or not further Federal support were provided to the States. LEAA grants made a significant contribution to the relatively rapid development of State CCH systems during the last 12 years. OTA research has identified the following three areas as possible priorities for further funding: 1) improving court disposition reporting systems on a nationwide basis; 2) upgrading criminal history record systems in the States that are operating manually or assisting those in the process of automating their name index and/or file; and 3) improving procedures in all States where necessary to assure the accuracy and completeness of criminal history information, to conduct audits of local users, to maintain and periodically review transaction logs, and to train employees and users. Surveillance Potentia l The flagging of criminal records is a common monitoring or surveillance practice and an accepted law enforcement tool. Placing a flag on a file helps law enforcement personnel to keep track of the location and activity of a suspect whenever there is a police contact. At the State level, both manual and automated files are used for flagging. Although this practice differs from State to State, the most frequent application seems to be for parole violators and wanted persons. At the Federal level, Ident records are usually flagged by using a wanted notice for persons with an outstanding arrest warrant or a flash notice for persons placed on probation or parole. With respect to NCIC, since hot files are flags by definition, all wanted or missing persons and stolen property records included in NCIC represent flags to law enforcement and criminal justice users. Concern has been expressed about the possible use of a national CCH system by Federal agenciesand particularly the FBIfor monitoring or surveillance of the lawful activities of individuals or organizations. To understand this concern, one must remember that the debate over a national CCH system began in the late 1960s and early 1970s, a time when the FBI was engaged in domestic political intelligence and surveillance activities with respect to, for example, civil rights and anti-Vietnam War leaders and groups. Also during the early 1970s, the FBI made very limited use of NCIC for intelligence purposes which, although strictly law enforcement in nature, had not been authorized by Congress. Since that time, the FBI has rejected all requests or proposals for intelligence use of NCIC.* During the course of the OTA study, FBI officials have repeatedly stated to Congress and to OTA that they will not permit Ident or NCIC to be used for unauthorized purposes of any kind. FBI officials believe that a national CCH would not have any significant surveillance potential and would represent little, if any, danger to law-abiding citizens. Strong and independent policy control over a national CCH system and tight restrictions on noncriminal justice access, coupled with outside audit and explicit statutory guidelines for operations, would help protect against the possibility-however remotethat As of s ep t em b er 1982, the Department of Justice and the FBI had approved but not yet implemented a U.S. Secret Service proposal to establish an NCIC file on persons judged to represent a potential threat to protectees, including the President. This could involve the use of NCIC to gather intelligence data on or track individuals not formally charged with a criminal offense.

PAGE 26

16 l An Assessment of Alternatives for a Nationa/ Computerized Crimina/ History System a national CCH system could be used at some point in the future in violation of first amendment or other constitutional rights. In comments to OTA, various criminal justice officials have suggested a statutory prohibition on intelligence use of III or any other national CCH system. On the other hand, some State officials have noted that there may be legitimate intelligence and surveillance applications, and that these possibilities should not be abandoned solely because of their sensitivity. Message Switchin g As noted earlier, unless all criminal history records were stored in one place (e.g., a national CCH repository) a national CCH system would require some electronic means to transfer records (and inquiries for such records) among and between the various State and Federal repositories and participating agencies. The transfer or switching of messages from one State to another through the NCIC computer has been a point of controversy over the last 12 years. Some message switching alternatives have raised questions about the impact on Federal-State relations and the potential for monitoring and surveillance. For example, in 1973, the FBI proposed to have NCIC assume all law enforcement message switching (not just NCIC/CCH traffic), including messages sent over NLETS. As a result, Congress has denied the FBI authority to perform message switching, defined as the technique of receiving a message, storing it in a computer until the proper line is available, and then retransmitting, with no direct connection between the incoming and outgoing lines. Department of Justice Appropriation Authorization Act of 1980. More specifically, the Department of Justice (DOJ) was prohibited, without explicit approval of the House and Senate Judiciary Committees of Congress, from utilizing equipment to create a message switching system linking State and local law enforcement data banks through equipment under the control of DOJ or the FBI. In addition, congressional approvals in 1979 and 1980 of the FBIs requests to upgrade NCIC computer technology were conditioned on the FBIs commitment not to use such technology for message switching. There are several message switching alternatives for III. First, inquiries could be switched via NCIC, with records returned via the NLETS message switching system. This approach was used in the III pilot and Phase 1 tests. The routing of inquiries through NCIC has been termed automatic inquiry referral and is a form of partial message switching. Second, both inquiries and records could be switched via NCIC. Third, both inquiries and records could be switched via NLETS. Fourth, records could be switched via NLETS and inquiries via NCIC or NLETS. Fifth, the use of NCIC or NLETS could be optional for switching of both inquiries and records. OTA has not evaluated these alternatives in detail, although all appear to be technically feasible. In making a complete evaluation, message formats and purpose codes, costs to the States and the Federal Government, response time, and message privacy and security all need to be considered. In any event, any DOJ or FBI message switching role in a fully operational III (or other national CCH system) would probably require congressional approval. Ibid. Congressional Policy Consideration s As noted earlier, the emerging consensus to ensure that beneficial impacts are maxiamong Federal and State law enforcement and mized and adverse impacts are controlled or criminal history record repository officials supminimized. ports the national index concept known as III. However, full implementation of III (or any Policy Contro l other national CCH system) raises a number Considerable debate has focused on which of issues that warrant congressional attention agency or organization should have direct

PAGE 27

Ch. 1Summary l 1 7 policy control over a national CCH system. Suggestions include a consortium of States, a broadened and strengthened NCICS Advisory Policy Board, an independent board, and/ or the FBI. For example, a broadened and strengthened Advisory Policy Board could include greater representation from the prosecutorial, judicial, correctional, and public defender sectors of the criminal justice community than at present, and could include an advise and consent role, at least with respect to State and local participation in a national system. There are many other possibilities, but the key issue is how to devise a mechanism that will effectively represent the interests of the diverse users of a national system, and afford them a strong and possibly controlling policy role. File Size and Content Under the 111 concept, the national index would include only names and identifying information (e.g., height, weight, social security number, and State and Federal criminal identification numbers). Proposals have been made to limit the index to entries on violent or very serious offenders, that is, for crimes included in the FBI Crime Index. However, this would exclude entries for drug, weapons, drunk driving, and other offenses generally considered to be serious but not included in the FBI Crime Index. At the other extreme, a totally unrestricted index could include entries on as many as 36 million persons. Other national index issues include the need for policies on limited retention periods for some entries and on the handling of juvenile offender records. Record Qualit y With a national index, the FBI would no longer maintain non-Federal records, and the problems of record quality in Ident and NCIC/CCH would be reduced. However, the quality of records maintained by the States, as well as the quality of any index based on those records, would still be a matter of concern. Record quality could be strengthened by tightening the disposition reporting requirements and/or requiring confirmation of records lacking disposition data with the originating agency prior to any dissemination. In the opinion of some, the latter requirement would be costly and impractical. The progress made by many States in recent years indicates that improved disposition reporting is possible, but continued improvement would require a significant further commitment measured in manpower, dollars, and system improvements at the State and local levels. Noncriminal Justice Acces s Significant noncriminal justice use of Federal and State criminal history record systems, coupled with widely varying State statutes defining authorized users and State policies on sealing and purging, has generated concern about control of access to criminal history records. Noncriminal justice access to a national index could be prohibited, although this would conflict with many Federal and State laws. Noncriminal justice access could be permitted, but only under stronger Federal guidelines than presently exist. A dual index could be established, one for criminal justice use and a second for noncriminal justice use, perhaps with the latter based on disposition or conviction information only. Even under the status quo, access to a national index would require complicated safeguards (which are technically feasible with a computer-based system) to be consistent with the wide variety of existing State laws and regulations, and would require some means to resolve conflicts between State laws, and between Federal statutes and Executive orders and State laws. Oversight and Audi t The purposes of new oversight mechanisms would be to help assure Congress, the public, and others that a national index (or any other national CCH system) is operating within the boundaries of law and regulation, and to help identify any problems that may emerge. Oversight is closely linked to system audit. Several possibilities have been suggested. First, Congress could require an annual management report on the operation of a national CCH system. Second, Congress could require periodic

PAGE 28

18 l An Assessment of Alternatives for a National Computerized Criminal History system audits of Federal and State CCH files to help ensure compliance with whatever system standards may be established. To keep costs down, the audits would presumably be conducted by sampling Federal and State files on a rotating and perhaps unannounced schedule. Any Federal audit authority, whether by GAO or some other body, would appear to require new Federal legislation and/or regulations. Public Participation NCICS APB is the only direct avenue of public participation in the governance of the existing NCIC/CCH system. However, at present APB does not include representation from the general public or from public defenders. Public defenders feel strongly that they should be represented on any policy board established for a national CCH system and that defense interests should have access to that system. The experience of Alameda County, Calif., where public defenders are considered to be part of the criminal justice community, has been that public participation in oversight can help ensure accountability of criminal justice record systems and can be beneficial in terms of system performance. Comprehensive Legislatio n Legislation represents one of the strongest measures to provide Federal direction and ensure accountability and control. It could provide explicit authority for a national index or other national CCH system, and include statutory guidelines for its operation and use. In addition to the areas listed above, legislation could establish access, review, and challenge procedures; criminal penalties; privacy standards; funding for computer-based user audits and disposition monitoring procedures; and uniform crime codes and criminal history record formats. Legislation could also cover areas discussed earlier such as intelligence use, message switching, and funding for development of court disposition reporting and State criminal history record systems. III Development Pla n In order to develop important additional data from the III test now under way, Congress may wish to consider whether the plan should be revised so that: 1) some or all of the participating States can be tested with no NCIC message switching as well as with partial message switching (known as automatic inquiry referral); and 2) record quality research can be conducted. AIDS/CCH Consolidatio n At present, the Ident/AIDS and NCIC/CCH files duplicate each other to a significant and growing extent. Any AIDS/CCH consolidation is likely to have a significant impact on the cost of FBI crimin al history and identification services and could be an integral part of a national CCH system. Congress may wish to request the preparation of several alternative consolidation plans, including the possible creation of a new National Criminal Information and Identification Division of the FBI which would combine Ident, NCIC, and related activities. Congress may also wish to examine the pros and cons of shifting management of a national CCH system to a new bureau within DOJ or elsewhere.

PAGE 29

Chapter 2 Nature and Origins of Criminal History Record s

PAGE 30

Contents Page Chapter Summary . . . . . . . . . . . 21 Nature of Criminal History Records . . . . . . . 21 Origins of Criminal History Records . . . . . . . 26 1. 2. 3 0 4. 5 0 Overview Facsimile Facsimile S ummary Facsimile o f o f of FIGURES Criminal History Records System Manual Ident Rap Sheet . . CCH Record and Explanation Sample Record Response .. ... . . ... of AIDS Record. . . . . . Bertillon Equipment . . . . . . . . . . . . . . . Response CCH . . . . . . . . . . . . Page 23 24 25 26 27 I

PAGE 31

. Chapter 2 Nature and Origins of Criminal History Records Chapte r Until the 1850s, criminal history records in the United States consisted largely of informal notes maintained by local police officers. Since that time, however, such records have become more formalized, centralized, widely used, and technology-based. Originally, criminal history records were known as rap sheets. The information they contained about subject individuals ranged from personal identifiers (e.g., height, weight, eye color, and/or identification numbers) to arrests (date, jurisdiction, and charges). Information contained in criminal history records also may include court disposition of charges, sentencing, incarceration, and the like. The actual content of criminal history files varies widely. The applicable laws, traditions, and management practices in a particular jurisdictionas well as disparities in the completeness, timeliness, and/or accuracy of criminal history information submitted to State and Federal repositories all contribute to variation in record content. Summar y These records are still maintained in local police departments, as well as in central repositories established by the Federal Government and 49 of the 50 States. Since 1924, the Federal Bureau of Investigation (FBI) has maintained a national criminal history record repository in its Identification Division (known as Ident). While police were the earliest users of criminal history records, such records now are used to varying degrees at all stages of the criminal justice process by prosecutors, public defenders, judges, probation officials, and the like. The technology of criminal records has evolved from paper-and-pencil, to manual filing, to computerized information systems; and from rogues galleries and daguerreotypes, to fingerprinting, to automated fingerprint classification and search systems. Nature of Criminal History Record s Criminal history records exist on several levels. Criminal history information generally originates with criminal justice agencies directly involved with the processing of specific persons. For example, the police maintain records (known as blotters) on individuals arrested and the date, time, charges, and related information. Fingerprints are normally taken when a suspect is booked in order to establish positive identification. Police blotters are usually compiled chronologically. Likewise, the courts maintain records of the results of judicial proceedings compiled chronologically by case or docket number, including the disposition and, if applicable, the charges of conviction and sentence imposed. Correctional agencies maintain records on individuals incarcerated, including intake information, facility assignment, and the like. This study does not focus on the original records of law enforcement, judicial, correctional, and other criminal justice agencies (e.g., prosecutional, parole/probation); rather it concentrates on the consolidation of this information in criminal history records organized by name 21 J 1, : { J ~ 4

PAGE 32

22 An Assessment of Alternatives for a National Computerized Crimina/ History System (and/or identification number) and maintained at the State and Federal levels. As will be discussed later, the Federal Government and all but one State maintain central criminal history record repositories. These repositories include both record information and related fingerprint data that are used (along with other personal descriptors) to make positive identification. The various levels of criminal history records are illustrated in figure 1. These criminal history records originally were known as rap sheets and included primarily the subjects personal identifiers and arrest information. However, a broader definition of criminal history records has emerged. As currently defined in the Code of Federal Regulations, criminal history records include information collected by criminal justice agencies on individuals consisting of identifiable descriptions and notations of arrest, detentions, indictments, information, or other formal criminal charges, and any disposition arising therefrom, sentencing, correctional supervision, and release. Thus, the definition of criminal history records (although still frequently referred to as rap sheets) used in this study includes information concerning an individuals arrests for violation of criminal laws and the disposition of charges (including charges dropped, dismissed, or of which the subject is acquitted). For convicted individuals, crimin al history records include information on sentencing, correctional supervision, and release. They also note the arresting jurisdiction and, where applicable, the correctional institution. Finally, these records include local, State, and/or Federal criminal identification numbers assigned to the individual involved, along with other personal identifiers (e.g., height, weight, and eye color). The records of repeat offenders contain criminal history information for each violation. Title 28 Code of Federal Regulations, pt. 20, subpt. A. The actual content of criminal history files varies widely, depending on the applicable laws, tradition, and management practices in a particular jurisdiction. For example, some States seal or purge information concerning arrests that do not result in a conviction. z Other States maintain such nonconviction information as a permanent part of the criminal history record. The many variations in State law that affect the content of criminal history records are reviewed in chapter 6. Also, as discussed in chapters 8 and 9, disposition reporting levels vary widely among the States. At the Federal level, the FBI has developed a standard format for criminal history records in Ident and in the computerized criminal history (CCH) file of the National Crime Information Center (NCIC). The Ident and CCH criminal history record formats are illustrated in figures 2 and 3. The I dent manual records currently are being converted to an automated format called the Automated Identification Division System, illustrated in figure 4. The content of the Ident and CCH records also varies widely, largely due to disparities in the completeness, timeliness, and/or accuracy of criminal history information provided to the FBI by Federal, State, and local criminal justice agencies. The South Carolina statute is an example of an especially strong automatic purging procedure. The South Carolina Code (S. Carolina Code sec. 17-1-40) requires agencies to automatically purge arrest records upon notification that charges were dropped or that the subject was acquitted. SEARCH Group, Inc. Sealing and Purging of Criminal History Record Information, technical report No. 27, Sacramento, Calif., April 1981, p. 13. As defined in 28 CFR 20.3(k), nonconviction information means arrest information without disposition if an interval of one year has elapsed from the date of arrest and no active prosecution of the charge is pending; or information disclosing that the police have elected not to refer a matter to a prosecutor, or that a prosecutor has elected not to commence criminal pr oceedings, or that pr oceedings have been indefinitely postponed, as well as all acquittals and all dismissals. See SEARCH, Sealing and Purging, op. cit., pp. 13-16, for discussion of variations in State law on sealing and purging.

PAGE 33

Ch. 2Nature and Origins of Criminal History Records 23 Figure 1 .Overview of Criminal History Records System Federal Other Other level Federal systems Federal users and (e.g., TECS a ) interstate users State level Local level Other In-State State systems users (e.g., NLETS d ) 49 State identification bureaus, of which 27 have a CCH e file and 7 have an automated name Other Local Local systems users (e.g., REGIS f ) Street level g index Type of { Date Disposition Initial assignment criminal Name Charges Release date history Arrest Sentence information Charges generated Fingerprints (illustrative) NOTES: a TECS = Treasury Enforcement Communication System. bldent/AIDS = Manual and Automated Identification Division System records (including fingerprints) maintained by the FBIs Identification Division. cNcIc/ccH = FB I' S National Crime Information Center computerized criminal history records. dNLETS = National Law Enforcement Telecommunications SyStem. eCCH = Computerized Criminal History. f REGIS = Regional governmental information systems which frequently transmit criminal history information (as in the St. Louis, Missouri region). gMay also include prosecution files and records maintained by pretrial diversion and probationparole agencies. SOURCE Off Ice of Technology Assessment adapted from Sarwar A Kashmerl, REJIS-A New Concept for Regional Criminal Justice Agencies !n LEAA, Pro ceedings of the Second International Symposium on Criminal Justice Information and Statistics Systems, Washington, D C 1974, p 380

PAGE 34

24 l An Assessment of Alternatives for a National Computerized Criminal History System Figure 2.FacsimiIe of Manual Ident Rap Sheet UNITED STATES DEPARTMENT OF JUSTICE 3-15-79 125 FEDERAL BUREAU OF INVESTIGATION IDENTIFICATION DIVISION WASHINGTON, D.C. 20531 Use of the following FBI record, NUMBER 000 000 A IS REGULATED BY LAW. It is furnished FOR OFFICIAL USE ONLY and should ONLY BE USED FOR PURPOSE REQUESTED. When further explanation of arrest charge or disposition is needed, communicate directly with the agency that contributed the fingerprint. PD Peoria IL PD Daytona Beach FL s o Ore Con IL Rec & Class Ctr Joliet IL PD Peoria IL Rec & Class Ctr Joliet IL (Source : Provided July 1979 Name And Number John Lee Doe 34653 John Doe ID104200 SID FL4261893 John L Doe John Lee Doe C61778 John L Doe 34653 John Doe C61778 Arrested OR Received 8-12-74 4-21-75 5-2-76 1-76 8-3-78 9-29-78 Division, Federal I CHARGE OMVWI shopliftin g burglar y 2 count s burglar y the f t the f t SUBJECTS CRIMINAL ALSO RECORDED IN FILE a 1 Bureau of investigation DISPOSiTION charge dism 12-18-74 4-29-7 5 Sent sus p 30 day s 6-10-7 6 1 yr IL Dept o f Correction s 1 y r 9-1-78 1 yr 6 mos to 4 yr. 6 mos guilty IL Dept of Corrction s 1 yr 6 mos to 4 yr 6 mos HISTORY NCIC-CCH SOURCE Provided by Identification Division, Federal Bureau of Investigation, July 1979

PAGE 35

Ch. 2Nature and Origins of Criminal History Records l 2 5 .. ----.. Figure 3. Facsimile of CCH Record and Explanation Sample Response CCH Summary Record (QH) 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 4001HEADERXXXXO1234 MD10126OO NCIC S UMMARY MULTIPLE STATE FBI/835172Kll 03/01/77 EH DOE, JOHN M W NH DoB/120345 HGT/511 WGT/165 EYE/BRO HAI/BLK SKN/LBR SMT/TAT L ARM SOC/375125249 MNU/AS-375125249 FPC/121011C0141159TTCI13 Comment / E p IL E PTI S TAT Is ARROW THRU HEART ADDITIONAL IDENTIFIERS 4002 HEADE RXXXX01234 MD10126OO SID/MD34567812 DOB/0ll147 031147 031149 SMT/CRIP L HND CON LENSES DENT UP LO SOC/421381928 382411221 120551394 MNU/MC-1428316 VA-C1234567 AKA/DOE, JOHN J/HENRY, JOHN/JOHN, HENRY DLU/602876 TOTAL ARRESTS 2 4003HEADE RXXXX01234 MD10126OO CHARGES CONVICTIONS OFFENSE 1 1 ASSAULT 1 1 DANGEROUS DRUGS 1 0 ATTEMPT SEX OFFENSES LAST ARREST STATUS (INCLUDED ABOVE) 012074 MDBPDOOOO BALTIMORE CITY PD MD COURT STATUS (I N C LUDED ABOVE)32 MD SUP CT NO 1 BALTIMORE 400HEADE RXXXX01234 MD10126OO 01 3603 ATTEMPTED HOMOSEXUAL ACT W ITH Boy-CONVICTED DISP DATE/020274 CONFINE/6M 02 1314 AGGRAV ASSLT-GUN--CONVICTED DISP DATE/020274 CONFINE/6M cc APPEAL DATE030274 SUPP COURT STATUS01 ACQUITTED CUSTODY STATUS 4L05HEADE RXXX X 0123 4 MD10126OO MDO02035C HOUSE OF CORRECTION JESSUP MD 030374 RECEIVED MDO02035C HOUSE OF CORRECTION JESSUP MD 090374 DISCHARGED END BASED ON NAME CHECK 4b SOURCE Provided by the National Crime lnformation Center, Federal Bureau of investigation, October 1978

PAGE 36

26 l An Assessment of Alternatives for a National Computerized Criminal History System Figure 4. Facsimile of AIDS Record UNITED STATES DEPARTMENT OF JUSTICE FEDERAL BUREAU OF INVESTIGATION IDENTIFICATION DIVISION WASHINGTON, D.C. 20537 Use of the following FBI record, NUMBER 000 000 N 1 is REGULATED BY LAW. It Is furnished for OFFICIAL USE ONLY and should ONLY BE USED FOR PURPOSE REQUESTED. When further explanation of arrest charge or disposition is needed, communicate directlv with the aqencv that contributed the fingerprints. Contributor of Identificr (0RI) Name Case Numher (OCA) NATIONAL CRIME INFORMATION I LO72O70 0 PD PEORIA I L 3465 3 11064010 0 PO DAYTONA BEACH, F L 1D-104 200 IL071 000 0 SO OREGON I L IL0990 15 c RECLASS CTR JOL I ET C61 770 ILO7207 0 0 PD PEORIA,I L 3465 3 IL 09901 5C REC LCLASS CTR JOL I ET C6177 8 Subject : Name State Number (SID) CENTER FGPT DOE JOHN LE E DOE JOH N FL426189 3 DOE JOHN L DOE JOHN LE E DOE JOHN L DOE JOHN 4C CLASS 08/ 1 2/7 4 04/2 1 /7 5 O 5/0 2/7 6 0 7/0 1/7 6 O 8/0 3/7 8 09/2 9/7 0 .C Charge D D ISPO sition C-O MVW I D-CHARGE D I S M I SSE DI 2 / 1 8 /7 C SHOPL I F T I NG DCON V I C T ED -04 / 29 / 75 SENT SUSP-30 D C-BURGLARY2 COUNTS D-CONV I C TE D 06 / 10 / 76 CONFINEMENT1 Y IL DEPT OF CORRECTI0N S CBURGLARY D-CONVICTED CONFINEMENT 1Y C THEFT DCONVICTED09/01/7 S CONF I N EMENT -1 Y 6M-4 Y 6M GUILT Y IL DEPT OF CORRECTI0N S C-T HEFT DCON V I CTED CONF I N EM EN T -1 Y 6 M4 Y 6 M SOURCE. Provided by the Identificatlon Division Federal Bureau of Investlgation, July 1979 Origins of Criminal History Record s Until the mid-1850s, criminal history reccity police began to centralize recordspriords in the United States consisted largely of marily rap sheetsand establish more consisinformal, anecdotal notes maintained by local tent reporting formats. police officers. However, as the larger cities in the Northeast such as Boston and New At the same time, city police needed more York created formal police departments, the reliable ways to identify suspects and crimi-

PAGE 37

Ch. 2Nature and Origins of Criminal History Records l 2 7 nals. In the mid-19th century, the police first established rogues galleries and used daguerreotypes to identify criminals. In the 1880s, these methods were supplemented by the Bertillon system of identification, which used body measurements, standardized photographs, notations of skin color and markings, and thumbline impressions, as shown in figure 5. In the early 1900s, fingerprinting evolved as the most reliable identification system; it remains the basic underpinning g of criminal history records at all levels of government. Also in the late 19th and early 20th centuries, the increased mobility of criminals fostered by the railroad and automobile expanded the need for sharing of criminal history records among different jurisdictions. The National Chiefs of Police Union, later to become the International Association of Chiefs of Police (IACP), was among the first to support a more formal institutionalized means for exchanging records. Partly because the States had not yet developed a statewide role in criminal justice recordkeeping, IACP supported a national bureau of criminal identification, and even maintained a prototype bureau itself for several years. In 1923, at the urging of IACP among others, IACP bureau files, along with the Federal criminal identification files (maintained at the Federal prison in Leavenworth, Kans.), were transferred to the U.S. Department of Justice (DOJ). Congress created the Identification Division in 1924, and passed legislation in 1929 making Ident a permanent part of the FBI within DOJ. By the 1940s, most of the States had developed State bureaus of identification, and the basic framework for State and Federal repositories of criminal identification and arrest records had taken shape. By 1949, 32 States had established identification bureaus, and as of September 1981, 49 of the 50 States had a Figure 5. Bertillon Equipment Bertillon Equipment ( n A I The Bertillon System of Identification is now used by the principal cldes of the U-ited States and foreign countries. The descriptive pamphlet mailed upon request Eat. 1544 SHARP & SMITH lnc. 1904 92 Wabash Avenue (Two doors North of Washington Street) Chicago, Ill. SOURCE Reproduced from Identification Wanted Development of the American Criminal ldentification System 1893193, International Association of Chiefs of Pollee, Police Management and Operations Divisions, Gaithersburg, Md 1977

PAGE 38

28 An Assessment of Alternatives for a National Computerized Criminal History System State identification bureau established by law.5 While the early use of rap sheets was primarily by the police, by the 1960s and 1970s criminal history records were increasingly viewed as important at all stages of the criminal justice process. Thus, the needs of prosecutors, public defenders, judges, magistrates, Nevada is the one State without a State identification bureau. Based on September 1981 survey conducted by NCIC staff. See NCIC staff paper prepared for the Nov. 3-4, 1981, meeting of the NCIC Advisory Policy Board Subcommittee on the Interstate Identification Index, Topic #6, pp. 1-2. and probation and correctional officials were all considered along with those of police and investigatory officers.6 c For a detailed discussion of the criminal identification system, see Donald C. Dilworth, Identification Wanted: Development of the American Criminal Identification System 18931943, International Association of Chiefs of Police, Gaithersburg, Md., 1977. Also see Donald A. Marchand, et al., A History and Background Assessment of the National Crime Information Center and Computerized Criminal History Program, Bureau of Governmental Research and Service, University of South Carolina, June 1979, Section I, Law Enforcement Record-Keeping in the United States: 1850 s-1950 s, pp. 6-37.

PAGE 39

.. Chapter 3 Evolution to Computerized Criminal History Records

PAGE 40

Content s Chapter Summary . . . . . . . . . . . 31 Problems With Manual Files . . . . . . . . . 31 Concern Over Multi-State Offenders . . . . . . . 32 c riminal Justice Reform and Computers . . . . . . . 33 Birth of the National Crime Information Center . . . . . 34 Origins of the Computerized Criminal History Program.. . . . 35 TABLES Table No. l. Percent Multi-State Offenders unselected Criminal History Record Files . . . . . . . . . . . . 33 2. Initial Participating Agencies for the National Crime Information Center . . . . . . . . . . . . . 35

PAGE 41

Chapter 3 Evolution to Computerized Criminal History Records Chapte r Until the 1960s, criminal history records were maintained on paper in manual files. The evolution to computerized recordkeeping reflects several factors, including growing problems with manual files, concern over multiState offenders, a rising crime rate, efforts to reform the criminal justice process, and the availability of new technology. Manual files have grown so large and cumbersome that response time has become a serious problem. For example, the Federal Bureau of Investigation (FBI) Identification Divisions (Ident) file has grown from 2 million fingerprint cards in 1924 to about 175 million in 1981, with an average of 27,392 cards received each day. The typical Ident response time during 1981 was in the range of 27 workdays for all crimin al record inquiries (including fingerprint and name checks). Another justification for centralized, computerized files is the mobility of criminals. Computer-based information systems can help speed the exchange of criminal history records among the States. OTA verified that multiState offenders account for about 30 percent of the Ident criminal record file and 33 percent of the National Crime Information Center/ Computerized Criminal History (NCIC/CCH) file. With Federal offenders excluded, about 12 percent of the records in the NCIC/CCH file are for multi-State offenders. Summar y Several study commissions established in response to rising crime rates have emphasized the role of computers in the reform of the criminal justice process. One result was the enactment of the Omnibus Crime Control and Safe Streets Act of 1968, which established the Law Enforcement Assistance Administration (LEAA). One priority of LEAA was the development of CCH systems in the States. At the Federal level, the FBI had wanted for years to automate criminal records. The computer technology of the 1960s provided the first practical opportunity. Thus, 1963 marked the initial FBI use of computer technology to process individual criminal records for Federal offenders, and led eventually to the establishment of NCIC. This center began operations on January 27, 1967, to create automated files of wanted persons and stolen vehicles. Automating criminal history records proved to be much more difficult. Questions immediately arose about who should operate a CCH system and what information should be maintained at the Federal and State levels of government. In 1970, after Project SEARCH (System for Electronic Analysis and Retrieval of Criminal Histories) began development of a national CCH system with LEAA funding, the U.S. Attorney General assigned management responsibilities for CCH to the FBI rather than to LEAA, a joint LEAA/FBI entity, or a consortium of States. Problems With Manual File s Until the 1960s, criminal history records were maintained on paper, and were created and updated manually (typewriter or handwriting). While this approach appeared to be satisfactory for many years, the increasing number of records eventually created a severe 31

PAGE 42

32 An Assessment of Alternatives for a Nationa/ Computerized Crimina/ History System problem. Manual processing of paper records is extremely time-consuming and labor-intensive, thus costly. As staffing and funding limits were reached, delays in obtaining and updating records increased significantly. Idents experience exemplifies the immensity of the problem. Starting with about 2 million fingerprint cards in 1924, the number grew to about 76 million in 1943, and to 170 million in 1979. As of October 1981, Ident held about 175 million fingerprint cards representing records on about 65 million persons. Of this total, 78 million cards representing 21 million individuals were in the Ident criminal file. During fiscal year 1981, Ident received an average of 27,392 fingerprint cards daily, of which about 12,684 involved criminal offenders. 1 Each card has to be recorded, examined for data completeness, searched against the master criminal file to locate any previous record, classified by fingerprint characteristics if a new print, and entered into an existing criminal history record (or a new record if the individual has no previous record). A copy of the record is then forwarded by mail (or occasionally by teletype or facsimile in case of urgent need) to the requesting agency. Based on data supplied by letter to OTA dated Oct. 30, 1981, from the Deputy Assistant Director, FBI Identification Division. During fiscal year 1981, I dent received a total of 6,848,043 fingerprint cards, of which 3,171,102 were from Federal, State, and local criminal justice agencies. Thus, based on 250 workdays, an average of 27,392 cards were received daily. Of these, a daily average of 12,684 were from criminal justice agencies. Reports from the field indicate that when an arresting agency forwards a fingerprint card to the FBI, it typically takes several weeks to receive a response. z Surveys conducted for the FBI in 1979 and 1980 indicated that the average Ident response time for processing of fingerprint cards was in the range of 36 workdays. 3 As of July and October 1981, the FBI estimated that Ident internal processing time (excluding mailing time) was averaging 27 and 25 workdays, respectively, for all categories of inquiries (both fingerprint checks and name checks). As of July 1982, processing time had improved, at least temporarily, to about 13 days, due to Idents suspension of record checks for federally chartered or insured banking institutions and State and local employment and licensing authorities. 4 The problem of slow response from manual systems is a key factor supporting the evolution to computerized systems. See Excerpts From Representative Viewpoints of State Criminal Justice Officials Regarding the Need for a Nationwide Interchange Facility, March 6, 1978, reprinted as app. B to U.S. Congress, Office of Technology Assessment, Preliminary Assessment of the National Crime Information Center and Computerized Criminal History System, Washington, D. C., December 1978, p. 71. 3Jet Propulsion Laboratory, FBI Fingerprint Identification Automation Study: AIDS III Evaluation Report, Volume I, California Institute of Technology, Pasadena, Calif., Nov. 15, 1980, pp. 1-1 to 1-3; prepared for U.S. Department of Justice, Federal Bureau of Investigation. Oct. 1, 1981, letter to all I dent fingerprint contributors from Nick F. Stames of the FBI, p. 2. Mar. 26, 1982, letter to OTA from Conrad Banner of the FBI; personal communication with Conrad Banner, July 30, 1982. Ident plans to restore these services on Oct. 1, 1982. Concern Over Multi-State Offender s The problem of multi-State offenders has in the United States have multi-State records been recognized for decades. It was one reason (that is, have been arrested in more than one for the initial establishment of a national reState during their criminal careers). More repository of criminal history records in the FBI cently, based on the composition of the NCIC/ to facilitate the exchange of records among difCCH file as of August 1, 1981, the FBI estiferent States. mated that 33 percent of all offenders have However, until relatively recently the actual multi-State records.5 An OTA study of crimipercentage of multi-State offenders was not See NCIC staff paper prepared for the Nov. 3-4, 1981, meetknown. In 1974, based on a sample of Ident ing of the NCIC Advisory Panel Board Subcommittee on th e Interstate Identification Index, Topic #3, p. 7. As of Aug. 1, records, the FBI concluded that approximate1981, 610,502 of the 1,841,776 total records in the NCIC/CCH ly 30 percent of the persons arrested annually file (or 33.15 percent) were multi-State.

PAGE 43

Ch. 3Evo/ution to Computerized Crimina/ History Records l 3 3 nal history files provides a confirmation and interpretation of the 1974 and 1981 FBI estimates. As summarized in table 1 based on a 1979 survey, OTA found that about 30.4 percent of individuals in the FBIs Ident file have arrests in more than one State, which agrees very closely with the earlier FBI figure of 30 percent. However, all Federal offenders are included in both the Ident and NCIC/CCH files. If Federal multi-State offenders are excluded, the percentage of multi-State records in the NCIC/CCH file drops from 33 to 12 percent.6 As of August 1, 1981, the percentage of NCIC/ CCH multi-State records for the eight States with records in the NCIC/CCH file fell within the 3-to 36-percent range, as shown in table 1. Thus, multi-State offenders represent a significant part of the total criminal population. Whether this multi-State population is com6Ibid. Excluding the 448,860 Federal offender records counted as multi-State, the remaining 161,642 multi-State records repre sent 11.6 percent of the 1,392,916 State records in the NCICi CCH file. Table 1 .Percent Multistate Offenders in Selected Criminal History Record Files Percent multi-State Selected criminal history record files offenders FBI Identification Division File b . 30.4 /0 FBI NCIC/CCH File c . . . 33.2 FBI NCIC/CCH File Excluding Federal Offenders c . . . 11. 6 States with records in the NCIC/CCH file c Nebraska. . . . . . 36.1 0/0 Michigan . . . . . 16.4 North Carolina . . . . 16.1 Florida. . . . . . . 13.8 lowa . . . . . . 11.0 Virginia . . . . . . 8.4 South Carolina . . . . 2.6 Texas . . . . . . 2.5 apercent of offenders in the file with arrests in two or more States bbased on a 1979 OTA Survey of record quality For the 168 Ident records With verifiable arrest events (see ch. 8), the number of different States i n which arrests occurred was distributed as follows. 1 State (117 records out of 168), 2 States (29 records), 3 States (8 records), 4 States (7 records), 5 States (3 records), 6 States (1 record), 7 States (2 records), 8 States (1 record) cBased on NCIC/CCH file size and composition as Of Aug. 1, 1981. SOURCE Off Ice of Technology Assessment and Federal Bureau of Investigation posed of criminals whose crimes are more serious or less serious than those of the general criminal population could not be determined from information available to OTA. Criminal Justice Reform and Computer s In the mid-1960s, the continuous rise in the incidence of crime, coupled with political and social tensions (e.g., over civil rights, urban renewal, and the Vietnam War), led to renewed efforts to reform the criminal justice process. In 1965, President Lyndon B. Johnson established a Presidents Commission on Law Enforcement and Administration of Justice to probe the causes of crime and recommend ways to improve its prevention and control. Two years later, in its 1967 report, the commission found serious deficiencies in criminal justice information in general and criminal history record systems in particular. The commission suggested that criminal justice could benefit dramatically from computer-based information systems. President Commission on Law Enforcement and Administration of Justice, The Challenge of Crime in a Free Society (Washington, D. C.: U.S. Government Printing Office, 1967), p. 266. The commission also concluded that the criminal justice process: 1) suffered from extreme decentralization, fragmentation, and a general lack of coordination of the agencies involved, and 2) was seriously overburdened partly owing to a general tendency in the United States toward overcriminalization, i.e., to prescribe criminal justice solutions to what were essentially social and moral problems. Nevertheless, the commissions emphasis on the use of computer and communication technology in the criminal justice process helped set the agenda for subsequent legislative and administrative initiatives with respect to computerized criminal history records. Foremost among these initiatives was the passage of the Omnibus Crime Control and Safe Streets Act of 1968 and the establishment of LEAA. Title I of this act represented the Federal Governments first comprehensive

PAGE 44

34 l An Assessment of Alternatives for a National Computerized Criminal History System grant-in-aid program for reform and modernization of the criminal justice process. The act provided for a block grant approach to Federal funding, recognizing that prevention and control of crime was basically a State and local responsibility. LEAA was established to administer the block grant program and to work closely with the States and localities in improving the administration of criminal justice at every level. Following the recommendations of the Presidents Commission, LEAA gave relatively high priority to grants for the development of computerized crimin al justice information and statistical systems in the States, including computerized criminal history record systems. From 1970 to 1980, LEAA awarded about $207 million in categorical grants to the States for criminal justice information systems, although only about $39 million was earmarked specifically for CCH-related systems. 8 An additional $200 million to $400 million in LEAA block grants to the States is estimated to have been spent on information systems.9 These grant programs ended in fiscal year 1981. 8Based on data from the Office of Justice Assistance Research and Statistics, U.S. Department of Justice. See ch. 5 for discussion. Based on data and analysis provided in a Sept. 9, 1981, letter to OTA from Tom Dalton of Seattle University. See ch. 5 for discussion. Birth of the National Crime Information Center Almost from its inception, Ident has experimented with new techniques to process fingerprints and rap sheets more efficiently. For example, as early as 1934 the FBI tried a punchcard and sorting system for searching fingerprints, but the technology at that time could not handle the large number of records in the Ident files. It was not until the development of third generation computer technology in the early 1960s that the FBIs goals of automating criminal records became feasible. For the first time computer technology made it possible to electronically store hundreds of thousands or even millions of records, and to process record requests and updates almost instantaneously. The FBI first used computer technology to process individual criminal records in 1963. In a Careers in Crime study, the criminal identification records of 194,000 Federal offenders were computerized and regularly updated as new information came into the FBI on arrests, convictions, and other criminal justice transactions. The study found that criminal activity increasingly was spilling over local government boundaries. It also concluded that the existing mechanisms for the exchange between local and State jurisdictions of wanted persons and stolen vehicle information, as well as rap sheets, were too slow and incomplete.l0 As a consequence, the Uniform Crime Reports (UCR) section of the FBI Criminal Records Division recognized a need to initiate the use of new computer technology. Computers could automate criminal records, allowing remote access to States and localities through appropriate communication lines. In 1965, the UCR section sought approval to proceed with the development of a national crime information system. Initially, it would be used for the rapid exchange of wanted person and stolen property information, with criminal history information to be added later. On January 20, 1966, the Attorney General approved the development of a national crime information center in the FBI. In his memorandum to FBI Director Hoover, the Attorney General cited the FBIs collection and exchange of criminal records with local police organizations as sufficient authority to establish the center.11 10 Alan F. Westin and Michael A. Baker, Data Banks in a Free Society (New York: Quadrangle Books, 1972), p. 51. ] Nicholas deB. Katzenbach, Attorney General of the United States, Memorandum for J. Edgar Hoover, Director, Federal Bureau of Investigation, National Crime Information Center, Jan. 20, 1966.

PAGE 45

Ch. 3Evolution to Computerized Criminal History Records l 3 5 Subsequent to the Attorney Generals approval, several planning meetings were held in which the FBI, the International Association of Chiefs of Police, local police departments and other law enforcement agencies (primarily State police) participated. The planning process helped determine the files to be initially included, criteria for entry of records into those files, and operational procedures to ensure that participating local and State information systems would be technically compatible with the national center. On January 27, 1967, NCIC began operation with approximately 23,000 records of wanted persons and stolen property in its computer files. The 15 initial participating agencies are listed in table 2. By September 1967, the FBI reported that over 260,000 records were on file on stolen veTable 2.initial Participating Agencies for the National Crime Information Center Washington, D. C., Police Department Maryland State Police, Pikesville Pennsylvania State Police, Harrisburg Philadelphia, Pennsylvania Police Department New York State Police Department New York City Police Department Boston, Mass. Police Department Chicago, Ill. Police Department St. Louis, Mo. Police Department Denver Field Division of the FBI California Highway Patrol, Sacramento Texas Department of Public Safety, Austin New Orleans, La, Police Department Georgia State Patrol, Atlanta Virginia State Police, Richmond SOURCE Off Ice of Technology Assessment hicles and license plates, stolen guns, stolen articles, and wanted persons. NCIC was operating 7 days a week and 22 hours a day, averaging 10,000 entries and inquiries daily. Origins of the Computerized Criminal History Progra m The way was now prepared for the development of a national computerized criminal history program. In 1967, the Presidents Commission on Law Enforcement and Administration of Justice strongly endorsed the use of computers in the criminal justice process. In 1968, Congress passed the Omnibus Crime Control and Safe Streets Act, which established LEAA to help the States improve their administration of criminal justice. Priority was to be placed on upgrading criminal justice information systems at the State and local levels. In 1967, when the FBI began operating NCIC, the first computerized national criminal justice information system, criminal history files were initially excluded. This was a deliberate decision in order to hold off until the stolen property and wanted person files were implemented. The feasibility of automating criminal history records had already been demonstrated and was never doubted. As it turned out, implementation of the CCH file proved to be much more difficult for a variety of reasons, including disagreements within the FBI and between the FBI and LEAA. In 1968 and 1969, the FBI established a working group made up of NCIC and Ident staff to begin developing plans for automating criminal history records in the FBI. As a result of disagreements over the impact of a CCH file on the maintenance of rap sheets in Ident, the FBI planning effort was temporarily halted. The initiative then moved to LEAA. LEAA was already receiving a number of grant applications from States seeking funds to develop computerized criminal justice information systems. LEAA agreed there was a need for such systems, but was concerned that each State might go off in its own direction, leaving us with a bewildering complex of independent and incompatible programs. 2 In ad z Project SEARCH, National symposium on CriminaJ Justice Information and Statistics Systems, Sacramento, Calif., 1970, p. 10.

PAGE 46

36 l An Assessment of Alternatives for a National Computerized Criminal History System dition, the total funds requested far exceeded LEAAs budget for this purpose. LEAA therefore decided to fund a new organization called Project SEARCH to develop and demonstrate that a computerized criminal offender file, containing data from all segments of criminal justice, can be standardized and exchanged between States on a timely basis. "1 3 States originally participating in Project SEARCH included Arizona, California, Connecticut, Florida, Maryland, Michigan, Minnesota, New York, Texas, and Washington. Within 14 months, working with LEAA grants, Project SEARCH developed a computerized rap sheet format and completed an online demonstration of the interstate exchange of Crimin al history records. The demonstration linked six States with a central computer in Michigans State Police headquarters. In this way, Project SEARCH and LEAA proved that it was technically feasible not only to automate manual history records, but to use a computerized system for the interstate exchange of criminal histories. However, questions arose about who should operate the system and what computerized criminal history information should be maintained at the Federal and State levels. During 1970, the pros and cons of different alternatives were analyzed and debated by the Attorney General, the FBI, and LEAA, and later the Office of Management and Budget.1 4 project SEARCH Newsletter, Sacramento, Calif., 1969, vol. 1, issue 1. 14See DonaId A. Marchand, et al., A Histury and Background Assessment of the National Crime Information Gmter and On December 10, 1970, the Attorney General decided that the FBI would take over management responsibility for the CCH system, rather than LEAA, a joint LEAA/FBI entity, or a consortium of States. The FBI named the system the Computerized Criminal History (CCH) program and operated it as part of NCIC, using NCIC computers and communication lines. The CCH program began operations on November 29, 1971, joining wanted persons and stolen property files maintained in the NCIC. On an interim basis, the CCH file was to contain the detailed criminal history of each offender whose record was entered by the States into the system. Eventually, under the singleState/multi-State plan adopted by the FBI, NCIC/CCH would maintain only summary data in the form of an index of single-State offenders, while the States would maintain detailed records. For multi-State offenders and Federal offenders, NCIC/CCH would maintain the detailed records. Due to a variety of issues addressed later in this report, neither the single-State/multiState plan nor any other NCIC/CCH alternative has been fully implemented.l s Computerized Criminal History Program, Bureau of Governmental Research and Service, University of South Carolina, June 1979, sec. III, The Computerized Criminal History Program: Its Origins and Initial Implementation, especially pp. 78-86, for details of this debate. sFor a complete discussion of the history and background of NCIC/CCH, see Ibid. and Donald A. Marchand, The PoLtics of Privacy, C.bmputers and Criminal Justice R&or& (Arlington, Va.: Information Resources Press, 1980), especially chs. 4 and 6.

PAGE 47

Chapter 4 Description of the National Crime Information Center in Context

PAGE 48

. Content s Chapter Summary . . . . . . . . . . . 39 NCIC . . . . . . . . . . . . . . 39 Related Systems . . . . . . . . . . . 39 What NCIC Is. . . . . . . . . . . . . 40 System Description . . . . . . . . . . . 40 NCIC Files . . . . . . . . . . . . 42 Volume of NCIC Transactions . . . . . . . . . 43 CCH Participation . . . . . . . . . . . 43 Related Systems . . . . . . . . . . . . 44 Identification Division System . . . . . . . . 44 National Law Enforcement Telecommunications System. . . . 46 Treasury Enforcement Communication System . . . . . 45 Justice Telecommunications System . . . . . . . 46 State and Local Systems . . . . . . . . . 46 TABLES Table No. 3. 4. 5. 6. Number of Records Included in NCIC, by File. . . . . . States With Full Participation in the NCIC/CCH Program . . Number of States With On-Line Computerized Criminal History Information Systems . . . . Status of State Criminal HistorySystems FIGURE -Figure No. 6. NCIC Network . . . . . August . ............. . . . . Page 42 43 47 48 Page 41

PAGE 49

. Chapter 4 Description of the National Crime Information Center in Context Chapte r The National Crime Information Center (NCIC) uses computer and telecommunication technology for collecting, storing, retrieving, transmitting, and disseminating criminal justice information. NCIC interfaces with and complements many other criminal justice information systems at the Federal, State, and local levels. Thus, NCIC maybe viewed as one member of a family of systems. NCI C NCIC contains 10 files. Eight of these, the so-called hot files, furnish an electronic bulletin board capability used by law enforcement agencies to list persons or properties (e.g., vehicles, guns, and securities) that are wanted, missing, or stolen. The ninth file is the computerized criminal history (CCH); the tenth is the Criminalistics Laboratory Information System (CLIS) file which contains no information about individuals. As of October 1981, NCIC included about 9.3 million records.4 million hot file records and 1.9 million CCH file records. However, about 90 percent of NCIC traffic is for use of the stolen vehicles/plates and wanted/missing persons files. CCH use involves about 4.4 millions transactions annually, or about 3.5 percent of total NCIC traffic. The low level of CCH use compared with hot file use is due in part to the small number of States that are fully participating in the CCH program. Whereas all 50 States can both enter data into and retrieve data from the hot files, as of December 1981 only 8 are authorized to enter data into the CCH file. Only 15 States have ever fully participated in CCH, with never more than 13 at any one time. However, in addition to the 8 fully participating States, 41 Summar y others participate in CCH on a limited basis (retrieve data entered by others) and have signed a management control terminal agreement with NCIC. NCIC has 137 direct communication lines to law enforcement and criminal justice agencies, including 79 State and Federal agencies, the Federal Bureau of Investigation (FBI) headquarters, 27 FBI field offices, and 2 FBI metropolitan resident agents. An estimated 64,000 other Federal, State, and local law enforcement and criminal justice agencies are entitled to access NCIC over these lines. Depending on the State, these agencies may include, for example, local police departments, U.S. magistrates, district attorneys, courts, county jails, State hospitals, and parole boards. Related System s NCIC interfaces with a large number of other criminal justice information systems. The following are particularly important: The FBI Identification Division (Ident) fingerprint record repository is integral to NCIC/CCH, since every CCH entry must reference an FBI identification number initially assigned by Ident based on positive fingerprint identification. The Automated Identification Division System (AIDS) is the FBIs effort to automate Ident, and in October 1981 included about 5.8 million criminal history records. All first offender records are entered into AIDS. Ident is not maintaining manual rap sheets on any individual who is in the AIDS file. When a criminal history record is needed it is generated by computer. As of October 1981, 58 percent of the records 39

PAGE 50

40 l An Assessment of Alternatives for a National Computerized Criminal History System -. in the NCIC/CCH file were also maintained in AIDS. l National La w Enforcement Telecommunications System (NLETS) is a computerized message switching network linking local, State, and Federal law enforcement agencies. Operated by a nonprofit corporation controlled by the States, NLETS does not hold or manage record files. It is a principal means by which user agencies verify data from the NCIC hot files. NLETS can also be used to transfer records from the NCIC/CCH file and between State CCH files. l Treasury Enforcement t Communication System (TECS) is a large communication network and computerized data base managed by the Customs Service in the Treasury Department. TECS supports over 1,400 terminals in, for example, the 50 States, Puerto Rico, and Canada, and serves various Federal agencies and Interpol (the International Police Organization). TECS users can access the NCIC/ CCH file to obtain summary records online. Full records can be produced offline and delivered to TECS users who are authorized to receive them. Justice Telecommunications System (JUST) provides computerized administrative message service to Department of Justice (DOJ) offices in Washington, D. C., and to 329 department offices in 169 cities nationwide. JUST has a direct link to NCIC and offline linkages to the Department of Defense AUTODIN network and the Department of State Diplomatic Network. State and local systems are also highly automated. As of August 1982, 27 States had CCH files, and another 7 had an automated name index. At the local level, most major metropolitan police departments use computer-based systems (19 such departments have direct lines to NCIC). Thus, automated systems are clearly prevalent at the local as well as State and Federal levels, and NCIC is only one of a rather large family of computer-based criminal justice information systems. What NCIC I s System Descriptio n NCIC is a computer-based national information system whose principal function is to support law enforcement and criminal justice activities. Managed and operated by the FBI, NCIC uses computers and telecommunication technology for collecting, storing, retrieving, transmitting, and disseminating criminal justice information among government agencies at the Federal, State, and local levels, and among some private organizations. The center is located in the FBIs computer facility in Washington, D. C., and includes a telecommunication network that reaches automated or manual teletype terminals in all of the 50 States, the District of Columbia, Canada, Puerto Rico, and the Virgin Islands, as shown in figure 6. As of October 1981, the NCIC network had 137 direct communication lines to law enforcement and criminal justice agencies, including 79 State and Federal agencies, FBI headquarters, 27 FBI field offices, and 2 FBI metropolitan resident agents.1 As shown in figure 6, State agencies with direct lines include primarily State police or highway patrols or departments of public safety, justice, or criminal identification. Nine Federal agencies (listed in fig. 6) have a direct line to NCIC. Although only 79 State and Federal agencies have a direct line to NCIC, an estimated 64,000 other Federal, State, and local law enFederal Bureau of Investigation, Interstate Identification Index (III): Background and Findings for July-September 1981 Phase I Pilot l?rojec~ Dec. 4, 1981, p. 22.

PAGE 51

Ch. 4Description of the /Vationa/ Crime Information Center in Context l 4 1 I z .. 5 Y . o l -. --- c 0 z 0 z al > E -5 c 0 % z

PAGE 52

42 An Assessment of Alternatives for a National Computerized Criminal History System forcement and criminal justice agencies are entitled to access NCIC over these lines. 2 NCIC users may retrieve data from the files, modify existing records, and/or add new data to the files in accordance with the relationship each has established with NCIC. Not all users are permitted to perform all functions. The rules for participating are described in the NCIC operating manual. 3 The NCIC system interfaces with and complements other systems for processing and disseminating law enforcement and criminal justice data. Some are operated by other Federal agencies, and many have been implemented at the State and local levels. The degree to which each is automated varies from system to system. NCIC File s The NCIC system provides access to data contained in 10 files. Eight of these files, the so-called hot files, furnish a bulletin board capability that is used by law enforcement agencies to list people or properties that are wanted, missing, or stolen. The ninth file, the CCH file, contains archival criminal history data most often used for criminal justice ac2 FBI estimate, NCIC operating hfiwIud pt. 10. P 13. Federal Bureau of Investigation, Technical Services Division, National Crime Information Center Section, NCIC Operating ManuaL updated periodically. tivities subsequent to apprehension of an individual. A tenth file, CLIS, provides access to a data base of laboratory information. Although maintained on and accessed through the NCIC system, in most respects CLIS is a separate information system, except that it shares NCIC computer and telecommunication facilities. CLIS contains no information about individuals. The number of records included in the various NCIC files (except for CLIS) is shown in table 3. These files contained over 9 million records as of October 1, 1981. The CCH file is the second largest (exceeded only by the stolen securities file), even though at present only eight States are authorized to enter criminal history record information into the CCH file. Most inquiries of the hot files are made by law enforcement and investigative personnel seeking to determine whether a specific person or item of property is wanted. Such an inquiry may occur, for example, when a traffic officer stops a vehicle and seeks to determine whether it has been reported stolen or whether the driver is wanted for possible involvement in criminal activity. In all cases, NCIC operating procedures require that information obtained from a hot file be verified with the agency that originated the record before police take any action based on it. Table 3.Number of Records Included in NCIC, by File Number of records as of Percent of total File 6/1/79 12/31/79 12/31/80 10/1/81 NCIC records Hot Files: 1. Stolen Vehicles a . . . . . 970,714 1,108,574 1,174,639 1,163,771 12.50/o 2. Stolen Articles . . . . . 1,091,461 1,502,209 1,562,284 1,427,535 15.4 3. Stolen Guns. . . . . . 1,337,310 1,426,008 1,574,959 1,674,814 18.0 4. Stolen License Plates . . . 397,706 499,868 551,373 543,173 5.8 5. Wanted Persons b . . . . 148,644 162,128 179,044 190,159 2.1 6. Stolen Securities . . . . 1,998,778 2,189,594 2,303,716 2,361,971 25.4 7. Stolen Boats . . . . . 17,615 21,277 24,707 22,807 0.2 8. Missing Persons . . . . . 21,535 22,722 23,406 24,640 0.3 Subtotal . . . . . . 5,983,763 6,932,380 7,394,128 7,408,870 79.7 /0 Computerized Criminal Histories File: 9. CCH . . . . . . . 1,482,017 1,606,837 1,706,955 1,885,457 20.3 Total . . . . . . . 7,465,780 8,539,217 9,101,083 9,294,327 100.0 /0 Includes vehicle parts, felony vehicles, airplanes, and trailers bDoeS not {rlclude Carladlan warrants, which totaled 183 as of Oct 1, 1981 SOURCE Federal Bureau of Invest lgatton

PAGE 53

Ch. 4Description of the Nationa/ Crime Informafion Center in Context l 4 3 Most inquiries of the CCH files are generated by criminal investigations or by criminal justice actions following an individuals detention. In these cases, data pertaining to a specific individual are requested. The data obtained are then used in conjunction with steps in the criminal justice process such as arraignment, determination of sentence, and the decision to grant parole. To obtain a CCH summary record, an agency must provide its own identifier plus a valid basis for searching the file. A search may be based solely on a State or FBI identification number, on a name and social security or other identification number, or on name, sex, race, and date of birth information. To obtain a full record, the requesting agency must provide a State or Federal identification number for the subject. Volume of NCIC Transaction s In terms of number of transactions, the bulk of the traffic processed by the NCIC system is concerned with processing data in and for the hot files. Use of the hot files is dominated by law enforcement and criminal justice personnel engaged in tactical operations. Information from the hot files is used primarily to guide actions taken prior to the arrest of an individual or the seizure of a piece of property. For the month of September 1981, NCIC transactions totaled 10,270,500, averaging 342,350 daily. This represents a traffic growth of about 28 percent over the last 2 1/2 years. Transactions for the stolen vehicles/plates and wanted/missing persons files account for roughly 90 percent of the total NCIC traffic. The CCH file is used primarily in postarrest situations and represents a very small part of total NCIC traffic. In September 1981, CCH traffic accounted for about 3.5 percent of the NCIC monthly total. At this rate, CCH use would involve about 4.4 million transactions annually. The low level of CCH traffic compared with hot file traffic is also due in part to the small number of States fully participating in the CCH program. CCH Participatio n There are two levels of participation in the CCH program. Full participation permits the organization to add data to the file as well as \ FBI, NCIC ,Section, NCIC Newsletter, October 1981, p. 2. *Th e averaw d~y NCIC traffic totaled 266,479 tmsactions in April 1979. Thus, the daily average increased 28.5 percent by September 1981. retrieve data from it. Such users are responsible for entering data into the files and maintaining the records they have entered. This can require considerable resources from the participant. Generally, States have been hesitant to participate fully in the CCH program. The maximum number of fully participating States that has been reached is 13. As of December 1981, only eight States were full participants, as listed in table 4. Less than full participation allows the user to access the data in the CCH file, but not to contribute to it. Users at this level are still required to meet the basic criteria established for participation and to execute a management control agreement that the rules of NCIC will Table 4.States With Full Participation in the NCICICCH Program Currently active (as of December 1981) Florida Iowa Michigan Nebraska North Carolina South Carolina Texas Virginia Previously active (withdrew from full participation prior to December 1981) Arizona California Illinois Minnesota New York Ohio Pennsylvania SOURCE Federal Bureau of Inves; tgatlon

PAGE 54

44 An Assessment of Alternatives for a National Computerized Criminal History System be followed. For example, the agreement includes provisions requiring the user to safeguard the data and limit its distribution to those authorized to receive it. (See ch. 6 for further details.) As of December 1981, only KanRelate d NCIC interfaces with a large number of criminal justice information systems operated by State, local, and Federal agencies. A selected sample of such systems is discussed below to illustrate mutual dependencies among systems designed to support criminal justice and law enforcement activities. Identification Division Syste m As of December 1981, Ident maintained files containing approximately 175 million fingerprint cards representing 65 million individuals. Of the total number of cards, 78 million representing 21 million individuals were in Idents criminal file, and 96 million representing 44 million people were in Idents civil file. In addition, Ident maintains files of criminal history data (in the form of rap sheets) on individuals who are included in the criminal fingerprint file. The criminal fingerprint file operated by Ident is key to the operation of the automated NCIC/CCH file. It is this file that is searched when an FBI identification number is required for entering a record in the NCIC/CCH file. Ident maintains manual criminal history records in parallel with the computerized records in the NCIC/CCH file. Ident also initiates the process of creating NCIC/CCH records for all Federal offenders. In addition, when Ident receives a fingerprint card from a State that is not a full participant in the NCIC/CCH program on a subject who already has a record in NCIC/CCH, Ident initiates the procedures to update the CCH record. Thus, Idents operations are integral to the NCIC/CCH system. sas had no agreement and therefore was not allowed access to the CCH file. 5 FBI, III: Background and Findings, op. cit., p. 24. System s In an effort to automate Ident record processing, the FBI has been developing, since the early 1970s, a three-phased system called AIDS (Automated Identification Division System). The first phase, AIDS-I, was implemented in August 1973 and supports a computerized data base containing the records of first offenders arrested since that time. This automated file has grown at the rate of approximately 750,000 records per year or 3,000 records per workday, and in October 1981 totaled about 5.8 million records. 6 Ident is not maintaining manual rap sheets on any individual who is in the AIDS file. When a rap sheet is needed, it is generated by computer. The second phase, AIDS-II, became operational in October 1979 and added the capability for automated name searching of the computerized arrest record file (AIDS-I). It is already handling roughly 45 percent of Idents name searching operations. In AIDS-III, not yet implemented, fingerprint readers will be used to automate the matching of fingerprint cards submitted to Ident against the criminal fingerprint file. As of October 1981, the prints of 14.6 million individuals born in 1929 or later had been converted, representing about 70 percent of the criminal file. However, automated fingerprint searches were run on only about 17 percent of the file due to equipment and staffing limitations. Automated processing of low-quality prints, such as latent prints found at the scene of a crime, is expected to be possible through the use of semiautomatic fingerprint reader equipment. 8NCIC Staff Paper prepared for the Nov. 3-4, 1981, meeting of the NCIC Advisory Policy Board Subcommittee on the Interstate Identification Index, Topic #7, p. 5.

PAGE 55

Ch. 4Description of the National Crime Information Center in Context l 4 5 The development of AIDS is intended as a means to automate Ident operations, not to replace or overlap the functions that are performed by NCIC/CCH. However, as presently structured the operation of the NCIC/CCH file will depend on the operation of AIDS to establish the identity of subjects before an NCIC/CCH record can be entered or updated. For example, the process of creating a record in the NCIC/CCH file for individuals who have committed their first Federal offense has been automated. An interface between AIDS and NCIC/CCH creates a copy of the AIDS record for the NCIC/CCH file. Similarly, whenever a record in AIDS for a Federal offender is updated, the related record in the NCIC/CCH file is automatically updated also. As of June 1979,647,990 records in AIDS had been duplicated in NCIC/CCH, which at that time represented 44 percent of total CCH records. As of October 1981, the percentage of NCIC/CCH records also in AIDS had increased to about 58 percent. 7 Thus, at least in the short run, the FBI is operating two systemsIdent/AIDS and NCIC/CCHthat maintain criminal history records on individuals, although alternatives for consolidation of these systems are being considered.* National Law Enforcemen t Telecommunications Syste m (NLETS ) NLETS is a computerized message switching network linking local, State, and Federal law enforcement agencies for the purpose of information exchange. Operated by a nonprofit corporation controlled by the States, NLETS does not hold or manage data files. It is a communication network functioning in much the same way as the public switched Telex network to carry messages between various users. NLETS plays an important role in the operation of the NCIC hot files. NCIC operating Ibid., p. 6. As of October 1981, about 1.085 million (or 57.6 percent) of the 1.885 million total NC IC/CCH records were also in AIDS. ** di9cu99ion in ch. 10. procedures require verification of data obtained from its hot files before action is taken based on that data. NLETS is a principal means by which such verification is accomplished. In such cases, NLETS is used to query directly the agency that would have information confirming whether an individual or item of property listed in the NCIC hot files is wanted or stolen. NLETS could be used to transmit criminal history information in the absence of NCIC/ CCH or any other system designed expressly for this purpose. There is no easy way to be certain of the contents of the messages moving over NLETS, although NLETS use statistics indicate that 2 percent of messages relate to criminal records. However, it is possible that some criminal history information is also being transmitted in the form of administrative messages, which account for about 17 percent of NLETS use. Florida and the FBI completed a pilot project in 1981 using both NLETS and CCH, as described in chapter 10. And NLETS is an integral part of the ongoing test of the Interstate Identification Index (III) concept. Treasury Enforcemen t Communication Syste m (TECS ) TECS is a large communication network and data base that supports over 1,400 terminals in, for example, the 50 States, Puerto Rico, and Canada. Managed by the Customs Service, it serves a variety of law enforcement and criminal justice agencies, including the Bureau of Alcohol, Tobacco, and Firearms; the Internal Revenue Service; and the National Central Bureau of the International Police Organization (Interpol); as well as the Bureau of Customs.TECS also serves the Drug Enforcement Administration, the Immigration and Natural ization Service, the U.S. Department of State, and the U.S. Coast Guard. TECS is used by customs officers and other officials to ascertain whether or not certain individuals or items of property are of interest to or wanted by law enforcement, c riminal jus-

PAGE 56

46 l An Assessment of Alternatives for a National Computerized Criminal History System tice, or related agencies. One of the main uses of the system is the checking of vehicles at border crossings to see if they are stolen, have been used in the commi ssion of a crime, or are associated with an individual who is wanted. The TECS data base is also used to determine if there is any reason to detain an individual at a port of entry or exit, including airports with international arrival service, or to prevent someone from crossing the border. Among the data available to TECS are pointers to the NCIC hot files. The indices maintained on TECS and the corresponding files of NCIC are maintained in real-time and coordinated on a regular basis. Records in the TECS index but not in NCIC files are deleted from the TECS index, while entries that are in NCIC files but not in the TECS index are added to TECS. A lookout file of persons and property of interest to Customs and the other users of TECS is maintained independently and is not coordinated with NCIC files. Items in this file may not be included in NCIC files. Users of TECS can access the NCIC/CCH file, but with certain restrictions. Only the NCIC/CCH summary record will be printed out at a TECS terminal. If the complete record is required, the TECS user must arrange to have it printed out at a primary NCIC/CCH user terminal and forwarded through the mails. Foreign inquiry by Interpol to the NCIC/CCH data must pass through a human operator who is an American law enforcement agent resident in the National Central Bureau (Washington, D. C.) office before it can be processed by the CCH system. The future configuration of TECS is under review by the current administration. Justice Telecommunication s System (JUST ) The DOJs JUST provides computerized administrative message service to department offices in Washington, D. C., and to approximately 329 DOJ offices in 169 cities nationwide. JUST provides a direct linkage to NCIC, enabling DOJ offices (such as the U.S. Marshals Service) to make inquiries against the NCIC data base. Under this arrangement, department offices first make inquiries against their own data bases at the DOJ computer facility and then, if necessary, route additional inquiries through the JUST facility to NCIC. JUST also has offline linkages to the Department of Defense AUTODIN network and the Department of State Diplomatic Network in support of international law enforcement activities. State and Local System s The capabilities of criminal justice information systems at the State and local levels range from handwritten manual files that are kept in the desks of local police departments to computerized systems that are considerably more comprehensive in terms of information content than those operated by the FBI. In September 1981, the FBI conducted a survey of all State criminal justice record repositories. Selected results, summarized in table 5, indicate that about one-half of the States have online computerized criminal history files and roughly another one-quarter have an automated name index. However, the 25 States with an online CCH file account for about 80 percent of all fingerprint cards submitted to State and Federal repositories. The 12 States with completely manual systems account for only 5 percent of the total. As of September 1981, the 25 States with an online CCH file maintained collectively a total of about 10.9 million full CCH records, compared with 5.8 million CCH records in the AIDS file and 1.9 million in the NCIC/CCH file. The States also maintain a large number of manual criminal history records, estimated at roughly 25 million to 30 million records in 1979. Available evidence suggests that the rate of dissemination for computerized records is significantly higher than for manual records. In 1979, while CCH and manual record disseminations were about equal in absolute numbers

PAGE 57

Ch. 4Description of the /Vationa Crime /nfomnation Center in Context 47 Table 5.Number of States With On-Line Computerized Criminal History Information Systems On-line CCH file a Automated name index Manual file only Total Number of States b ., . . . 25 (50/0) 13 (260/o) 12 (24%) 50 (100/0) Number of fingerprint cards received by States c . . . . . 3.37 million (81 0 / 0 ) 0.57 million (14/0) 0.22 million (5/0) 4.16 million (100/0) Number of fingerprint cards submitted to FBl d . . . 2.35 million (81 %) 0.40 million (14/0) 0.16 million (5%) 2.91 million (100/0) EXCIUde~ s~ate~ ~lth ~ff.llne CCH flier ~,, wa~hlngton and MaWland, Including these t~~ States, the 27 states accounted for about 85 percent of all criminal fingerprint cards submitted to State and Federal repositories bBased on septernber 1961 survey of State crlmlnal history record repositories ccrlmlnal fingerprint cards received annually by State reposltorles, State estimates dcriminal fingerprint cards submitted by States to the FBI dur!ng fiscal year 1981; FBI estimates SOURCE Federal Bureau of Investigation, September 1981 survey of State crlmlnal record repositories (roughly 5 million each), these disseminations accounted for more than half of all CCH records maintained but less than one-fifth of all manual records. 8 This is due in part to the record automation policies of many States (21 States as of September 1981 9 ) which require that the records of current offenders be converted first. Most States do not convert existing manual files until an activity (e.g., arrest) occurs. Also, records can be updated more easily and disseminated more quickly once they are computerized. At the local level, most major metropolitan police departments use computer-based criminal justice information systems. Nineteen such departments have direct lines to NCIC. Literally hundreds of other police departments have access to NCIC through metropolitan or State agency terminals. The exact number of police departments with computerized information systems is not known. However, as early as 1974, a survey of police departments in cities with populations exceeding 50,000 found that 56 percent (193 of the 326 departments responding) were using computers. Criminal justice recordkeeping was reported as the second most important use of computers, followed closely by police patrol and inquiry systems based on hot files of wanted persons and stolen property. OTA 50-State Survey conducted in 1979-80. See appendix B for list of State officials responding. NCIC Staff Paper prepared for the Nov. 3-4, 1981, meeting of the NCIC APB Subcommittee on II 1, Topic 46, p.3. Kent W. Colton, The Use and Impact of Police Computer Technology, in Kenneth Kraemer and John King, Cbmputers in Local Government: Police and Fire (Pennsauken, N. J.: Auerbach, 1981), pp. 2-4. In a 1979 OTA survey of the 50 States, 11 OTA found that 34 States reported a total of about 9,000 terminals at the State and local levels with direct access to criminal history information in State repositories. Several States contacted in a 1982 followup survey indicated that the number of terminals had increased since 1979; for example, from 244 to 700 terminals in Massachusetts, 70 to 206 in New Jersey, 80 to 102 in South Carolina, and 225 to 250 in Virginia. In 37 of 49 States reporting, law enforcement officers on patrol duty can gain access to criminal history information in State files through local police and patrol inquiry systems. Thus, the clear implication is that automated systems are prevalent at the local as well as at State and Federal levels. In a followup survey, OTA found that as of August 1982,27 States had a CCH file (including an automated name index), 7 States had an automated index, and 16 States were operating manually. The discrepancies when compared with the results of the September 1981 FBI survey are explained as follows. Two States (Washington and Maryland) with CCH files were counted by the FBI as automated index States since the CCH files did not permit online interstate access. Five States (Idaho, Maine, New Mexico, Pennsylvania, and South Dakota) counted as automated index States by the FBI were still in the process of implementation as of August 1982. One State (Massachusetts) counted by the FBI as manual actually has an operational automated name index. The OTA results are summarized OTA 50-State Survey.

PAGE 58

48 l An Assessment of Alternatives for a National Computerized Criminal History System in table 6. For those 12 States in the process and/or uncertainties in staffing and funding. of implementing an automated name index With full implementation, all but four States and/or CCH file, the estimated time to comwould have at least an automated name index; pletion ranged from 1 month, to 1 year, to an two of these four remaining manual States do indefinite time period, due largely to variations have plans to automate. Table 6.Status of State Criminal History Systems, August 1982 Automated Manual file CCH file name index only Totals Operational Number of States . . . 27 (54/0) 7 (14%) 16 (320/o) 50 (100%) Under implementation Number of States . . . 2 a 1O b Totals after implementation Number of States . . . 29 (580/o) 17 (34%) 4 (8/0) 50 (100/0) a Estimated time to complete implementation: 1 month (1 State), and 6 months (1 State). Estimated time to complete implementation: 3 months (1 State), 6 months (2 States), 9 months (1 State), 1 year (1 State), indefinite (3 States), and unknown (2 States). SOURCE: Office of Technology Assessment 50-State survey, 1982 followup.

PAGE 59

Chapter 5 NCIC Technology. and Costs -%

PAGE 60

-,.) Contents Technology . costs . . NCIC Technology Chapter Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hardware Previously Used by NCIC . . . . . . . Hardware Upgrade and Message Switching . .......?. . Status of Hardware and Software ,, NCIC Costs . . . . . costs costs to the FBI..... . . Upgrade . . . . . . . . . . . . . l . . . . . . . to the States . . . . . . . . . TABLES Page 51 51 51 52 52 52 54 55 55 56 TableNo. Page 7. NCIC Direct Costs, Fiscal Years 1972-81 . . . . . . 56 8. LEAA Grants to States for Comprehensive Data Systems and Statistical Programs, Total by Fiscal Year . . . . . 57 9. LEAA Grants to States for CCH-Related Systems, Total by Fiscal Year and by State . . . . . . . . . . . 57 IO. CCH/OBTS Operational Costs for 1978 by State.. . . . . 58

PAGE 61

Chapter 5 NCIC Technology and Costs Chapter Summar y Technolog y The computer and communication technologies used by the National Crime Information Center (NCIC) have been the subject of considerable controversy in recent years. Part of the controversy relates to fundamental concerns about the need for and impacts of NCIC, and particularly the computerized criminal history (CCH) fileconcerns that have been intertwined with technology issues. Another aspect relates to operational problems experienced with the NCIC system, such as unscheduled downtime and slow response to inquiries, owing in part to the use of old equipment. The need for updated equipment conflicted with the concern that new computer technology would make it easier for NCIC to engage in unauthorized functions such as message switching (i.e., the transfer or switching of messages from one State to another through the NCIC computer). Late in 1979, the Federal Bureau of Investigation (FBI) received authority from Congress to upgrade the NCIC communications controller (a device that controls and manages the flow of messages into and out of the computer). While any new controller has the potential to be programed for message switching, the congressional authorization strictly prohibited any message switching applications and required a periodic General Accounting Office (GAO) audit to ensure FBI compliance. Early in 1980, the FBI, with congressional concurrence, took action to upgrade the main computer, subject to the same conditions. These hardware upgrades, combined with operating system software enhancements that have now been implemented, have improved the quality of NCIC service and reduced downtime. The FBI recognizes that NCIC applications software also needs to be upgraded. While there are no firm plans as yet, the FBI has initiated a review of NCIC system needs for the next 5 years, including the possibility of a major system redesign with further hardware as well as software improvements. In July 1982, the FBI announced a major computer procurement to, among other things, further upgrade the NCIC computers. cost s The total financial cost to the Nation of operating NCIC is shared by the FBI and the users. The FBI pays for the central computer facilities (including administrative, operational, and programing costs) and the communication links, while the users pay for the terminals and the costs of gathering, inputting, and processing the data. The Federal budget covers the costs to the FBI and to the Federal agencies that use NCIC. State and local budgets cover most of the remaining costs, although in the past these have been partially underwritten by the Federal Government through grants from the Law Enforcement Assistance Administration (LEAA) and others. Although NCIC cost components can be identified, quantifying them is very difficult. NCIC direct costs have increased from $2.9 million in fiscal year 1972 to an estimated $6.1 million in fiscal year 1981. Because the FBI Identification Division (Ident) is indispensable to the operation of the CCH file, a portion of Idents costs for criminal file activities should also be allocated as an NCIC cost. The FBI has not estimated what this allocation might be. Since fiscal year 1970, the Federal Government through LEAA has provided about $207 million in categorical grants to the States for comprehensive data systems and statistical 51

PAGE 62

52 l An Assessment of Alternatives for a National Computerized Criminal History System programs. About $39 million of this total was Federal grants cover only part of the costs for CCH-related systems. However, these of implementing and operating State CCH sysgrants peaked in 1976 and ended in fiscal year terns. Estimates of the full costs vary widely 1981. In addition, some portion of LEAA and have not been independently verified by block grants to the States were used for crimOTA. inal information, telecommunications, and record systems. NCIC Technolog y Hardware Previousl y Used by NCI C Until 1980, the FBI leased two IBM 360/65 computers, first marketed in 1965, from the Itel Corp. for use by NCIC. Each 360/65 had 2 million characters of memory. The second computer was normally used by the FBI to meet internal requirements for batch processing and by NCIC as a backup for the first computer. NCIC peripheral equipment included a number of tape and disc memory storage devices leased from various vendors and two IBM 2703 nonprogramable communication controllers. These controllers managed the NCIC communication lines. Only one was active at any one time, with the second serving as a backup. The age of the main computers and the peripheral devices, the technology on which they were based, and the design limitations of this equipment all had implications for the operation of NCIC. Age: Significant hardware difficulties experienced by NCIC suggested that the central computers were reaching the point where maintaining the units was no longer cost effective. The vendor offered to replace the installed 360/65 equipment at no cost to the FBI because the cost of maintaining these obsolete computers had become excessive. Technology: The IBM 360/65 used core memory, which is an obsolete technology. Although core memories are still used by computers with useful life remaining, the memories in modern machines are based on solid-state technology and are cheaper, require less power, and are more reliable. Design: An IBM 360/65 is not particularly efficient when used for applications that require the data bases to be accessed randomly, as is the case with NCIC. In addition, this computer was not designed to support NCIC teleprocessing needs. Because the communication controller was not programmable, it could not continue operation if the main computer failed. Thus, the match between the NCIC applications and the 360/65 computers was not a good one. Hardware Upgrade an d Message Switchin g The computer and communication technologies used by NCIC have been the subject of considerable controversy in recent years. Part of the controversy relates to fundamental concerns about the need for and impacts of NCIC, and particularly the CCH fileconcerns that have been intertwined with technology issues. Operational problems experienced with the NCIC system, such as unscheduled computer downtime and delayed response to inquiries, had been increasing due to the use of relatively old equipment. But the need for updated equipment conflicted with concerns expressed in Congress and elsewhere that new computer technology would make it easier for NCIC to engage in such unauthorized functions as message switching (i.e., the transfer or switching of messages from one State to another through the NCIC computer).

PAGE 63

Ch. 5NCIC Technology and Costs l 5 3 To a significant extent, the message switching controversy has served as a technological focal point for several of the policy issues discussed later in this report. Among these issues are whether an NCIC message switching function would place the FBI in an inappropriate (and perhaps illegal) role with respect to State and local criminal justice activities; and whether NCIC message switching would make it at least technically possible for the FBI to monitor and perhaps control criminal justice communications among the States. Over the last decade, the question of NCIC message switching has involved strong differences of opinion among the FBI, LEAA, the Attorney Generals office, the Office of Management and Budget, various congressional committees, a number of States, and other groups such as SEARCH Group, Inc., and the American Civil Liberties Union. 1 As a result, Congress denied to the FBI the authority for NCIC to perform message switching. A restriction placed in the Department of Justice Appropriation Authorization Act prohibited the use of funds for message switching. In this context, message switching was defined as the technique of receiving a message, storing it in a computer until the proper outgoing line is available, and then retransmitting, with no direct connection between the incoming and outgoing lines. More specifically, the Department of Justice (DOJ), including the FBI, was prohibited, absent approval of the House and Senate Judiciary Committees of Congress, from utilizing equipment to create a message switching system linking State and local law enforcement data banks through equipment under the control of DOJ or the FBI. 3 1 For a detailed discussion of the message switching controversy, see Donald A. Marchand, et al., A History and Background Assessment of the National Crime Information Center and Computerized Criminal History Program, Bureau of Governmental Research and Service, University of South Carolina, June 1979, sec. IV, The Message-Switching Issue and the Recent Policy Debate Over the CCH Program, pp. 122-167. 2 See U.S. Congress, House of Representatives, House Report 96-628, 96th Cong., NO\. 16, 1979. 3 Ibid. In mid-1979, the FBI sought approval from Congress for acquisition of a new front-end processor (also known as a communications controller), a device that controls and manages the flow of messages into and out of the computer. In order to evaluate this request, the Senate Judiciary Committee asked both the Office of Technology Assessment (OTA) and the Institute for Law and Social Research (INSLAW) for an outside analysis. 4 OTA and INSLAW agreed that the existing NCIC communications controller was obsolete and that upgraded equipment would improve NCIC performance. However, both OTA and INSLAW noted that the capability for message switching is inherent in a state-of-theart communications controller and message switching could be implemented at any time through software modification. Therefore, the use of the new controller would need to be governed by management, procedural, and perhaps legal means. 5 In September 1979, the FBI received authority from the Senate Judiciary Committee to proceed with the acquisition of a new NCIC communications controller. However, the authorization was conditioned on the FBIs commitment: 1) to lease (for a period no longer than 2 years) rather than buy the controller; 2) not to acquire a message switching option with the controller or to message switch data between States; 3) to program the system to assign requests for data from the NCIC/CCH file the very lowest priority; and 4) to permit biannual GAO audits to ensure compliance. Later in 1979, the FBI requested authority to upgrade the NCIC host computer. An OTA analysis found that (t)he central processor (host computer) used for NCIC is the IBM 360/65, an early third generation machine that June 26, 1979, letter from the Chairman of the Senate Committee on the Judiciary to the OTA Director. See Aug. 10, 1979, letter and attachment from the OTA Director to the Chairman of the Senate Committee on the Judiciary; and July 30, 1979, letter and enclosure from the INSLAW President to the Chairman of the Senate Committee on the Judiciary. 6 Sept. 12, 1979, letter from the Chairman of the Senate Committee on the Judiciary to the Director of the FBI.

PAGE 64

54 l An Assessment of Alternatives for a National Computerized Criminal History System .- was first marketed in 1965 and is now obsolete Operating statistics indicate that the processor is approaching the end of its useful life and may soon have to be replaced, even if no change is permitted in the character of its applications. Early in 1980, based in part on the OTA findings and with congressional concurrence, the FBI took action to upgrade the host computer, subject to the same conditions stipulated for the communications controller. 8 Status of Hardware an d Software Upgrad e In May 1980, the two obsolete computers were replaced with two National Advanced Systems (NAS) 9 AS/5000 computers. Each of the new computers has 4 million characters of memory, double that of the 360/65. Also, the disc drives have been upgraded and now provide more cost-effective storage of online data. In May 1981, new operating system software (multiple virtual storage) was installed. Finally, in October 1981, the two obsolete communication controllers were replaced with two CC I Model CC80 controllers. The upgrading of the host computers, disc drives, operating system, and the communication controllers substantially increased the computer power available to NCIC. These changes have improved the quality of service and minimized downtime. Unscheduled monthly downtime averaged about 1.6 percent for the 12-month period ending October 1981. 10 With respect to applications software, the FBI recognizes that NCIC application programs will also need to be upgraded at some future time. The use of a higher level programing language would enhance the maintainabilU.S. Congress, Office of Technology Assessment, draft paper on NCIC Technology, Sec. 1.2.1.1, Description of the System: Central Processor and Memory, January 1980. Jan. 11, 1980, letter from the FBI Director to the Senate Judiciary Committee Chairman. National Advanced Systems (NAS) acquired Itel, the previous vendor. For the months of November 1980 through October 1981, unscheduled downtime (in percent of hours in the month) was 1.1, 1.0, 0.5, 3.0, 1.1, 1.5, 1.5, 2.0, 2,4, 1.2, 1.6, and 1.7 percent. Data from NCIC Newsletters. ity of NCIC application programs and make it easier to recruit the necessary programing personnel. In addition, upgraded applications software would help reduce the substantial programing backlog that presently exists. As of December 1981, the NCIC staff listed 14 non-CCH and 11 CCH programing priorities, many of which are not scheduled for implementation until 1983 and beyond. 11 While there are no firm plans as yet to reprogram the NCIC applications software, the FBI has initiated a review of NCIC system needs for the next 5 years, including the possibility of a major system redesign with further hardware as well as software improvements. 12 By comparison, the Ident Automated Identification Division System (AIDS) uses hardware similar to that used by NCIC. However, the AIDS host computer (an NAS AS/5-3) is accessible only within the FBI headquarters building through 10 minicomputers (IV-Phase Model 4-70), whereas the NCIC host computer is accessible via 137 communication lines to Federal, State, and local criminal justice agencies. NCIC and AIDS peripheral equipment is generally comparable, except for equipment associated with the AIDS automated fingerprint reader systems. An AIDS operating system software upgrade to multiple virtual storage (already implemented by NCIC) is scheduled to be completed in the near future. 13 In July 1982, the FBI announced a major computer procurement. The two AS/5000 computers used by NCIC will be replaced with two IBM 3033S computers, and the AS/5-3 used by AIDS will be replaced with an IBM 3033N. In addition, the operating system will be upgraded to MVS/SPl.3. 14 1lNCIC staff paper prepared for the Dec. 9-10, 1981, meeting of the NCIC Advisory Policy Board, Topic #8, pp. 1-3. NCIC System Report presented by Kier Boyd of the FBI at the June 17-18, 1981, meeting of the NCIC Advisory Policy Board. For a more detailed technical description of AIDS and NCIC, see FBI, Interstate Identification Index: BackgrourId and Findings for July-September 1981 Phase I Pilot Project, Dec. 4, 1981, pp. 18-25. Telephone conversations with Kier Boyd and Gordon Zacrep of the FBI, July 30, 1982.

PAGE 65

Ch. 5NCIC Technology and Costs l 5 5 NCI C The total cost to the Nation of operating NCIC is shared by the FBI and the users. The FBI pays for the central computer facilities (including administrative, operational, and programing costs) and the communication links, while the users pay for the terminals and the costs of gathering, inputting, and processing the data at the Federal, State, and local levels. The Federal budget covers the costs to the FBI and to the Federal agencies that use NCIC. State and local budgets cover most of the remaining costs, although in the past the Federal Government has partially underwritten the costs to States and localities through grants from LEAA and others. The total cost of NCIC includes the indirect costs incurred by supporting activities and systems, as well as those costs directly chargeable. For example, the CCH file is heavily dependent on support from various State, local, and Federal agencies for its operation. Records can be entered into the CCH file only after an FBI number has been obtained, and only Ident can assign that number. Records can be entered solely by authorized criminal justice agencies, and State and local criminal history systems are major sources of entries. The costs of operating NCIC that are common to both the hot files and CCH file include: l l l l an allocated portion of the costs of the FBI computer facility and communication lines, including both hardware (on a lease or purchase basis) and operating personnel; the FBIs cost of developing and maintaining the NCIC software; the cost to State and local governments and Federal user agencies for terminal equipment and operators; and the cost of personnel in user agencies who record and format the data for NCIC input and processing. The following additional costs are incurred by the CCH file: l l l A portion of the FBIs cost of operating I dent. A portion of the costs incurred by State and local governments for operating their own criminal history record systems. Many of these systems are automated and include costs that are similar to those of the FBI computer center. A portion of the cost of the activities of the courts, correctional authorities, and other criminal justice agencies to support criminal history record systems on which NCIC depends. Although the components of NCIC costs can be identified, it is very difficult to quantify them. For example, the costs to the FBI of operating computer and communication facilities are broken down by organizational unit, not by function or programmatic activity. Thus, while OTA has been able to identify the direct costs of NCIC (i.e., the costs of the NCIC section and related technical support within the FBI Technical Services Division), indirect costs such as those incurred by Ident to support the CCH file are not readily identifiable. Similarly, at the State level the funding for operating criminal justice information systems comes from a variety of sources and is not broken down by function. While OTA has documented the level of Federal funding to the States through LEAA categorical grants for CCH and related activities, the portion of State revenues and Federal block grants devoted to hot files or CCH files is not known. OTA has not attempted to quantify costs at either State or local levels. Costs to the FB I The costs of the NCIC section and related automated data processing (ADP) and telecommunications support are shown in table 7. Over the 10-year period from fiscal year 1972 through fiscal year 1981, NCIC costs have increased about 110 percent at an aver-

PAGE 66

56 l An Assessment of Alternatives for a National Computerized Criminal History System Table 7.NCIC Direct Costs, Fiscal Years 1972.81 (thousands of dollars) FY FY FY FY FY FY FY FY FY FY Personnel (manpower): NCIC section Agent work years a . . . . . 6 10 11 14 15 12 9 9 8 8 Support work years. . . . . 24 42 39 106 102 111 96 99 99 99 Total work years . . . . . 30 52 50 120 117 123 105 108 107 107 ADP and telecommunications support Operations work years . . . . N/A N/A N/A 17 17 19 19 19 20 20 System maintenance work years . N/A N/A N/A 5 5 6 6 7 8 8 System development work years . N/A N/A N/A 7 8 8 Total work years . . . . . 22 22 25 25 33 36 36 Personnel (costs): NCIC section . . . . . . $360 $684 $735 $1,476 $1,556 $1,713 $1,624 $1,757 $1,644 $1,800 ADP and telecommunications support . N/A N/A N/A 396 418 475 500 660 800 900 Nonpersonnel costs: NCIC section . . . . . . N/A N/A N/A 155 143 165 163 169 170 180 ADP and telecommunications Support b NCIC telecommunications network services . . . . . . 900 900 940 1,000 1,000 1,000 1,000 1,000 1,000 1,000 FBI NC!C terminals . . . . 130 130 130 140 140 140 110 100 100 100 Computer center space . . . N/A N/A N/A 140 180 180 200 200 225 200 ADPE rental and maintenance . . 1,100 1,100 1,200 1,300 1,400 1,500 1,610 1,610 1,543 1,960 Equipment purchase . . . . 750 c Estimated costs d . . . . . 439 479 530 Total NCIC costs . . . . $2,929 $3,193 $3,535 $4,607 $4,837 $5,173 $5,207 6,246 $5,482 6,140 aA work year is defined as an equivalent full-year employee. bADP and telecommunication SU pp O rt ~ resources are estimated as a prorated portion of FBI Computer Center resources. Except for the $750,000 for purchase of replacement telecommunications control equipment In FY 1979, all ADP/telecommunication funding for NCIC is implicitly included in the FBIs centralized ADP/telecommunt. cation budget. cFunds allocated for the replacement of telecommunication control equipment, but never spent d Estimates of costs for Which data were not available. Computed by dividing the sum of the costs for which data were available for each of the Years estimated by 0.85, a factor derived from the data that were available for fiscal years 1975 through 1979. SOURCE Office of Technology Assessment and Federal Bureau of Investigation age rate of about 9 percent annually. However, personnel costs for the NCIC section have increased almost 400 percent. There has also been a sizable, although not as large, increase in personnel costs for ADP and telecommunications support. Other costs have increased at or below the rate of inflation. Since Ident is indispensable to the operation of the CCH file, a portion of Idents costs for criminal justice activities (which totaled about $58.7 million in fiscal year 1980) should be allocated as an NCIC cost. The FBI has not estimated what this allocation might be. Costs to the State s Some portion of the expenditures of State agencies for hot files and CCH use should be included in the overall costs of NCIC. However, there is really no sound basis on which to identify this portion. First, as noted in chapter 4, there is wide variation in the levels of State participation in NCIC. While all States can contribute to the hot files, only eight are currently permitted to contribute records to the CCH file. Some States make relatively heavy use of NCIC, while others use it only minimally. Thus, the exact basis for allocating costs to NCICeven if known would be difficult to determine. Second, the level of automation of State criminal justice information systems also varies widely. Therefore, the costs to the States are also likely to be highly variable, depending on the type of system. Third, some portion of State funding has been provided by the Federal Government, primarily through LEAA. Since fiscal year 1970, LEAA has provided about $207 million in categorical grants to the States for comprehensive data systems and statistical programs, as shown in table 8.

PAGE 67

Ch. 5NCIC Technology and Costs l 5 7 Table 8.LEAA Grants to States for Comprehensive Data Systems and Statistical Programs, Total by Fiscal Year Fiscal year Total amount 1969 ., . . . . ... $ 0 1970 ... . . . . . . 1,000,000 1971 . . . . . . . 4,000,000 1972 . . . . . . . 9,700,000 1973 21,200,000 1974 ....:::::::::::: :::::::::: : 24,000,000 1975 . . . . . . . 26,000,000 1976 . . . . . . . 31,622,000 1977 . . . . . . . 21,152,000 1978 . . . . . . . 16,000,000 1979 . . . . . . . 21,290,000 1980 . . . . . . . 15,000,000 1981 . . . . . . . 16,275,000 Total . . . . . . . $207,239,000 a Includes transition quarter grants totaling $6 mllllon SOURCE Office of Justice Assistance Research and Statlstlcs, U S Department of JustIce About $39 million of this amount has been for CCH-related systems, including offenderbased transaction systems (OBTS) to record key events about individuals as they pass through the criminal justice process. As indicated in table 9, CCH grants peaked in 1976 and ended in 1981. From 1970 to 1980, 145 CCH-related grants were awarded to 35 different States. However, 10 States receiving 57 grants accounted for about two-thirds of the total funds awarded to all States. In addition, some portion of LEAA block grants to the States was used for criminal justice systems. While OTA did not attempt to estimate this amount, an independent analysis provided to OTA suggests that an additional $200 million to $400 million of block grants funds were spent on law enforcement telecommunications systems, criminal justice agency record systems, and criminal justice information systems. 15 Federal grants account for only part of the cost of implementing and operating a States CCH files. Estimates of the full costs vary widely and have not been independently verified by OTA.A 1975 study by lNSLAWestimated the total cost of a fully developed CCH (in which all States were full participants) to be $361 million in current dollars (adjusted for inflation over the 10-year development period 1975-84). Of this total, the State share was es- . Data and analysis provided to OTA by Sept. 9, 1981, letter and enclosures from Tom Dalton of Seattle University, Table 9. LEAA Grants to States for CCH-Related Systems, a Total by Fiscal Year and by State Fiscal year Number of grants Total amount 1969 .., . ... 0 $ 0 1970 ......., . 2 123,975 1971 . . . 0 0 1972 . . . 10 2,714,105 1973 . . . 5 1,379,531 1974 . . . 21 5,875,968 1975 . . . 17 7,068,913 1976 ......, . 36 9,931,835 1977 . . . 24 4,545,147 1978 ....., . . 15 4,726,194 1979 ...., . . 12 2,650,308 1980 ..., ... ... 3 274,756 1981 . . . 0 0 Total . . . 145 $39,290,732 aIncludirlg offender-based transaction statistics and computerized crlmlnal history systems SOURCE Office of Justice Assistance, Research, and Statlstics, US Department of Justice timated at $274 million ($44 million for development and $230 million for operation) and the Federal share at $87 million (for FBI/Ident and NCIC/CCH). 16 A 1979 study by the National Center for State Courts (NCSC) attempted to determine development and operational costs for State CCH systems. The results were fragmentary, but provided a basis for concluding that the 1975 estimates were probably low. For example, INSLAW projected that 15 States would be full CCH participants in 1978, with a combined CCH/OBTS operational cost in that year of $17.7 million in 1978 dollars. By comparison, NCSC found that the actual 1978 CCH/ OBTS operational costs for the 15 States listed in table 10 were reported to total about $42 million, 18 more than double the INSLAW projection. Some of the difference may be attributable to assumptions about the rate of inflation (assumed by INSLAW to be 26 percent over the 3-year period 1975 to 1978). 19 However, the comparison does suggest that operating costs in 1978 were significantly higher than projected by INSLAW. .. .Institute for Law and Social Research, Costs and Benefits of the Comprehensive Data Svstem Program, prepared for I.EAA, June 1975, vol. 1: Summary, pp. 12, 26. Ibid., pp. 12, 25, 26, National Center for State Courts, A Re\iew of OB7S and C(7H Program Requirements in the Judiciarjr, Williamsburg, \a., 1979, pp. 86, 131. ]nstitute for I,aw and Social Research, Costs and Benefits of the Comprehensi\e Data System Program, prepared for I. I? AA, June 1975, vol. I: Summary.

PAGE 68

58 l An Assessment of Alternatives for a National Computerized Criminal History System Table 10.CCH/OBTS Operational Costs a for 1978 by State (in 1978 dollars) State Operating cost Alabama. . . . . . . . $ 480,000 Arizona. . . . . . ., . 5,000,000 California. . . . . . . 14,826,000 Delaware . . . . . ., . 1,241,000 District of Columbia . . . . 1,500,000 Hawaii . . . . . . . 1,000,000 Maryland . . . . . . . 2,000,000 Michigan . . . . . . 1,731,000 Minnesota . . . . . . 1,250,000 Missouri . . . . . . . 600,000 New York..... . . . . . 8,000,000 North Carolina . . . . . 2,500,000 Oklahoma . . . . . . 200,000 Texas . . . . . . . 1,097,000 Virginia . . . . . . . 195,000 15-State total . . . . . $41,620,000 a includes combined costs for Computerized Criminal History Systems and Offender-Based Transaction Systems (OBTS) SOURCE National Center for State Courts, A Review of OBTS and CCH Prograrn Requirements in the Judiciary, 1979, pp 06, 131 In general, NCSC encountered great difficulty in obtaining reliable cost data for the study. State operating personnel were frequently unable to reconstruct the requested data from available records. The operation of CCH and OBTS systems was often so intertwined with that of other criminal justice systems that the proper allocation of costs was almost impossible. In addition, some portion of State identification bureau costs (estimated at $60 million for fiscal year 1980 20 ) should be allocated as an NCIC cost. Finally, no systematic data were available on the costs to the various localities of preparing and submitting CCH/OBTS information for use by State systems. ~International Association for Identification, RmctionalRe quirements and Systems Development Plan for State Identification Bureaus: Executive Summary of Findings and Recommendations, Utica, NY., October 1980, p. 1.

PAGE 69

Chapter 6 Legal/Regulatory Framework for NCIC, Ident, and State CCH Systems

PAGE 70

Contents Chapter Summary. . . . . . . . . . . . Federal Statutes and Regulations.. . . . . . . . . Authority to Operate Ident and NCIC . . . . . . . Record Content . . . . . . . . . . . Record Updating . . . . . . . . . . . Record Dissemination. . . . . . . . . . . Freedom of Information and Privacy Act . . . . . . NCIC Operating Policies and Procedures . . . . . . Selected NCIC Hot File Operating Procedures . . . . . Selected CCH File Operating Procedures . . . . . . Federal Agency Orders or Procedures . . . . . . . Federal and State Court Rulings . . . . . . . . State Statutes and Regulations . . . . . . . . Early Efforts of Project SEARCH and LEAA . . . . . Implementing LEAA Privacy and Security Regulations . . . State Statutes and Regulations as of June 1981 . . . . . Initiatives to Enact Comprehensive Federal Legislation . . . Page 61 62 62 62 63 64 65 66 66 66 67 67 69 69 70 71 73 TABLES Table No. Page 11. Federal Agency Orders or Procedures for NCIC . . . . 67 12. Illustrative Federal/State Court Rulings on Criminal Records . 68 13. Categories of State Statutes and Regulations . . . . . 72 14. Survey Comparison of Changes in State Statutes/Regulations by Category . . . . . . . . . . . . . 73

PAGE 71

Chapter 6 Legal/Regulatory Framework fo r NCIC, Ident, and State CCH Systems Chapte r The management and use of criminal justice information in the United States are governed by a variety of Federal, State, and local statutes, regulations, and executive (or agency) orders, and Federal and State court rulings. Overall, Federal statutes and regulations have: l l l l granted basic authority to the Attorney General and the Federal Bureau of Investigation (FBI) for operating its Identification Division (Ident) and the National Crime Information Center (NCIC); established standards for use of the various FBI criminal justice information systems. defined a range of authorized users of Federal systems (e.g., the Office of Personnel Management for Federal employee background checks by authority of Executive Order No. 10450); and established standards for use of State criminal history systems funded in whole or in part by the Law Enforcement Assistance Administration (LEAA). During the 1970s, LEAA funding and the pioneering research of Project SEARCH (System for Electronic Analysis and Retrieval of Criminal Histories) played a significant role in stimulating substantial progress in development of State statutes and regulations for use of State criminal record systems. However, direct LEAA funding has now ended, and with it the option of terminating Federal funds for noncompliance (the primary penalty authorized by Congress). In general, Congress has provided broad discretion to the FBI and LEAA in defining standards for the interstate collection, mainteSummar y nance, and dissemination of criminal history information. Until the 1970s, title 28, United States Code (USC), sec. 534 provided the sole legislative direction at the Federal level. Congressional initiatives to pass comprehensive criminal justice information legislation in the early 1970s were not successful, but led to the enactment of an amendment to the Crime Control Act of 1973 requiring LEAA to issue detailed privacy and security regulations for State and local criminal history information systems (which appear as title 28, Code of Federal Regulations (CFR), pt. 20, subpt. B). Regulations for Federal systems (Ident and NCIC/ CCH) and the interstate exchange of criminal history record information are set forth in title 28, CFR, part 20, subpart C. The responsibility for enforcing management and use standards for criminal justice information is left largely up to the States, localities, and other users. For example, while LEAA regulations established standards for record quality and security, LEAA conducted little actual monitoring of State compliance, but did engage in an active program of publishing guidelines, model State codes, and the like. While the FBI is authorized to terminate Ident and/or NCIC services to users who violate regulations, compliance is largely voluntary. At present, the FBI program to monitor compliance includes computer edits and quality checks of information from contributing agencies that is maintained in FBI files, but does not include direct audits of State or user files and record disseminations. Indeed, the FBI has never had the authority to conduct such audits. 61

PAGE 72

62 An Assessment of Alternatives for a National Computerized Criminal History System In the early 1970s, efforts to enact comprehensive legislation, such as the Criminal Justice Information Systems Security and Privacy Act of 1971 or the Criminal Justice InAct of 1974, were not successful; nor were initiatives in the late 1970s to include criminal justice information system standards as part of the proposed FBI charter legislation. formation Control and Protection of Privacy Federal Statutes and Regulation s Authority to Operat e Ident and NCI C The FBI has statutory authority to establish and maintain criminal history files in Ident and NCIC. (28 USC 534 (1968)). In part, this statute authorizes the Attorney General to acquire, collect, classify, and preserve criminal identification, crime, and other records, and to exchange them with authorized officials of Federal, State, and local law enforcement agencies, and with penal and other institutions. The Attorney General has delegated this authority to the Director of the FBI in title 28, CFR, section 0.85. In addition, a 1973 amendment to the Omnibus Crime Control and Safe Streets Act of 1968, Public Law No. 90-351, 82 Stat. 200 (1968), adding a section 524 (42 USC 3771), directs the executive branch to assure the adequate provision of privacy and security of criminal history information (reorganized by Public Law No. 96-157, 818,93 Stat. 1212 (1979) as 42 USC 3789g (Supp. 1980)). The privacy and security regulations in 28 CFR part 20 (1975) were issued pursuant to this congressional directive. Record Conten t The information that may be stored in criminal history records maintained by Ident and NCIC is described in 28 CFR 20.2 (1975), and includes identifiable descriptions and notations of arrests, detentions, indictments, informations, or other formal criminal charges, and This section is based on app. A to Jet Propulsion Laboratory, FBI Fingerprint Identification Automation Study: AIDS III Evaluation Report Volume VI: Environmental Analysis, California Institute of Technology, Pasadena, Nov. 15, 1980, prepared for the L S. Department of Justice, Federal Bureau of Investigation. any disposition arising therefrom, and details as to sentencing, correctional supervision, and release. Only information on serious and/or significant offenses may be stored in these records (28 CFR 20.32, 1975). Specifically excluded are the nonserious offenses of drunkenness, vagrancy, disturbing the peace, curfew violation, loitering, false fire alarm, nonspecific charges of suspicion or investigation, and traffic violations (other than manslaughter, driving under the influence of drugs or liquor, and hit and run). Offenses committed by juvenile offenders are also specifically excluded unless the juvenile is tried in court as an adult. Menard v. Saxbe, 498 F.2d 1017 (1974) resulted in judicial examination of the content of the FBIs criminal history files. It involved a suit against the FBI for expungement of a State (California) arrest record retained by the FBI. It had been established at the State level that there was no probable cause for the arrest, and the status of the proceeding was changed from arrest to detention. The FBI had been so notified, and had amended its record to show that the subject encounter with the police was not considered to be an arrest under California law, and that no formal proceedings had been brought. The court determined that once the FBI was notified that the subject was not involved in the criminal justice process, it had no authority to retain the record in the criminal files, even though the record accurately portrayed the events as they had occurred. The controlling statute (28 USC 534) only authorizes the storage of information about formal criminal proceedings in the criminal files. The court stated that the FBI has a responsibility to assure that it does not disseminate criminal records containing inappropriate information. The decision was

PAGE 73

Ch. 6Legal/Regulatory Framework for NCIC, Ident, and State CCH Systems l 63 carefully grounded on statutory considerations, but the court left as an open question the extent to which this decision is mandated by the U.S. Constitution. One unresolved problem that arises from this decision is what to do with the fingerprints of suspects who undergo pretrial diversion. This alternative to the usual judicial process is sometimes used when the U.S. Attorney determines that the suspects infraction of the law was due to an unfortunate set of circumstances and is not likely to be repeated. Sometimes the suspect is formally arrested, sometimes not. Instead of going through the usual criminal process, the suspect agrees to a set of conditions, which usually involve some type of restitution to the victim and a period of probation. If these obligations are successfully fulfilled, the charges are either dismissed or never brought. The unresolved question is whether pretrial diversion qualifies as a formal criminal process under 28 USC 534 when the suspect is not actually arrested. The FBI retains such records now, but its authority to do so is uncertain after the Menard decision. The FBI has requested legislative direction in this matter, but so far none has been forthcoming. Record Updatin g On May 20, 1975, the Department of Justice (DOJ) issued a regulation prohibiting dissemination of arrest information more than a year old unless accompanied by a disposition when no active prosecution of the charge is known to be pending (28 CFR 20.33, 1975). The prohibition does not apply to records released for criminal justice purposes or to authorized Federal agencies. It came in the wake of Tarleton v. Saxbe 507 F.2d 1116 (1974) in which the court expressed concern about the impairment of an individuals liberty that results when that person stands accused of a crime. It noted that the reason for the constitutional guarantee of a speedy trial is to mitigate this restriction of the accuseds liberty, and the court suggested that the lower court inquire into what justifications, if any, exist for the FBIs failure to indicate dispositions within a reasonable time after arrest. Two years later, the district court order in Tarleton v. Saxbe 407 F. Supp. 1083 (1976) directed the FBI to conduct a feasibility study of system procedures that would enable it to keep disposition entries in its criminal records reasonably current. By the time the study was conducted, the FBI had solved the immediate problem by promulgating 28 CFR 20.33 (1975). Most of the systems and procedures suggested by the study for keeping the disposition data more current were designed for use in a computerized system. Regulation 28 CFR 20.37 (1975) makes it the responsibility of each criminal justice agency contributing data to FBI criminal history record information systems to assure that information is kept complete, accurate, and current. It calls for a disposition to be submitted within 120 days after it has occurred. However, the only sanction available for enforcing this policy is regulation 28 CFR 20.38 (1975) that permits DOJ to cancel its criminal record services to any agency that fails to comply with its regulations. Pursuant to 28 CFR 20.32 (1975), Ident and NCIC do not record minor and juvenile offenses. Although this regulation went into effect in June 1975, NCIC has had such a policy since November 29, 1971, and Ident since February 9, 1973. The regulation itself does not require the FBI to expunge information on minor offenses previously compiled. However, the district courts order in Tarleton v. Saxbe 407 F. Supp. 1083 (1976) required the FBI to delete from the record, prior to dissemination, all information relating to nonserious offenses. The FBI is deleting these offenses from requested records as they are sent out. The FBI currently expunges and seals records pursuant to State and Federal court orders. The authority for sealing the record of a person who has been found guilty of unlawful possession of a controlled substance is found in 21 USC 844(b)(1) (1972). If the subject individual has not previously been convicted of violating any Federal narcotics laws, the court may, after trial or entry of a guilty plea, place the person on probation without entering a

PAGE 74

64 l An Assessment of Alternatives for a National Computerized Criminal History System judgment of guilty. If the person does not violate any conditions of the probation, the court may dismiss the proceedings. DOJ retains a record solely to determine first offender status. As of July 1981, 35 States provide procedures whereby subjects can have nonconvictions purged from their records, and 24 provide a procedure for purging records of convictions. Twenty States provide for sealing of records of nonconviction and 22 provide for sealing of convictions. 2 For example, Arkansas provides for purging all records relating to a crime wherein the person has been acquitted or the charges dismissed (Ark. Stat. Ann. 5-1109, 1975). This State also provides for the sequestering of records of first offenders so that they are available only to law enforcement and judicial officials (Ark. Stat. Ann. 43-1231, 1975). When either procedure takes place, the court sends a copy of the order to the Arkansas State Identification Division and the FBI Identification Division. In comparison, the California Penal Code allows a defendant who has been acquitted to file a motion to seal rather than purge the record of arrest and acquittal (Cal. Penal Code 851.8, Deering Supp., 1980). As in Arkansas, a copy of the judges order sealing the record is forwarded to law enforcement agencies, including the FBI. On September 24, 1973, DOJ instituted (by DOJ Order 556-73) a procedure by which individuals, upon request and verification of identity, may review the criminal history information maintained on them. Individuals may apply to the contributor of the information to make any changes in the record. If the contributor corrects the record it must notify the FBI, and the FBI will make any changes necessary in accordance with the corrections (28 CFR 20.34, 1975). Record Disseminatio n Recipients of criminal history information are limited by 28 USC 534 (1968) to law en2 SEARCH Group, Inc., !fkds in State Security and Privacy Legislation, Sacramento, Calif., November 1981, p. 5, prepared for the U.S. Department of Justice, Bureau of Justice Statistics. forcement agencies, penal, and other institutions. In 1971, the district court for the District of Columbia, in deciding Menard v. Mitchell 328 F. Supp. 718 (1971), held that other institutions refer to other official criminal justice and law enforcement institutions only. Prior to this decision, the FBI had been providing criminal history records to States for employment and licensing checks. Immediately after this decision, Congress responded by passing the Departments of State, Justice, and Commerce, the Judiciary, and Related Agencies Appropriation Act, 1973, Public Law No. 92-544, 2, 86 Stat. 1109 (1972) allowing the FBI to disseminate criminal history information to officials of federally chartered or insured banking institutions. Public Law No. 92-544 also permits dissemination to State and local government agencies for purposes of employment and licensing if the check is authorized by a Federal or State statute and approved by the Attorney General. Then, in 1975 Congress amended the Securities Exchange Act, 17 (15 USC 78q(F)(2)) to require every member of a national securities exchange, and every broker, dealer, registered transfer agent, and registered clearinghouse agency to undergo an FBI criminal history check. The dissemination of criminal histories to authorized Federal agencies is permitted pursuant to Federal statute or Executive order, 28 CFR 20.33(2) (1975). For example, Executive order 10450 requires a national security investigation of prospective civilian officers or employees in any department or agency of the Federal Government. In most cases the investigation includes at least a national agency check (including a check of FBI files) and written inquiries to local law enforcement agencies. In effect, the order authorizes dissemination of criminal history record information to Federal agencies for use in background investigations, whether national agency checks or full field investigations. This authority has also been established for military employees or applicants (Executive order 12065) and for certain employees of defense contractors (Executive order 10865).

PAGE 75

Ch. 6 Legal/Regulatory Framework for NCIC, Ident, and State CCH Systems l 65 As a consequence of Menard v. Mitchell, DOJ has strictly construed the statutes governing dissemination of criminal history files. It has revised its earlier position under 28 USC 534 (1968) and now refuses to allow access, directly or through State law enforcement agencies, to railroad police and campus police. Even though these groups may be authorized by State statute to investigate crimes or apprehend criminals, DOJ does not find them to be authorized Government officials under the meaning of 28 USC 534. It has also refused, under Public Law 92-544, to provide criminal history records to State boards of bar examiners when the board is established by rule of the State supreme court rather than by a statute. Once the criminal history records leave the FBIs control, one sanction available to enforce FBI dissemination policies is 28 CFR 20.33 (b)(1975). This regulation provides that the exchange of criminal history record information with authorized recipients is subject to cancellation if dissemination is made outside the receiving department or related agencies. Also, certain civil and criminal penalties are provided under the Privacy Act of 1974. Freedom of Informatio n and Privacy Ac t Under the Freedom of Information Act, Public Law No. 89-487, 80 Stat. 250 (codified at 5 USC 552, 1977), all Government agencies are required to supply copies of their records to any member of the public who requests them (5 USC 552(a)(3)). It has been established that this act applies to computer tapes to the same extent that it applies to other records (Long v. U.S. IRS 596 F.2d 362, 1979). However, the act provides several categories of exemptions: 1) matters that are exempt under another statute, if the statute leaves the agency no discretion or supplies particular criteria for applying the exemption, may be withheld from the public (5 USC 552(b)(3)); 2) if disclosure of a file would constitute a clearly unwarranted invasion of personal privacy it need not be disclosed (5 USC 552(b)(6)); and 3) investigatory records compiled for law enforcement purposes are exempt if release would constitute an unwarranted invasion of personal privacy (5 USC 552( b)(6C). Note that the privacy standard for these records is less strict than the privacy standard for other records. There are other exemptions covering law enforcement records, but they are of limited application (see 5 USC 552(b)(7)). If the agency invokes any one of these exceptions, it must release any reasonably separable portion after deleting the exempt portions. Regulations promulgated pursuant to this statute allow the Attorney General to exempt the whole system of FBI criminal records from public disclosure. This exemption, which is noted in DOJ regulations (28 CFR 16.10, 1973), is uniformly applied to exempt all criminal histories from disclosure. The Privacy Act of 1974 (Public Law 93-579 codified in part at 5 USC 552a, 1977) was passed shortly after the Freedom of Information Act. Its purpose is to protect the privacy interests of individuals by regulating the collection, maintenance, use, and dissemination of personal information by Federal agencies. The Privacy Act requirements apply to all Federal agency systems including Ident and NCIC, except where the head of an agency (in this case the Attorney General) may exercise certain exemptions for systems of records maintained for the enforcement of criminal laws. The Attorney General has exercised specific exemptions, particularly for access and challenge procedures. However, alternate procedures are provided in 28 CFR 20.34, which establishes the right of individuals to have access to and review their own criminal history record information maintained by Ident or NCIC, and to seek correction by the source agency if the information is believed to be incorrect or incomplete. Individuals may also direct a record challenge to the FBI, who will then forward the challenge to the source agency. The FBI will make any changes necessary in the Ident or NCIC files if proper notification is received from the source agency.

PAGE 76

66 l An Assessment of Alternatives for a National Computerized criminal History System NCIC Operating Policies and Procedure s The FBI Director has approved a set of NCIC operating policies and procedures; these embody the statutory-regulatory framework discussed above, but go considerably further in some areas. The policies and procedures are based in part on recommendations from the NCIC Advisory Policy Board, and are included in the NCIC Operating Manual distributed to NCIC terminal operators. The manual is updated and revised periodically as needed. Selected NCIC Hot Fil e Operating Procedure s Each record in an NCIC file is identified with the originating agency. The NCIC Operating Manual emphasizes repeatedly that agencies that enter records into NCIC are responsible for record accuracy, timeliness, and completeness. The FBI does assume responsibility for those records entered by the FBI. In addition, the FBIas system managerhelps maintain the integrity of the system through: 1) automatic computer edits that reject certain types of errors in data; 2) automatic purging of records after they are on file for a prescribed period of time; 3) quality control checks by FBI personnel; and 4) periodically furnishing lists of all records on file for validation by the originating agencies. The manual also emphasizes that an NCIC hit may not be probable cause for arrest. NCIC only provides one more piece of information to be evaluated by the officer along with other facts in determining if there is sufficient legal basis for probable cause to arrest a person or seize property. 5 An immediate confirmation with the originating agency is necessary to ensure the validity of the hit before an arrest or seizure is made. The manual points out that NCIC is an informational tool. It is no substitute for professional judgment. 3 NCIC Operating Manual Introduction, p. 7. Ibid. 5 1 bid., p. 2. NCIC information must be evaluated along with other facts known to the criminal justice official. Finally, NCIC procedures place some limitations on what can be entered into files. For example, before entering a record into the wanted persons file, the entering agency is required to determine, to the maximum extent possible, if extradition will be authorized. If not, the record should not be entered. G The manual further provides detailed procedures for correcting errors and for sending and receiving messages with the various hot files. Selected CCH Fil e Operating Procedure s As with the NCIC hot files, each criminal justice agency contributing data to CCH is responsible for assuring that information on individuals is kept complete, accurate, and current. For all arrest data included in such records, disposition data should also be included to the maximum extent feasible and submitted to CCH within 120 days after the disposition has occurred. 7 Unlike the hot files, CCH operating procedures require that all criminal justice agencies seeking direct access to CCH execute a written agreement with the FBI Director. This agreement commits the agency to abide by all CCH rules, policies, and procedures. 8 These procedures were approved by the NCIC Advisory Panel Board and adopted by the FBI Director. The CCH operating procedures specify the kinds of criminal history information that may be entered into the CCH file, require continuous checks by the FBI and States on the accuracy of records in the file, and define the right of an individual with a record in the CCH file to review that record and seek correction if the information is believed to be inaccurate or in. Ibid., p. 7. 7 Ibid., pt. 10, p. 7, same as 28 CFR 20.37. Ibid., same as 28 CFR 20.36.

PAGE 77

Ch. 6Legal/Regulatory Framework for NCIC, /dent, and State CCH Systems l 67 complete. In addition, they also define who may have direct access to CCH records and the limitations on the use of such records. With respect to system security, systems that interface directly with NCIC are required to be under the management control of criminal justice agencies. The procedures also establish a set of physical, technical, and personnel security measures required of all agencies having access to CCH. These measures include logging all transactions against the CCH file, screening and verifying all CCH inquiries, placing all terminals in secure locations, and screening all terminal operators. Finally, the procedures define the role of the NCIC Advisory Policy Board, particularly with regard to establishing criteria for purging records, for secondary access to CCH, and for the organization and administration of CCH. With respect to the last, all rules govemTable 11 .Federal Agency ing direct terminal access to the CCH file apply equally to Federal and State agencies. In addition, such agencies must permit an Advisory Board-appointed inspection team to conduct inquiries concerning any alleged security violations. 9 Federal Agency Order s or Procedure s In addition to the NCIC operating policies and procedures, some Federal agencies have their own orders or procedures for using NCIC. OTA conducted a partial survey of Federal users to identify the range of operating policies and procedures that govern the use of NCIC. Illustrative results of this survey are summarized in table 11. For further details, see Ibid., pp. 15-27. Orders or Procedures for NCIC Agency Policy/procedure Bureau of Indian Affairs (BIA) U.S. Department of the lnterior a Internal Revenue Service (IRS) U.S. Department of the Treasury b Postal Inspection Service U.S. Postal Service c Federal Prison System (FPS) U.S. Department of Justice d United States Marshals Service U.S. Department of Justice e None. Adheres to policies and procedures of agency operating terminal. Both Criminal Investigative Division (CID) and Internal Security Division (IS) have detailed operating procedures, e.g., CID procedures require NCIC be queried when evaluating possible tax fraud. NCIC entries are limited to IRS fugitives, and permitted only at the CID National Office terminal in Washington, D.C. Fugitives are purged from NCIC when apprehended or when matter is dismissed by Federal courts. Part 11, ch. 1, sec. 18 of Confidential/ Field Manual. Authority provided by 39 USC 404(a)(7) and 18 USC 3061. NCIC access by written agreement with FBI. FPS program statements 1070.1 and 1231.1 and NCIC operating manual. USM Order 2423.1, ch. 3; e.g., arrest warrants issued to U.S. Marshal by a Federal court are to be screened to determine if the USMS retains the primary responsibility for their entry into NCIC. Warrant information will be forwarded via Justice Telecommunication System (JUST) within 48 hours to the USMS Communications Center for entry into NCIC according to the NCIC operating manual, a oct 16, 1979, memorandum ICI OT A from Dlvtslon of Law Enforcement Services, Bureau of Indian Affairs, U S Department of the Interior b Oct 2, 1979, letter tO OTA from Deputy commissioner, Interna[ Revenue Service, U S Department of the Treasury c Dec 24, 1 979, letter to OTA from the Chief Postal Inspector, U S postal Semlce d Sept 18, 1979, memorandum to OTA from the D!rector, Federal Prison Service, U.S Depafiment of Just Ice e Sept 18, 1979, memorandum to O TA from Director, U S Marshals Serwce, U S Department of Justice SOURCE Off Ice of Technology Assessment Federal and State Court Ruling s State and Federal courts have focused priof identification and arrest records by police marily on the collection, use, and maintenance at the local and State levels. As discussed ear-

PAGE 78

68 An Assessment of Alternatives for a National Computerized Criminal History System lier, some more recent cases (e.g., Menard v. Saxbe, Tarleton v. Saxbe, Menard v. Mitchell) have begun to focus on the recordkeeping policies and practices of the FBI. ing. Judicial rulings have lacked a consistent direction, as illustrated in table 12. This is 1O For further discussion of judicial rulings, see Donald A Mmchand, et al., Histnry and Background Assessment of the National Crime Information Cknter and Computerized Criminal History Program, Bureau of Governmental Research and Service, University of South Carolina, June 1979, sec. V, Regulating the Use of Criminal History Records in the United States: Overview of Activities, pp. 168-175. In general, however, the activity of the State and Federal courts has been infrequent and uncertain throughout the 100-year history of law enforcement and criminal justice recordkeepTable 12.lllustrative Federal/State Court Rulings on Criminal Records Individual Public safety Year court Cas e Ruling rights and welfare 1906 1941 1944 1945 1946 1966 1967 1967 1968-72 1970 1971 1974 1974 1976 1979 Supreme Court Louisiana Itzkovitch v. Whitaker Ruled for the defendant. Police could not post picture in x rogues gallery since it violated defendants personal rights because he had never been convicted. Kansas City Police restrained from disseminating photoX graphs and fingerprints of defendent within State and nationwide. In absence of controlling statute, police had discretion to destroy fingerprints, photographs, and measurements of those accused but not convicted, No justification for taking identification records in advance of conviction, except to identify person charged or to recapture a fugitive. Absent a statute, police had discretion to maintain and operate record systems for identification, even for those acquitted of misdemeanors. Absent State statute, police could retain arrest records whether accused was acquitted, discharged, or released. County officials should return fines and expunge police and court records connected with arrests and prosecutions intended to intimidate black citizens who wished to vote. Ordered fingerprints and photographs destroyed that were taken when defendant was arrested for refusing to submit to military induction. Decisions generally favored defendants involved in illegal and mass arrests or arrests not leading to conviction. Generally aimed at local or State police departments, not Ident. Arrest alone did not justify maintenance of fingerprints or record by State or Ident. Where probable cause for arrest exists, court would not order expungement by FBI, but would limit disclosure to nonlaw enforcement officials for employment purposes. FBI had no authority to retain record since arrest was changed to detention, FBI could retain neutral identification records. FBI had duty to prevent dissemination of inaccurate arrest and conviction records, and had to take reason. able precautions to prevent inaccuracy and incompleteness of records, Court held that the police had a right to publicize a record of an official act, such as an arrest, without exposing State or Federal officials to lawsuits for civil rights invasion. Court found a violation of sixth, eight, and 14th x amendment rights when arrest information without otherwise available disposition was used in setting bail. Supreme Court Missouri State v. Harris x court of Chancery New Jersey Court of Chancery New Jersey Supreme Court Indiana Fernicola v. Keenan McGovern v. Van Riper State v. Tyndall x x U.S. Court of Appeals Herschel v. Dyra U.S. v. McLeod US. Court of Appeals Alabama US. District Court Puerto Rico U.S. v Kalish (a) U.S. Court of Appeals District of Columbia U.S. District Court District of Columbia Menard v. Mitchell Menard v, Mitchell x U.S. Court of Appeals District of Columbia Menard v, Saxbe U.S. Court of Appeals District of Columbia Tar/ton v. Saxbe U.S. Supreme Court Paul v, Davis x U.S. District Court New York Tatum v. Rogers See, for example, Hughes v. Rizzo, 282 F Supp. 881 (1968), Morrow v District of Columbia, 417 F 2nd 728 (1989), Wheeler v Goodman, 306 F Supp 58 (1969) SOURCE Office of Technology Assessment.

PAGE 79

Ch. 6Legal/Regulatory Framework for NCIC, /dent, and State CCH Systems l 69 due in part to the limited involvement of the U.S. Supreme Court in this area. Most of the significant decisions have been made in State and lower Federal courts, and have varied widely in different States. Any trends in judicial decisionmaking have been more a product of the larger social and political movement toward expanding due process and other individual rights over the last 40 to 50 years, rather than the result of changes in judicial perspectives on criminal justice recordkeeping per se. In most criminal record cases, the balancing of individual rights of privacy and due process versus the maintenance of public safety and welfare has proven to be a difficult challenge to the courts. The tools that the courts have had at their disposal, such as injunctive relief and court orders to seal and expunge specific records, have been of limited effectiveness and used reluctantly. The courts have frequently sought legislative guidance. State Statutes and Regulation s The last 10 years have seen a dramatic increase in State statutes and regulations on criminal justice information systems. This is partly owing to the development of LEAA regulations (title 20, CFR, pt. 20) and State efforts to implement them. Early Efforts of Projec t SEARCH and LEA A In 1970, Project SEARCH (originally the System for Electronic Analysis and Retrieval of Criminal Histories) with LEAA funding developed a series of guidelines, model State statutes, and model administrative regulations for State and local CCH systems. 11 This effort was premised on the view that a nationally integrated CCH should be federated in nature; i.e., fundamentally dependent on State and local systems as opposed to one uniform national system. However, Project SEARCH recognized that such an approach would necessitate privacy and security standards at the State and local as well as Federal levels to uniformly protect individual rights and mitigate potential adverse social impacts. These early voluntary efforts produced some results. For example, four StatesAlaska, California, Iowa, and Massachusettsadopted the model State act and/or regulations in whole or in part. At the local and regional level, codes of ethics and self-imposed guidelines were adopted by some systems, such as the Santa Clara County (California) criminal justice information system and the Kansas City (Kansas) Alert II regional system. However, in 1971, concerned about the still limited acceptance of the Project SEARCH standards, LEAA required State plans to include provisions for privacy and security. In 1972, LEAA established the Comprehensive Data System (CDS) program that provided Federal dollars for CCH development, but made privacy and security plans a condition of funding. The CDS program was the primary means for LEAA to tie the development of local and State criminal justice information systems to a set of minimum standards for system development, privacy, and security. In July 1973, the LEAA-sponsored National Advisory ComSee Project SEARCH, Security and Privacy Considerations in (Mmimd History Information Systems, California Crime Technological Foundation, Sacramento, 1970; and A Model State Act for Criminal Offender Record Information, California Crime Technological Foundation, Sacramento, 1971. Also see Project SEARCH Committee on Security and Privacy, Model Adrninistrative Regulations for Criminal Offender Record Information, March 1972. See Donald Marchand, Criminal Justice Records and Civil Liberties: The State of Cdfornia Department of Justice, State of California, Sacramento, 1973, pp. 136-138, 358-366; and Melvin F, Bockelman, ALERT IIProgress Toward a Computerized Criminal Justice System, in Project SEARCH, Pro ceedings of the International Symposium on Criminal Justice Information and Statistics Systems, California Crime Technological Foundation, Sacramento, Calif., 1974, pp. 126, 131-2.

PAGE 80

70 l An Assessment of Alternatives for a National Computerized Criminal History System mission on Criminal Justice Standards and Goals adopted privacy and security standards that largely reflected Project SEARCH reports. 3 By March 1974, 33 States had indicated their desire to participate in the CDS program by submitting plans. Implementing LEAA Privac y and Security Regulation s In 1973, an amendment was added to the Omnibus Crime Control and Safe Streets Act of 1968 by Sen. Edward Kennedy requiring LEAA to promulgate regulations to provide safeguards for the privacy and security of criminal history record information. The Kennedy amendment followed a period of frustrating efforts by both the House and the Senate to pass legislation controlling the use of arrest records nationwide. During July 1973, Senator Kennedy tacked on his amendment to the primary piece of legislation supporting the LEAA program. While the measure was considered temporary by Congress in the light of anticipated efforts to pass more comprehensive legislation, it had considerable impact on LEAA and its relations with State and local criminal justice agencies. Section 524(b) of the Crime Control Act of 1973, as amended, provided that: All criminal history information collected, stored, or disseminated through support under this title shall contain, to the maximum extent feasible, disposition as well as arrest data where arrest data is included therein. The collection, storage and dissemination of such information shall take place under procedures reasonably designed to insure that all such information is kept current therein; the Admini stration shall assure that the security and privacy of all information shall only be -. g National Advisory Commission on Criminal Justice Standards and Goals, Report on the Criminal Justice System, U.S. Government Printing Office, Washington, D, C., 1973. 1 4 Richard W. Velde, LEAA Deputy Administrator for Policy Development, prepared statement in U.S. Congress, Senate Committee on the Judiciary, Subcommittee on Constitutional Rights, Criminal Justice Data Banks, Hearings, vol. I, 93d Cong., 2d sess., 1974, p. 301. used for law enforcement and criminal justice and other lawful purposes. In addition, an individual who believes that criminal history information concerning him contained in an automated system is inaccurate, incomplete, or maintained in violation of this title, shall, upon satisfactory verification of his identity, be entitled to review such information and to obtain a copy of it for the purpose of challenge or correction. 15 Following passage of the act with the Kennedy amendment, LEAA issued draft regulations in 1974 and held hearings in different parts of the country. On May 20, 1975, LEAA published its regulations, which required States accepting Federal funding to develop specific policies and procedures in five areas: 1) completeness and accuracy of records, 2) audit, 3) individual access and review, 4) limits on dissemination of records, and 5) security. 16 LEAA issued final regulations on March 19, 1976. The States experienced a number of problems in implementing the regulations including lack of resources, confusion in interpretation of the regulations, and lack of a State legislative mandate. More specifically, the following impediments to State implementation were identified in each of the five areas covered by the regulations: l l l Completeness and Accuracy: the lack of a clear and effective mandate, funds and/ or technical ability needed to introduce or improve an arrest and disposition reporting system, and sufficient time in which to do SO. Individual Access and Review: the lack of standardized, comprehensive policies, applicable to all impacted agencies in a State, which are supported by formalized procedures and the force of State law. Limitations on Dissemination: the lack of a statewide policy supported by formalOmnibus Crime Control and Safe Streets Act of 1968, Public Law 90-351,82 Stat. 200 (1968), adding sec. 524 (42 USC $ 3771). Carried forward by Justice Systems Improvement Act of 1979, Public Law 96-157, $ 818, 93 Stat. 1212 (1979) as 42 USC j 3789g (&lpp. 1980). Codified in 28 CFR 20, subpt. B. See Mitre Corp., Implementing the Federal Privacy and Security Regulations, McLean, Va., December 1977.

PAGE 81

Ch. 6Legal/Regulatory Framework for NCIC, Idenf, and State CCH Systems 71 ized mechanisms and procedures, that is promulgated, pursued and enforced by some responsible agency. Security: the lack of specific, statewide security standards and the resources required for the full implementation of these standards. l Audit: the lack of both a legislative mandate to conduct audits and the resources these audits will require. In 1978, LEAA issued two publications to assist the States in adopting information management policies for local and State criminal justice information systems. The first surveyed in detail the existing privacy and security statutes and administrative policies. 19 The second assessed the issues and difficulties that the States have confronted in 25 areas of information policy, and highlighted 4 State approaches to developing regulations. 20 In addition, LEAA and SEARCH Group, Inc. (a nonprofit corporation formed in 1975 with broader membership and interests than the original Project SEARCH) continued to provide policy, management, and technical assistance to State and local agencies. However, by 1980 almost all LEAA funding for State implementation had been phased out. LEAA itself was reorganized by the Justice Systems Improvement Act of 1979, with most prior functions terminated or transferred to other agencies by spring 1982. SEARCH Group, Inc. continues to receive research funds from DOJ Bureau of Justice Statistics. Ibid., Volume 1: Findings and Recommendations of an Eighteen State Assessment, p. ix. 191,F,AA, us, Department of Justice, Privacy and ~c~ritY of Criminal IIistory Information: A Compendium of State I&gislation, 1978. LEAA, U.S. Department of Justice, Privacy and Security of Criminal History Information: An Analysis of Privacy Issues, 1979. State Statutes and Regulations as of June 1981 The latest comprehensive survey of State statutes and regulations, conducted by SEARCH Group, Inc., and funded by the Bureau of Justice Statistics, documents substantial progress between 1974 and mid-1981. 2 State statutes and regulations are classified into 28 different categories described in table 13. 22 The methodology used to conduct this and similar prior surveys included library research and extensive contact with both the legislative information offices and record repositories of many States. Once the laws were collected, each States Attorney General was sent a copy of his States laws and attested to their completeness and accuracy. Responses served to correct any omissions or inaccuracies in the initial survey. 23 The survey results are summarized in table 14. They indicate that by 1981 over two-thirds of the States had statutes and/or regulations: 1) establishing a State regulatory authority (46 States in 1981 compared with 7 States in 1974); 2) placing some kind of restrictions on dissemination of criminal history information (51 States compared with 12 States); 3) establishing the rights of individuals to inspect their criminal history records (43 States compared with 12); 4) requiring agencies to ensure reasonably complete and accurate criminal history information, including timely disposition reporting (49 States compared with 14); 5) providing criminal sanctions for violation of privacy and security laws (39 States compared with 18); and 6) stipulating what criminal history records are to be open to the public (52 States in 1981 compared with 9 in 1974). SEARCH Group, Inc., Trends in State Skcurity and l+ivacy Legkdation, Sacramento, Calif., November 1981. The full results of the survey are available from the Bureau of Justice Statistics, U.S. Department of Justice. These are the same 27 categories used in the 1978 LEAA Compendium of Stat.e Legislation and a 1979 Supplement, with the addition of category 28, Establishment of a Central State Repository. LEAA, Ibid., 1979 Supplement, p. vii.

PAGE 82

72 An Assessment of Alternatives for a National Computerized Criminal History System Table 13.Categories of State Statutes and Regulations 1. State Regulatory Authority. A grant of power to a State agency to promulgate Statewide security and privacy regulations for criminal justice information systems. 2. Privacy and Security Council. A State board, committee, commission, or council whose primary statutory function is monitoring, evaluating, or supervising the confidentiality and security of criminal justice information. 3. Regulation of Dissemination. Restrictions on dissemination of criminal history information. 4. Right To Inspect. The right of an individual to examine his criminal history records. 5. Right To Challenge. The right to an administrative proceeding in which an individual may contest the accuracy or completeness of information pertaining to him. 6. Judicial Review of Challenged Information. The right of an individual to appeal an adverse agency decision concerning challenged information to a State court. 7. Purging. Nonconviction Information.The destruction or return to the individual of criminal justice information where no conviction has resulted from the event triggering the collection of the information. 8. Purging: Conviction /n formation. The destruction or return to an individual of criminal history information indicating a conviction. 9. Sealing: Nonconviction Information. The removal of criminal history information from active files where no conviction has resulted from the event triggering the collection of information. 10. Sealing: Conviction Information. The removal from active files of individual criminal history information indicating a conviction. 11. Removal of Disqualifications. The restoration of rights and privileges such as public employment to persons who have had criminal history records purged or sealed. 12. Right To State Nonexistence of a Record. The right to indicate in response to public or private inquiries the absence of criminal history in cases of arrest not leading to conviction or where an arrest or conviction record has been purged. 13. Research Access. The provision for and regulation of access to criminal justice information by outside researchers. 14. Accuracy and Completeness.A requirement that agencies institute procedures to ensure reasonably complete and accurate criminal history information, including the setting of deadlines for the reporting of prosecutorial and court dispositions. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. Dedication. The requirement that computer configurations be assigned exclusively to the criminal justice function. Civil Remedies. Statutory actions for damages or other relief resulting from violations of various privacy and security laws. Criminal Penalties.Criminal sanctions for a violation of various privacy and security laws. Public Records. Requirements that certain criminal history records maintained by the police or courts be open to the public. Separation of Files. Requirements that criminal history information be stored separate from investigative and intelligence information. Regulation of Intelligence Collection, Restrictions on the kind of intelligence information that may be collected and retained and/or prohibition on its storage in computerized systems. Regulation of Intelligence Dissemination. Restrictions on dissemination of intelligence information. Security. Requirements that criminal justice agencies institute procedures to protect their information systems from unauthorized disclosure, sabotage, and accidents. Transaction Logs. Records that must be maintained by criminal justice agencies indicating when and to whom criminal justice information is disseminated. Training of Employees. Security and privacy instruction that must be provided to employees handling criminal justice information. Listing of /formation Systems. A mandatory disclosure of the existence of all criminal justice information systems describing the information contained in such systems. Freedom of Information (Including Criminal Justice Information). Provisions for public access to government records that apply to criminal justice records. Freedom of information (Excluding Criminal Justice information). Provisions for public access to government records from which criminal justice records are specifically excluded. Central State Repository. Establishment of a bureau, agency, or other entity to collect and maintain criminal history records or criminal identification data for all criminal justice agencies in the State. SOURCE SEARCH Group, Inc LEAA, and Bureau of Just Ice Statistics

PAGE 83

Ch. 6.Lega/Regatoryry Framework for NCIC, Ident, and State CC/-/ Systems 73 Table 14.Survey Comparison of Changes in State Statutes/Regulations by Category a Item 1974 1977 1979 1981 Hem 1974 1977 1979 1981 1. State regulatory authority ... 7 38 42 46 2. Privacy and security council 2 10 13 21 3 Regulation of dissemination 24 40 44 51 4 Right to Inspect . 12 40 43 43 5. Right to challenge . 10 30 36 35 6. Judicial review of challenged information 10 20 22 18 7 Purging nonconviction information 20 23 28 35 8. Purg!ng conviction Information 7 13 19 24 9 Sealing nonconviction information 8 15 16 20 10 Sealing conviction information 7 20 21 22 11 Removal of dlsqualifications 6 22 22 27 12 Right to state nonexistence of a record 6 13 17 22 13 Researcher access. ... ., ., 6 12 14 21 14, Accuracy and completeness 14 41 45 49 15 Dedication ... ... ... . 2 3 3 2 16. Civil remedies ., ... 6 22 25 33 17. Criminal penalties . 18 35 39 39 18, Public records . 9 43 42 52 19 Separation of files ... ., 5 10 10 7 20. Regulation of intelligence collection. 3 10 10 13 21. Regulation of intelligence dissemination . . 7 24 25 19 22, Security ... ... . 12 26 31 32 23. Transaction logs . . ., 6 11 27 29 24. Training of employees. ., 4 18 23 16 25. Listing of Information systems . 1 8 8 8 26. Freedom of Information Including Criminal Justic e (b) (b) 18 27 27. Freedom of Information excluding Criminal Justice . . (b) (b) 22 28. Central State repository (b) (b) (; 52 aThe flgure~ Presented are ~umulatlve and may Include statutes or regula !Ions previously enacted but excluded from prior su~eys b Data unavailable for these years SOURCE SEARCH Group, Inc Bureau of Justice Statistics and LEAA, U.S Department of JustIce Initiatives to Enact Comprehensiv e Federal Legislatio n 2 4 As noted earlier, the Kennedy amendment to the Omnibus Crime Control and Safe Streets Act of 1973, and more recently the restrictions on NCIC hardware procurements and prohibitions on FBI message switching included in DOJ Appropriations Acts, have been interim actions aimed at dealing with spe cific problems until more comprehensive legislation could be enacted. During the decadelong debate, congressional initiatives and executive branch proposals for comprehensive legislation on criminal justice information systems have not produced such legislation. As early as 1970, Congress approved an amendment to the Omnibus Crime Control and Safe Streets Act, sponsored by Sen. Charles Mathias, which required LEAA to submit legislation to ensure the integrity and accuracy of criminal justice information systems funded in whole or in part by the Federal Government, and protecting the constitutional rights of all persons covered or affected by the act. In 1971, Sen. Roman Hruska intro. ZJFO~ ~ det~fled discussion, SEW Marchand, et al., ~isto~.} and Background Assessment, op. cit., pp. 192-202, and, more generally, pp. 72-167. duced S. 2546, The Criminal Justice Information Systems Security and Privacy Act of 1971 for DOJ in response to the Mathias amendment. This bill essentially would have codified the NCIC privacy and security policies and afforded substantial discretion to the Attorney General with respect to implementation. In 1972, Sen. Hruska introduced a similar bill, except that it provided for reversal of the Menard v. Mitchell decision. Both bills were referred to committee with no further action taken. In 1972 and 1973, Cong. Don Edwards introduced bills to establish privacy and security standards for the dissemination and use of criminal arrest records, and to regulate all State and local as well as Federal criminal justice information systems receiving Federal funds. Both bills were referred to committee and hearings were held, 25 but no further action was taken. _-_-___ ...- ~~see, for example, U.S. congress, House commitk~ on the Judiciary, Subcommittee No. 4, Securit.} and Pri}rac.} of Criminal Arrest Records, Hearings, 92d Cong., 2d sess.. N1 ar. 16, 22, 23, and Apr. 13 and 26, 1972.

PAGE 84

74 An Assessment of Alternatives for a National Computerized Criminal History System In February 1974, Sen. Hruska introduced S. 2964, The Criminal Justice Information Systems Act of 1974, on behalf of DOJ, and Sen. Sam Ervin, Jr. introduced S. 2963, The Criminal Justice Information Control and Protection of Privacy Act of 1974, on behalf of the Subcommittee on Constitutional Rights of the Senate Judiciary Committee. Both bills refleeted much of the work of Project SEARCH, the National Advisory Commission on Criminal Justice Standards and Goals, and the NCIC privacy and security policies. However, the Ervin bill took a more restrictive approach that would have limited all record disseminations to conviction information only and severely constrained noncriminal justice access. Also, the Ervin bill would have created a Federal Information Systems Board to be responsible for administration and enforcement, whereas the Hruska bill would have vested such authority in the Attorney General. Extensive hearings were held on both bills. 26 The result was a compromise bill introduced by Sen. Ervin in December 1974. No further action was taken that year. In 1975, Sen. John Tunney, then Chairman of the Senate Judiciary Subcommittee on Con stitutional Rights, and Cong. Don Edwards, Chairman of the House Judiciary Subcommittee on Civil and Constitutional Rights, reintroduced the original 1974 Ervin and Hruska bills and the Ervin compromise bill. Hearings were held in both the House and Senate. 27 Because 2% U.S. Congress, Senate Commit@ on the Judiciary, Subcommittee on Constitutional Rights, Criminal Justice 13ata Banks, Hearings 1974, vol. I, hearings, vol. II, app., 93d Cong., 2d sess., March 1974. see U.S. Congress, Senate Committee on the Judiciary Subcommittee on Constitutional Rights, Crimtial Justice Information and Prohwtion of fiivacy Act of 1975, 94th Cong., 1st sess., of continuing disagreements among DOJ, the International Association of Chiefs of Police, Project SEARCH, various State officials (e.g., the Attorney General of Massachusetts), and the American Civil Liberties Union, among others, no further action was taken on these bills or on a new compromise bill introduced by Sen. Tunney. 28 Since 1975, there have been no new congressional or executive branch initiatives for comprehensive legislation. The proposed FBI Charter legislation did make some limited reference to criminal justice information systems; and Senate Judiciary Committee hearings were held in late 1979 on sections 535(c), 536(d), and 536(e), the provisions of the Senate version (S. 1612) that related to the collection and dissemination of criminal history information. However, FBI Charter legislation was not enacted by the 96th Congress and is not under consideration by the 97th. Also, in 1980 and 1981, the Senate passed amendments to the DOJ Appropriations Authorization Act to mandate a new, comprehensive study of DOJ criminal justice information systems, and to reaffirm the congressional prohibition on message switching unless and until a message switching plan has been approved by the appropriate committees of Congress. 29 July 15 and 16, 1975; and U.S. Congress, House Committee on the Judiciary, Subcommittee on Civil and Constitutional Rights, Criminal Justice Information Llmtrol and Protection of Privacy Act of 1975, 94th Cong., 1st sess., July 14, 17, and Sept. 5, 1978. 2RFor further discussion of the legislative and policy history, see Marchand, et al., History and Background Assessmen& op. cit., and also Donald A. Marchand, The Pohtics of Privacy, Cbmputers, and Criminal Justice Records, Information Resources Press, Arlington, Va., 1980. g See S. 2377, sec. 113, 96th Cong., 2d sess.; and H.R. 4169, 97th Cong., 1st sess., and Senate Amendment No. 612 passed by Senate rollcall vote of 85-O on Nov. 12, 1981.

PAGE 85

Chapter 7 Use and Users of NCIC/CCH, Ident, and State CCH Systems

PAGE 86

Contents Chapter Summary. . . . . . Federal Government Users. . . . State and Local Users. . . . . International Users . . . . Private Users.. . . . . . Federal Government Users . . . NCIC/CCH File Use . . . . Ident: Criminal Justice Use..... . Ident: Noncriminal Justice Use..... . State CCH Systems Use. . . . State and Local Users . . . . NCIC/CCH Use . . . . . Ident: Criminal Justice Use..... . Ident: Noncriminal Justice Use..... . State CCH Systems Use. . . . International Users. . . . . . Private Users.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . TABLES Table No. 15. 16. Pattern of Use and Users: NCIC/CCH, Ident, and State CCH Systems . . . . . ........4 .. .. .., .. $ . . Noncriminal Justice Use of Ident by Federal Agencies, Fiscal Year 1981 . . . . . . . . . . . . Page 77 77 78 78 78 79 79 80 80 81 82 82 82 83 83 84 85 Page 77 81

PAGE 87

Chapter 7 Use and Users of NCIC/CCH Ident, and State CCH Systems Chapte r The two primary categories of criminal justice information system users are Federal Government agencies and State and local agencies. Foreign countries represent a third category, and private organizations a fourth. For each of these categories, OTA reviewed available data on the uses of the National Crime Information Center (NCIC) computerized criminal history (CCH), Identification Division (Ident), and State CCH systems. Patterns of use are summarized in table 15. Federal Government User s NCIC/CCH: Federal agencies collectively represent about 12 percent of total CCH trafTable 15.Patterns of Use and Users: Summar y fic. The relatively high Federal usage rate in part reflects the fact that all Federal offenders are included in the CCH file. Federal agencies use CCH information primarily for criminal justice purposes. Ident: All Federal agencies have access to Ident and account for one-third of Idents total trafficabout 3 percent by Federal law enforcement and criminal justice agencies and 30 percent by Federal noncriminal justice users. The large volume of Federal noncriminal justice use of Ident is attributed to several factors. First, Ident has a much more complete file than does CCH. Second, applicants for NCIC/CCH, Ident, and State CCH Systems 15A Criminal justice v. noncriminal justice use NC IC/CCH Ident State CCH .User agencies CJ a Non-CJ b Totals CJ Non-CJ Totals CJ Non-CJ Totals Federal State Local 12/0 22 /0 66 /0 (c) (c) (c) 12/0 3% 22% 66% 440/0 30 /0 23 0/0 33 /0 (d) 670/o 85 0/0 3 % 12 % 3 0/0 97% Totals 100/0 (c) 100 /0 47 /0 53 /0 1 00% 850/o 15/0 100 /0 15B. Law enforcement v. other criminal justice use NC IC/CCH Ident State CCH User agencies LE e Other CJ f Totals LE Other CJ Totals LE Other CJ Totals Federal 800/0 20 0/0 100 /0 80/0 920/o 100% (9) (9) State Local 480/o 100/0 520/o 1 00% 0% h 1 oo% 40% 600/0 100% 6670 34% ~CJ Cnmlnal Justice use (e g polIca, courts, corrections) Non.CJ = Noncrlmlnal justice use (e g employment and licensing, security checks) c = Negl!g!ble d = Precise data unavailable, but percentage estimated to be very small @LE Law enforcement use (e 9 Pollee sherlff~ Other CJ = Other crlmlnal justice user (e g prosecuting attorney, courts, probation parole) Not known County agency use Included with State agency use 100 0/0 SOURCES NCIC)CCH use percentages based on data from the July-September 1981 pilot test of the Interstate Iden!lf{catlon Index, data collected by FBI and Florlda Department of Law Enforcement, analysis and calculations by OTA Ident use percentages based on fiscal year 198f data collected by the FBI w!th the except!on of State/local law enforcement v other Crlmlnal Just Ice use data which are from U S Comptroller General F/ow Cr/rn/na/ Jusf/ce Agenc/es Use Crfrn~na/ F/(story /nformaf/err, U S General Accounting Office, Washington, D C Aug J974, p 34 State CCH use percentages based on data from 1979 OTA 50 State survey, and 1982 followup 77

PAGE 88

78 An Assessment of Alternatives for a National Computerized Criminal History System Federal positions, Federal contractors, military recruits, and national security personnel are routinely subjected to background investigations. Third, background investigations normally include fingerprint checks, which only Ident is equipped to process. The Department of Defense (DOD) and Office of Personnel Management (OPM) conduct over 95 percent of all background investigations. State CCH: Federal agency access to State CCH systems varies widely depending on the interpretation of State statutes. Federal requests account for about 23 percent of State CCH noncriminal justice traffic (about 3 percent of total State CCH traffic). State and Local User s NCIC/CCH: CCH file traffic from State and local agencies accounts for about 88 percent of total CCH traffic. In general, the probability of State and local users achieving a hit is not high because such a small fraction of current criminal records are contained in the CCH file. States that are fully participating in CCH use it most frequently. During July-September 1981, most State and local messages (about 67 percent) were from the eight States fully participating in CCH. Fifteen States had no CCH message traffic at all, and with few exceptions the remaining 27 had very little. ldent: State and local use accounts for about two-thirds of the total Ident workloadabout 44 percent by State and local law enforcement and criminal justice agencies, and about 23 percent by State and local noncriminal justice users. In order to reduce the current backlog and response time, Ident has suspended services to most State and local employment and licensing authorities until October 1, 1982, at which time user fees will be instituted. Ident does not maintain use statistics by originating agency. Interview results suggest, however, that criminal justice use is almost entirely in postarrest situations and that noncriminal justice use is primarily for employment and licensing purposes. Users are cautioned that Ident information is provided for official use only and should only be used for the purposes requested. The number and types of agencies that can access Ident vary greatly from State to State. State CCH: OTA found that about 85 percent of record requests were from criminal justice agencies and about 15 percent from noncriminal justice agencies. Of the 15 percent of CCH requests for noncriminal justice purposes, most were for State/local license applications, State/local employment checks, and Federal security checks. International User s NCIC: Canada is the only country permitted direct access to the NCIC hot files under a reciprocal assistance agreement between the Royal Canadian Mounted Police (RCMP) and NCIC. However, the RCMP cannot access the CCH file. Other foreign countries wishing to access NCIC must do so through the Drug Enforcement Administration (DEA), the official U.S. liaison with the International Police Organization (Interpol). The volume of international NCIC traffic apparently is quite small. Ident: Foreign use of Ident is also very limited. During fiscal year 1981, 2,556 fingerprint cards were submitted to Ident through the National Central Bureau of Interpol and by foreign police agencies. Almost all foreign users of Ident involve individuals (U.S. citizens abroad or foreign nationals giving a U.S. address) who have been arrested for narcotics or smuggling violations. Ident information is provided for criminal offenses only, not for political, religious, and social violations or complaints. Private User s With the exception of federally chartered or insured banking institutions and the securities industry, private organizations are not authorized access to criminal history information contained in Ident or NCIC. But in a majority of States, private organizations can lawfully obtain conviction information (and frequently arrest information as well) from State criminal history record files.

PAGE 89

Ch. 7Use and Users of NC/C/CCH, /dent, and State CCH Systems w 79 Federal Government User s NCIC/CCH File Us e Federal agencies, like all other agencies, must meet the following criteria used by the . FB1 1. 2. 3. m authorizing access: The agency must be a Government agency as required by title 28, U.S. Code, section 534. The agency must meet the definition of a criminal justice agency as contained in the Department of Justice regulations on Criminal Justice Information Systems (title 28, Code of Federal Regulations, pt. 20, subpt. A). Section 20.3 of these regulations defines a criminal justice agency as: courts, a government agency or any subunit thereof which performs the administration of criminal justice pursuant to a statute or executive order, and which allocates a substantial part of its annual budget to the administration of criminal justice. The administration of criminal justice means performance of any of the following activities: detection, apprehension, detention, correctional supervision, or rehabilitation of accused persons or criminal offenders. The administration of criminal justice shall include criminal identification activities and the collection, storage and dissemination of criminal history record information. An agency not meeting the qualifications set out in (2) above must meet the definition of an agency under management control of a criminal justice agency as contained in the CCH program background, concept, and policy as approved by the NCIC Advisory Policy Board. Management control includes the authority to set and enforce priorities; standards for the selection, supervision, and termination of personnel; and policy governing the operation of computers used to process criminal history record information. Management control includes, but is not limited to, the supervision of equipment, systems design, programing, and operating procedures. A noncriminal justice user agency must have a written agreement with the criminal justice agency that has management control. Federal agencies must execute a CCH agreement with the FBI in order to access CCH. Federal users account for about 12 percent of the CCH files message traffic, based on data from the July-September 1981 pilot test of the Interstate Identification Index (111). Federal agencies directly accessing CCH during the test period included the U.S. Secret Service, U.S. Department of Justice, Bureau of Customs (which provides an interconnection with Treasury Enforcement Communication System (TECS)), U.S. Postal Service, and FBI. Since all active Federal offenders are ultimately recorded in CCH, the file contains a disproportionate number of Federal as opposed to State offenders. For example, on August 31, 1981, Federal offenders accounted for 24.4 percent of the total records in the CCH file. The relatively high usage rate for Federal agencies is thus not surprising. Judging from CCH purpose codes entered by users for each message, essentially all Federal user requests for CCH information are for criminal justice purposes. Secondary dissemination for other uses beyond the terminal agency has not been measured and identified by OTA, except through anecdotal references. CCH purpose codes provide little insight into the actual recipients and end users of CCH information. All law enforcement and criminal justice use is grouped under the C Code Federal Bureau of Investigation, Interstate Identification Index: Background and Findings for Jul+v-September 1981 Phase I Pilot Project, Dec. 4, 1981, pp. 146, 153, 158, 159. The percent of Federal use was calculated by adding the 407 matching inquiries (hits, where an inquiry matched an Index entry) for Federal agencies without NLETS access to the approximate 982 matching inquiries (8.6 percent of 11,415) for Federal agencies with NLETS access, and dividing total Federal agency matching inquiries (1389) by total matching inquiries (11,415 + 410 = 11,825). 2NCIC staff paper prepared for Nov. 3-4, 1981, meeting of the NCIC Advisory Policy Board Subcommittee on the Interstate Identification Index, Topic x3, p. 7.

PAGE 90

80 An Assessment of Alternatives for a National Computerized Criminal History System (Criminal Justice); thus, prosecutors, courts, probation, parole, and correctional institutions are included, along with police and law enforcement, within one general term. However, agencies requesting CCH information are identified by an Originating Agency Identifier. Statistics on type of requesting agency were maintained during the III pilot test. The results suggest that Federal law enforcement agencies accounted for about 80 percent of total Federal traffic on NCIC/CCH. 3 The CCH file also serves some Federal agencies for noncriminal justice purposes (i.e., employment and security screening), but the volume of transactions for such purposes is very small. The bulk of Federal agency inquiries for noncriminal justice uses is directed to Ident. Ident: Criminal Justice Us e While direct NCIC/CCH access is limited to a small number of Federal agencies, Ident may be used by all Federal agencies for either criminal justice or employment screening purposes. During fiscal year 1981, approximately 172,000 fingerprint cards were submitted to Ident by Federal law enforcement and criminal justice agencies, compared with 3.0 million cards from State and local criminal justice agencies and a total of about 6.8 million cards received by Ident from all sources. Rap sheets are provided by Ident to Federal agencies for the official use of the receiving agency only. Section 534 of title 28 (U. S. C.) provides that the exchange of identification records is subject to cancellation if dissemination is made outside of the receiving department or related agencies. Each identification record contains a caveat stating that it is furnished for official use only and that it should only be used for the purpose requested. FBI, 111 Background and Findings, op. cit. The percent of Federal law enforcement agency use was calculated by adding the 401 hits for Federal law enforcement agencies without NLETS access to the approximate 696 inquiries (6.1 percent of 11,415) for Federal law enforcement agencies with NLETS access, and dividing by total Federal agency hits as above. A General Accounting Office (GAO) report on criminal history information provides one indicator of actual rap sheet use. According to the report, almost 95 percent of all fingerprint card submissions are intended to be used for postarrest purposes. 4 This is entirely logical, since fingerprints are seldom obtained from individuals who are not already in custody. Moreover, the Ident turnaround time of several weeks precludes most prearrest and arrest uses of rap sheets, except in atypical situations (ongoing investigations, etc.). Ident: Noncriminal Justice Use In 1981, Ident processed a monthly average of 175,000 fingerprint requests for Federal employment, contractors, and security clearances. Noncriminal justice Federal users rely on Ident for several reasons: 1) direct access NCIC terminals are located only in law enforcement or criminal justice agencies, so noncriminal justice users cannot easily access NC IC/CCH files; 2) NCIC policy requires that any response to an employment or licensing inquiry be furnished only through a criminal justice agency; 3) employment background investigations are normally conducted through fingerprint checks, which only I dent is equipped to process; and 4) Ident has a much more complete file than does CCH. During fiscal year 1981, Ident received about 2.1 million fingerprint card inquiries from Federal noncriminal justice users. This accounts for roughly 30 percent of all of Idents traffic. The large volume of Federal noncriminal justice use may be attributed to several factors. All applicants for Federal positions, as well as many persons who are employed under Federal contracts, are routinely subjected to background investigations. Likewise, DOD uses Ident services when conducting security investigations related to personnel occupying sensitive positions, and for background checks of all military recruits. The .. U.S. Comptroller General, How Criminal Justice Agencies Use Criminal History Information, U.S. General Accounting Office, Washington, D. C., August 1974, p. 34.

PAGE 91

Ch. 7Use and Users of NCIC/CCH, /dent, and State CCH Systems 81 Immigration and Naturalization Service also accounts for a significant volume of traffic to Ident. Table 16 summarizes the volume of fingerprint card submissions to Ident by all Federal noncriminal justice users in fiscal 1981. The largest type of noncriminal justice use among Federal agencies involves background and security investigations of Federal applicants, employees, and contractors. Although DOD and OPM account for over 95 percent of all background investigations, a few other agencies operate internal suitability and security programs. The FBI, Treasury Department, Department of State (DOS), and Central Intelligence Agency (CIA) are the most notable. The FBI is responsible for investigating the background of its own employees, as well as White House personnel, Presidential appointees, employees of the U.S. Attorneys Office, and certain applicants of the Nuclear Regulatory Commission. DOS conducts background investigations of Foreign Service and Office of the Secretary of State employees and applicants. Additionally, DOS performs overseas investigations for OPM and DOD, and Table 16. Noncriminal Justice Use of Ident by Federal Agencies, Fiscal Year 1981 Flngerprint cards Purpose of use submitte d Federal employment Army ., ., ., ., ., 210,145 Air Force ... ., 105,791 Navy, .,, 111,924 Marine Corps ., ............ 52,066 Coast Guard...................... ... 19,557 Off Ice of Personnel Management 334,941 Miscellaneous Federal agencies 38,605 Subtotal . ... ... 873,029 (41 .4%) Federally related employment (contractors, security clearances) Department of Energy 5,605 Defense Investigative Service ., 331,641 Department of Transportatio n 1,105 Other Federal agencies ., 73,450 Subtotal . ... ., ., 411,801 (19.60/, ) Nonemployment users Veterans Admlnlstratlon (establishing entitlements to benefits) ., 1,269 Immlgratlon and Naturalization Service (persons applylng for naturalizatlon, cltlzens applylng for adoption of foreign-born children, etc ) 820,742 Subtotal ., 822,011 (39%) Total Federal noncrimlnal justice fingerprint cards submltted 2,106,84 1 SOURCE Federal Bureau of Investlgatlon conducts limited checks of visa and passport applicants. The Treasury Department conducts suitability and security investigations for many of its own employees, especially those working for the Internal Revenue Service, U.S. Secret Service, and Bureau of Alcohol, Tobacco and Firearms. CIA conducts special background investigations of all its own employees. State CCH Systems Us e Executive Order No. 10450 specifically states that the processing of all applicants for Federal jobs must include written inquiries to local law enforcement agencies. Pursuant to this requirement, all employment-related investigations conducted by Federal agencies include written and/or personal contacts with State and/or local law enforcement authorities. In each of these investigations, Federal investigators must attempt to access criminal history record information contained in State and local files. Moreover, military recruiters routinely verify or investigate the background information provided by applicants for military service. Despite the obvious importance of State and local criminal history information to Federal suitability investigations, Federal agencies fre quently encounter difficulties in accessing nonFederal files. Access to State and local criminal history records is governed by many varying (and conflicting) statutes, procedures, and interpretations that make it virtually impossible for Federal agencies to obtain certain records. Federal officials cite the Privacy Act of 1974 and Law Enforcement Assistance Administration (LEAA) regulations as two causes of the inconsistent practices and procedures that govern Federal agency access to State and local criminal history files. In addition, access policies in some States do not specifically apply to local criminal history records. Furthermore, in States that have enacted statutes governing access, the interpretation is frequently left to local jurisdictions. One consequence has been a tendency for access to be harder for OPM and easier for DOD. Many jurisdictions deny OPM access on the basis

PAGE 92

82 l An Assessment of Alternatives for a Nationa/ Computerized Crimina/ History System that State laws (or LEAA regulations) only permit access for c riminal justice agencies, and OPM is not a criminal justice agency. On the other hand, many jurisdictions have used their discretion to define certain DOD units as criminal justice agencies, thus permitting access.5 The 1979 OTA survey of 50 States indicated that Federal security checks (including miliSee SEARCH Group, Inc., Federal Access to State and Imcai Criminal Justice information, Sacramento, Calif., March 1979. State an d NCIC/CCH Us e The possibility of State and local users achieving a hit on NCIC/CCH files is generally low because such a small fraction of current criminal records are contained in the CCH file. For fiscal year 1981, the eight States fully participating in (submitting records to) NCIC/ CCH accounted for only about 24 percent of all criminal fingerprint cards submitted to Ident in that year. 7 Based on data from the III pilot test, State agency users account for about 22 percent of CCH message traffic and local agencies about 66 percent. Most of these State and local messages (about 67 percent) were from the eight States fully participating in CCH at that time. Fifteen States had no CCH messages at all during the test period, and most of the remaining 27 States (with the notable exceptions of Oregon, Illinois, and California) had very few. 8 Most requests (about 94 percent during the test period 9 ) were for summary online criminal history records. Summary records were provided within 30 minutes 77 percent of the time and within 15 minutes 64 percent of the time. l0 Ident received 2,914,911 criminal fingerprint cards from States in fiscal year 1981. Of that total, 708,149 cards (or 24.3 percent) were received from the eight States fully participating in NCICICCH: Florida (272,400 fingerprint cards); 1owa (18,730); Michigan (55,727); Nebraska (8,259); North Carolina (48,800); South Carolina (83,560); Texas (156,804); and Virginia (63,869). Data from FBI, 111 Background and Findings, op. cit. I bid., Ibid., p. 145. Ibid., p. 162. tary) accounted for about 23 percent of noncriminal justice requests in States with CCH systems, and about 35 percent in those with manual systems. 6 Federal noncriminal justice use accounts for about 3 percent of total State CCH use.* 6 OTA 50-State Survey conducted in 1979-80. See appendix B for list of State officials responding. *C~culated by multiplying the Federal percentage of State CCH noncriminal justice use (23 percent) by the noncriminal justice percentage of total State CCH use (15 percent). Local User s Statistics maintained for the III pilot test indicated that State agency requests were split about evenly between law enforcement and other c riminal justice agencies (prosecuting attorneys, courts, corrections), but that local inquiries were almost entirely from law enforcement agencies.11 Like their Federal counterparts, and essentially for the same reasons, State and local noncriminal justice users rely almost entirely on Ident to provide centralized criminal history record information for licensing and employment purposes. Ident: Criminal Justice Us e For States lacking a well-developed centralized criminal history file, Ident provides a central repository and clearinghouse for criminal history record information. Even for States with a more sophisticated central repository, Ident provides a national repository for the criminal history records of multi-State and Federal offenders. During fiscal year 1981, about 3 million criminal fingerprint cards submitted by State and local agencies were processed. Ident does not maintain use statistics by originating agency because of the vast number (over 20,000) of organizations authorized to submit fingerprints and other criminal history infor] i I bid, p. 158, However, county level agencies were counted as State rather than local users.

PAGE 93

Ch. 7Use and Users of NCIC/CCH, /dent, and State CCH Systems l 83 mation. Results of interviews conducted for this study tended to confirm the earlier GAO findings that about 94 percent of fingerprint record use by State and local criminal justice agencies is for postarrest purposes, and about 40 percent of criminal justice use is by law enforcement agencies with the other 60 percent accounted for by prosecutorial, judicial, probation-parole, and correctional agencies. Ident: Noncriminal Justice Use Noncriminal justice use of Ident by State and local agencies and other authorized users* accounted for about 23 percent of its total workload during fiscal year 1981, when about 1.6 million fingerprint cards were received from these sources. On October 1, 1981, due to a large backlog of unprocessed cards (400,000 as of September 1981) and increasing processing time (up to 27 workdays in September 1981), Ident suspended services to federally chartered or insured banking institutions and State and local employment and licensing authorities. Ident plans to restore these services on October 1, 1982, charging a user fee of$12 per fingerprint card processed to cover costs. 13 Ident does not maintain separate figures for licensing and employment fingerprint card submissions because both types of record checks are performed pursuant to the same authority (Public Law 92-544). In addition, since 1957, the division has not retained the fingerprint cards submitted by State and local employment and licensing users. Unlike NCIC/CCH, Ident provides criminal history information directly to noncriminal justice users; previously approved mailing addresses are used to furnish responses. Such users are cautioned that the information is provided for official use only and should only be used for purposes requested. RedisU. S. Comptroller General, Criminal History Znforrnation, op. cit., pp. 12-14. *other authorized users include primarily federally chm~red or insured banking institutions and the securities industry. See Sept. 1 and Oct. 1, 1981, letters to all fingerprint contributors from Nick Stames, Assistant Director, FBI Identification Division. The suspension does not apply to employment in criminal justice agencies or to the securities industry. semination of criminal records provided by Ident is not permitted unless the requesting agency is already authorized to receive such records. However, once information becomes part of a State or local file it can be difficult to identify the information as having originally been provided by Ident, and limitations on dissemination are difficult to enforce. The number and types of State or local agencies permitted access to criminal history data vary greatly from State to State. 14 Som e States permit a wide variety of licensing authorities to use criminal history records (e.g., real estate commissions, alcoholic beverage boards, parimutuel racing commissions, licensing boards for such occupational groups as barbers, cosmetologists, psychologists, insurance agents, polygraph examiners, and adoption authorities). Other States forbid access to all agencies except those authorized by Federal statute (banking and securities industries). Also, State laws control the classification of peace officers, not all of whom are directly involved in conventional law enforcement duties. State CCH Systems Us e The extent to which in-State criminal histories satisfy the needs of State criminal justice agencies varies from one State to another. For example, some States do not require centralized reporting of fingerprint cards on criminal offenders. In addition, local police agencies in the State are not always consistent in notifying the State repository of arrests and dispositions. Consequently, some State criminal history files contain only a fraction of all known criminal offenders in the State. Given these constraints, law enforcement and criminal justice users in these States depend to a greater extent on the FBI even for in-State criminal history information. 15 See in general Steven W. Hays, et al., An Assessment of the Uses of Information in the National Crime Information Center and Computerized Criminal History Program, Bureau of Governmental Research and Service, University of South Carolina, October 1979, sees. 11, III, and IV, and especially app. B, State Repository Site Visits, pp. 172-322, which included California, Florida, Mississippi, South Carolina, and Texas. ibid.

PAGE 94

84 l An Assessment of Alternatives for a National Computerized Criminal History System ..-. .. In contrast, Californias criminal history file is estimated to be approximately 78-99 percent complete; i.e., about 78 percent of all reportable arrests and about 99 percent of all adult arrests in the State are included in the States repository. Thus, California users of criminal history records rely on the FBI primarily for out-of-State information. 16 The 1982 OTA 50-State survey followup found that about 85 percent of all requests to State CCH repositories were from criminal justice agencies, and about 15 percent were from noncriminal justice agencies. Based on 1979 Ibid., California site visit, pp. 272-282; personal communication with Nell Hutchinson, California Department of Justice, Aug. 30, 1982. data, about 66 percent of requests from criminal justice agencies were from law enforcement agencies and about 34 percent from other criminal justice agencies (primarily courts, prosecutors, probation-parole, and corrections). Of the requests from noncriminal justice agencies, about 49 percent were for license applications, 24 percent for employment checks, and 4 percent for State and local security checks. As noted earlier, about 23 percent of the requests from State CCH systems were for Federal security checks, whereas the comparable figure for manual systems was about 35 percent. 17 170TA 50-State Survey conducted in 1979-80. International User s Under certain conditions, information from NCIC hot files and from Ident is made available to government authorities in foreign countries. Canada is the only foreign country permitted to access the NCIC data base directly. Under a reciprocal assistance agreement, the RCMP have a terminal in their central headquarters and NCIC has access to the Canadian Police Information Center in Ottawa. The RCMP cannot access the CCH file. Other foreign countries wishing to access NCIC must do so through DEA, which is the official U.S. liaison with Interpol. DEA is responsible for determining whether or not Interpol requests are consistent with existing treaties and Federal legislation. Since Interpol inquiries are lumped together with Department of Justice message traffic, it is not possible to identify the exact volume of international NCIC traffic. However, FBI officials note that the volume of traffic is very low. Interpol receives significant criminal information assistance through TECS. However, Interpols interface with TECS does not allow it to access NCIC indirectly nor to access State and local criminal information files through the National Law Enforcement Telecommunications System (NLETS). The use of Ident data base by foreign countries is also very limited. During fiscal year 1981, 2,556 fingerprint cards were submitted to Ident under the International Exchange Program. The largest number of fingerprint card submissions (952) was received from foreign contributors through the National Central Bureau of Interpol. The largest number of direct submissions from foreign police agencies came from Canada (1,160) and Great Britain (355). In 1981, 47 different countries submitted fingerprints to Ident and an additional 30 countries were entitled to do so. Ident notes that in almost every case the fingerprint cards that are submitted involve U.S. citizens arrested in a foreign country, or foreign nationals arrested in a foreign country who, in the judgment of the contributing agency, would be of interest to U.S. authorities, usually because they give an address in the United States at the time of arrest. Almost all of these cases involve individuals who have been arrested for narcotics or smuggling violations. Records provided to international users are manually reviewed before distribution, and carry the caveat that the record is provided for official use only and should only be used for purpose requested. Information is

PAGE 95

Ch. 7Use and Users of NCIC/CCH, Ident, and State CCH Systems l 85 provided only if there is a legitimate criminal try is not a member of the International Exoffense involved. Political, religious, or other change Program, DOS is asked to make a final social violations or complaints are not honored determination as to whether or not criminal by Ident. In cases where the requesting counhistory information should be released. Private User s With the exception of federally chartered or insured banking institutions and the securities industry, private organizations are not authorized access to criminal history information contained in I dent or NCIC. But in a majority of States, private organizations can lawfully obtain conviction information (and frequently arrest information as well) from State criminal history record files. For example, as of mid-1981, 10 States plus the Virgin Islands provided statutory authority for private employers to obtain both conviction and nonconviction arrest data. l8 In addition, many State statutes regulate only the central State repository or records disseminated by the repository. A recent SEARCH study concluded that in most States, even some of those with comprehensive criminal record statutes, local police agencies are still free-absent a local ordinance to release to private employers whatever arrest or conviction data they choose to. l9 In Florida, the States Public Records Statute permits private access to criminal history files. As of June 1979, the following private organizations were among those listed as secondary users of the Florida Crime Information Center: Commercial Carrier Corp.; General Telephone Co.; Jacks Cookie Co.; Ryder Truck Lines, Inc.; United Parcel Service; Winn Dixie Stores, Inc.; Rinker Materials Corp.; and *See ch. 6. SEARCH Group, Inc., Privacy and the Private Ernplo}er, September 1981 draft, p. 33. g Ibid., pp. 34-35. Hughes Refrigerated Express, Inc. 20 This list did not include ad hoc requests from private individuals and organizations. According to Florida officials, the list of private sector secondary users has grown substantially in recent years. During fiscal year 1981, about 37,000 private sector record checks were processed. This represented about 25 percent of all applicant record checks for that year. 21 In South Carolina, the States Freedom of Information Act permits private access to conviction information. As of mid-1979, officials of the South Carolina Law Enforcement Division (SLED) estimated that the primary noncriminal justice users of conviction information were large companies for the purpose of employment screening. There was no easy way for SLED to determine specifically who had access to such information once it was released to local agencies. In Florida, NCIC/CCH out-of-State information is exempted from the Public Records Statute and disseminated only in accordance with laws of the State originating the information. In South Carolina, NCIC/CCH out-of-State information is not maintained in the State CCH file, other than to note that such information is available from Ident. OHays, Assessment of Uses, op. cit., Florida Site Visit, p. 290-291; site visit conducted June 26, 1979. Ibid., p. 293. Fiscal year 1981 data from FE31, III Background and Findings, op. cit., p. 77. Hays, Assessment of Uses, op. cit., South Carolina site ~risit, p. 310; site visit conducted May 28 and June 4 and 12, 1979.

PAGE 96

Chapter 8 Record Quality in Federal and State Criminal History Information Systems

PAGE 97

Contents Page Chapter Summary . . . . . . . . . . . 89 Methodology of Record Quality Research . . . . . . 89 Methodology Used for Federal Files + . . . . . . . 90 Accuracy of Federal Record Quality Estimates . . . . . 90 Methodology Used for State CCH Files . . . . . . 90 Findings of Record Quality Research . . . . . . . 91 FBI Criminal History Files . . . . . . . . . 91 State Criminal History Files . . . . . . . . .. ....93 Significance of Findings. . . . . . . . . . 94 TABLES Table No. Page 17. Record Quality of FBI Identification Division Criminal History File Disseminations, Based on 1979 Sample . . . . . 92 18. Record Quality of FBI NCIC/CCH File Disseminations, Based on 1979 Sample. . . . . . . . . . . . 93

PAGE 98

Chapter 8 Record Quality in Federal and State Criminal History Information System s Chapter Summar y As discussed in chapter 6, Federal and many State laws and regulations emphasize the importance of complete, accurate, and current criminal history information. The results of record quality research conducted by OTA and others indicate that while the quality of criminal history records has improved since 1970, significant problems remain. The results for the Federal Bureau of Investigations (FBI) Identification Division (Ident) file, the National Crime Information Center/Computerized Criminal History (NCIC/CCH) file, and State files are summarized below.* Ident: Based on a 1979 sample of arrest events, OTA record quality research found that about 30 percent of the Ident records that could be verified lacked a court disposition that had occurred and was confirmed by the district attorney in the local area responsible for prosecution. A 1980 study conducted by the Jet Propulsion Laboratory found that Ident receives dispositions for about 45 percent of the arrests reported. OTA found that about one fifth of the I dent arrest events sampled were inaccurate when compared with charging, disposition, and/or sentencing information in local records. *OTA a]so conducted record quality research on the NCIC Wanted Persons File. Although outside the primary focus of this report, the results are summarized in app. A. NCIC/CCH: OTA record quality research based on a 1979 sample of arrest events found that about 27 percent of the CCH records that could be verified lacked a court disposition that had occurred. About one-fifth of the arrest events were inaccurate with respect to charging, disposition, and/or sentencing information in local records. While it is possible that NCIC/CCH (and Ident) record quality has improved since 1979, OTA is not aware of any comparable research conducted by the FBI or others to document such improvements. States: Based on 1979 and 1982 50-State surveys and a 1979 sample of records selected from one major urban jurisdiction in each of three States, OTA record quality research found that the most significant problem involved the lack of disposition information. Survey results indicated a 65 percent average disposition reporting level for the 41 States responding in 1979. For the three urban jurisdictions sampled, disposition reporting was 58, 60, and 85 percent. In general, a comparison between 1970 and 1979 survey data shows some improvement in disposition reporting for all States and significant improvement for States with computerized (as opposed to manual) systems. Several States contacted by OTA have achieved further improvement in disposition reporting since 1979, but the overall average increased only marginally to 66 percent in 1982. Methodology of Record Quality Researc h The research reported here is the first sysboth Federal and State criminal history infortematic effort to measure record quality in mation systems. The State CCH systems were 89

PAGE 99

90 l An Assessment of Alternatives for a National Computerized Criminal History System included in this study because any future national CCH system will depend on criminal history information generated initially, and in some designs maintained, by the States. Therefore, it is important to obtain an estimate of the levels of record quality in State systems. Methodology Used fo r Federal File s The research reported here on record quality of the Ident criminal history file was based on a stratified proportional sample of criminal history records selected from the outgoing mailroom of the FBI during late July and early August 1979. A random sampling procedure could not be carried out on this file because no log or list of recent disseminations existed. The selection of records was weighted by the only known population parameter-the proportion of records requested by various States. Within States, records were selected in rough proportion to the number of requests made by local agencies during the week of selection. One recent arrest event was chosen from each of 400 criminal history records selected for intensive examination. The criminal history information for each arrest event selected was then sent to the district attorney in the local area responsible for prosecution of the case for full and complete verification of arrest, court disposition, sentencing, and correctional information. Information returned by local district attorneys was then compared with information recorded on the Ident criminal history record, and record accuracy, completeness, and ambiguity were evaluated. The statistical estimates of record quality in Idents manual criminal history files are technically generalizable to the population of Ident disseminations that took place in 1979. The results are not technically generalizable to the entire Ident criminal file. Thus, this research measured the quality of information being disseminated by Ident, not the quality of records that are stored but not disseminated. The research on record quality of the NCIC/ CCH file was based on a systematic sample of the NCIC/CCH transaction log, with a random start for the period January 1 to June 1, 1979. Four hundred recently disseminated criminal history records were selected as of August 12, 1979, and from each a recent arrest event was chosen for verification. The process used for verification of NCIC/CCH records was the same as for Ident. The results of this research are technically generalizable to the population of CCH disseminations that took place during 1979. They are not, however, technically generalizable to the entire NCIC/ CCH file. Accuracy of Federal Recor d Quality Estimate s The ability to estimate population parameters using randomly drawn samples is a function of sample size as well as the underlying distribution of the variable being estimated. In this research, the sample sizes were restricted due to limited resources. Moreover, the response rate of local authorities who verified information varied. For the NCIC/CCH and the Ident criminal history samples, the statistical estimates of the record quality features are generally accurate to within 6 percent (plus or minus 6 percent). That is, there is 95 percent confidence that the true population parameters of record quality lie within plus or minus 6 percent of the estimates given in the tables. Methodology Used fo r State CCH File s Record quality of State criminal history files was estimated through use of surveys of all States and through a sampling of State CCH use in three major urban jurisdictions. For the former, a written questionnaire was sent by OTA in 1979 to the Governors of all 50 States, the District of Columbia, and Puerto Rico. Forty-eight States and Puerto Rico responded. The questionnaires were filled out by the State officials listed in Appendix B. A followup telephone survey of all 50 States was conducted in August 1982, with update information provided by the officials listed in Appendix D. For

PAGE 100

. Ch. 8Record Quality in Federal and State Criminal History Information Systems l 9 1 urban areas, since resources limited the size and location of sampling, one major urban jurisdiction in each of the three States was selected. All three States maintain CCH systems that are among the more advanced in operation. Within each jurisdiction, information contained in approximately 500 CCH records used in recent prosecutions was compared with information in local court and district attorney manual records. It was found that district attorneys did not have a list of recently prosecuted cases in which a State CCH record was used, and could provide only rough guesses about the annual number of such cases. The research team was therefore required to work backwards in the manual files of the district attorney until 500 cases were found in FB I which a State CCH record had been used. This meant looking at cases prosecuted between 1975 and 1979 in one jurisdiction, while in two others the cases spanned a l-year period. The State record quality sampling studies are technically generalizable to the population of criminal cases prosecuted using a State criminal history record in the jurisdiction examined for the time period specified; they are not actually samples from a larger population, at least not in any systematic sense of the word sample. In addition, they are not technically generalizable to the entire population of recently prosecuted criminal cases in the States as a whole. There are differences in the completeness and accuracy among jurisdictions within a State largely owing to variations in local court reporting procedures. Findings of Record Quality Researc h Criminal History File s The results of the OTA record quality studies of the I dent and NCIC/CCH files are summarized in tables 17 and 18. For both files, the major record quality problems were: 1) no disposition information, and 2) inaccurate disposition, charging, or sentencing information, when compared with information in local records. For each arrest event, the evaluation was based on a direct comparison between the information in the Federal record and the information in local records. No disposition meant that a court disposition was shown in the local record, but not in the Federal record. Inaccurate meant that the disposition, charges, or sentence shown in the Federal record did not agree with the disposition, charges, or sentence shown in the local record. Thus, for 49 (or 29.6 percent) of the 168 verifiable Ident arrest events and 45 (or 27.2 percent) of the 165 verifiable NCIC/CCH arrest events, no disposition was reported even though the disposition had occurred at least 120 days earlier. For 34 (or 20.2 percent) of the 168 Ident arrest events and 32 (or 19.4 percent) of the 165 NCIC/CCH arrest events, information on disposition, charges, or sentence was inaccurate. Tables 17 and 18 count only one record quality problem per record, although many records exhibited more than one. For example, counting multiple problems, 20 (or 11.9 percent) of the 168 Ident arrest events and 11 (or 6.7 percent) of the 165 NCIC/CCH arrest events showed more dispositions than charges or more charges than dispositions, when compared with local records. Other studies on record quality have tended to confirm the OTA findings with respect to disposition reporting. For example, an FBI analysis found that, as of August 13, 1979, 39.4 percent of arrests in the NCIC/CCH file were without dispositions. A 1980 study conducted by the Jet Propulsion Laboratory found that Ident receives dispositions for about 45 percent of the arrests reported. 1 The differences are partially explained by the OTA Jet Propulsion Laboratory, FBI Fingerprint Identification Automation Study: AIDS 111 Evaluation Report, Volume Vl: Environmental Analysis, California Institute of Technology, Pasadena, Nov. 15, 1980, p. A-3, prepared for the U.S. Department of Justice, Federal Bureau of Investigation.

PAGE 101

92 An Assessment of Alternatives for a National computerized criminal History Syste m Table 17.Record Quality of FBI Identification Division Criminal History File Disseminations, Based on 1979 Sample Arrests in sample. . . . . . . . . . . . 400 Local agency responses . . . . . . . . . . 231 Arrests not verifiable because . . . . . . . . . 63 Pending or sealed . . . . . . . . . . . 19 No record locatable . . . . . . . . . . 37 No prosecution of arrest . . . . . . . . . 6 Fugitive . . . . . . . . . . . . . 1 Total arrest cases verified . . . . . . . . 168 Results: Actual disposition not recorded on Ident record . . . . . 68 Disposition occurred more than 120 days prior to 7124/79a .......49 Disposition occurred less than 120 days prior to 7/24/79 b . . 11 Disposition occurred after 7/24/79 . . . . . . . 2 Disposition data unknown . . . . . . . . . 6 Ident record otherwise incomplete when compared to local record. 12 Shows sentence but no conviction information.. . . . . 7 Shows conviction but not correctional information . . . . 5 Ident record inaccurate when compared to local record . . . 34 Disposition information does not agree . . . . . . 15 Charging information does not agree . . . . . . . 11 Sentencing information does not agree . . . . . . 8 Ident record ambiguous when compared to local record . . . 11 Shows more dispositions than charges or vice versa . . . 3 Other ambiguities . . . . . . . . . . . 8 Complete, accurate, unambiguous . . . . . . . 43 NOTE: Although many records exhibited more than one record quality problem, only one per record iscounted above Earhest dateof samphng was 7/24/79 aDisposltion date s w ere a s follows2/14/7g, 1/IEi/79,1/10/79, I/16/79,12/20/78,12/12/78, 12/4/78,11/7/789/1/78, 7127178, 7117178, 617t78< 5130178, 3J2W78,311178, Z17178, 1117178, 1~21177, 11114177, WW77,5/2W77, 4115177,4111177, 3/14/77,2/22/77,11/29/76, 7/23/76, 6/25/76,6/14/76,4/12/76, 219176,1112176,12123175, 8/29/75(2),8/8/75,5/15/75, 4/16175, 3/5/75, 1/30/75, 1212174, 3118174, 5111173, 11/6/72, 9128172, 6/16/68, 12/20/67, 7/1/65, and 12/11/64 bDispos!tion dates were as follows6 /29/79, 6/26/79,6/5/79,6/3/79, 5/21/79,5/10/79,5/7/79, 4/30/79, 4/27/79(2~ and 413179 SOURCE. OffIce of Technology Assessment methodology which removed arrest events from further consideration if they were found to be pending (case still active and no disposition had occurred) or sealed (disposition had occurred but was sealed for legal reasons); if there had been no prosecution of the arrest; or if no record (or docket) was locatable. This latter category was significant since 16 percent of the NCIC/CCH sample (55 0ut of 257) were not verifiable due to no record (or docket) locatable. No record locatable generally refleeted a police disposition; that is, the charges were dismissed after arrest but prior to arraignment. For some of the arrests, the dismissal of charges was not noted in the Federal record, and thus would have been included as no disposition reported had those arrests not been removed from further analysis. Thus, the OTA analysis tends to understate the true level of arrests without dispositions. With respect to the importance of the record quality problems, there is general agreement that lack of dispositions is aand perhaps the problem. The FBI points out, however, that, except for Federal offenders, the responsibility for submitting dispositions lies with State and local C riminal justice agencies. Both NCIC and Ident encourage prompt submission of dispositions, and, indeed, Federal regulations require that dispositions be reported to the central State repository within 90 days after the disposition has occurred, 2 and to the FBI criminal history record systems within 120 days. 3 However, these regulations are difficult to enforce and few sanctions are available. With respect to incorrect or ambiguous information, the FBI believes that this is largely CFR j 20.21(a)(l). 3 28 CFR f 20.37.

PAGE 102

Ch. 8Record Quality in Federal and State Criminal History Information Systems l 9 3 Table 18.Record Quality of FBI NCIC/CCH File Disseminations, Based on 1979 Sample Arrests in sample . . . . . . . . . . . . . . 400 Local agency responses . . : . . . . . . . . . . 92 Arrests not verifiable because Pending or sealed . . . . . . . . . . . . . 3 No record locatable . . . . . . . . . . . . . 55 No prosecution of arrest . . . . . . . . . . . . 9 Fugitive . . . . . . . . . . . . . . . . 1 No arrest data . . . . . . . . . . . . . . 24 Total arrest cases verified . . . . . . . . . . . 165 Results: Actual disposition not recorded on NCIC/CCH record . . . . . . 48 Disposition occurred more than 120 days prior to 8/12/79 a . . . . . 45 Disposition occurred less than 120 days prior to 8/12/79 b . . . . . 1 Disposition date unknown . . . . . . . . . . . . 2 NCIC/CCH record otherwise incomplete when compared to local record. . . 7 Shows sentence but no conviction information. . . . . . . . 2 Shows conviction but no correctional information . . . . . . 5 NCIC/CCH record inaccurate when compared to local record . . . . 32 Disposition information does not agree . . . . . . . . . 13 Charging information does not agree . . . . . . . . . 10 Sentencing information does not agree . . . . . . . . . 9 NCIC/CCH record ambiguous when compared to local record . . . . 4 Complete, accurate, unambiguous . . . . . . . . . 74 NOTE Although many records exhibited more than one record quality problem, only one per record ls counted above Date of sampling was 9/12/79 aDlsposltion dates were as follows 4/5/79,3/21/79, 11/15J78, 9/lEY78, 7/27/78,7/3/78, 6/29/78, 6/8178,6/2/78,5/11/78, 4/27/78,4/21/78, 4/4/78,2/15/78, 12/2/77, 11/10/77, 10/26/77, 10/25/77, 10/19/77, 10/18/77,9/16/77, 6/8/77,4/8/77,4/4/77, 1/7/77, 12/22/76, 10/19/76, 9/30/76,10/21/75, 10/13/75,3/24/75,10/8/74, 7/16/74, 3/9{74,1/29/74, 11/26/73,8/28/73,10/5/72, 9126172,818/72,7124/72, 5/6/71,7/15/70, 6/23/70, and 4/13/70 bThedlsposttlon date was 5/3/79 SOURCE Offlceof Technology Assessment. attributable to the realities of the criminal justice process. For example, arrest charges (based on probable cause standards) may differ significantly from prosecutor charges (based on the necessity of proving a case beyond a reasonable doubt), which in turn may differ from final charges (frequently reflecting the results of plea bargaining). Thus, charges may change as a person moves through the criminal justice process, but these changes may not always be reported to the FBI. Also, people whose familiarity with the criminal justice process is limited may have particular difficulty in understanding and interpreting criminal history records. Nonetheless, based on the OTA research, a significant portion of Ident and NCIC/CCH records disseminated in 1979 appear to be incomplete, inaccurate, and/or ambiguous when compared with information in local records. State Criminal History Files A comparison between the 1979 0TA 50State survey and a 1973 General Accounting Office (GAO) study (based on a 197050-State survey conducted by the Law Enforcement Assistance Administration (LEAA)) shows some improvement in the average disposition reporting level over that 9-year period. Using bracketed averages, the GAO study found the average disposition reporting level to be about 52 percent for the 49 States responding. In comparison, the OTA study found the average Use of bracketed averages was necessary since OTA did not have access to the original 1970 LEAA State-by-State survey data. Based on 1970 data, the number of States and disposition reporting levels were as follows: 31 States (less than 65 percent); 11 States (65 to 90 percent); 7 States (more than 90 percent). From U.S. Comptroller General, Development ofa Nation wide Criminal Data Exchange System-,lleed to Determine Cost and Improve Reporting, General Accounting Office, January 1973, p. 10.

PAGE 103

94 An Assessment of Alternatives for a National Computerized Criminal History System disposition reporting level to be about 65 percent for the 41 States responding. 5 However, the 1979 average for computerized States (with a CCH file and/or automated name index) as opposed to manual States was even higher (about 71 percent compared to 50 percent for manual States). Given that in 1970 only one State (New York) had a CCH system, the results indicate that most of the improvement in disposition reporting over the 1970-79 period was in States with computerized systems. For the three major urban jurisdictions studied with respect to use of their State CCH files, when compared to the results of the OTA 50-State survey, the disposition levels for two of the jurisdictions (about 58 and 61 percent of arrests with dispositions, respectively, based on the OTA record quality research) were below the 71 percent average reported by 29 States with computerized systems in 1979. The disposition level for one urban jurisdiction (about 85 percent) was considerably above the average. All three were above the 50 percent reported by 12 States with manual criminal history systems in 1979. Several States contacted by OTA have achieved further improvement in disposition reporting since 1978. For example, in North Carolina, a mandatory disposition reporting requirement has gone into effect and some efforts to correct incomplete records from the largest jurisdictions have been initiated. The North Carolina State Bureau of Investigation computerized in 1976 and has since improved disposition reporting from 26 percent (1976) to 56 percent (1981). As of 1981, this States computerized police information system indicated a disposition reporting rate of about 75 percent. In California, strengthened field efforts over the last 3 years have increased the disposition reporting level for felonies from 66.6 percent in 1978 to 70.8 percent in 1980. However, between 1979 and 1982, average disposition reporting levels for all States responding improved only marginally, to about 66 percent. bFrom OTA 50-State survey. With respect to other aspects of record quality (e.g., inaccuracy, ambiguity), research results available to OTA were not adequate to draw any statewide or nationwide conclusions. The State criminal history records sampled by OTA in three urban jurisdictions and by Richard Faust (Tatum v. Rogers, S. D. N. Y., 75 Civ. 2782) in one urban jurisdiction, given the nature and size of the samples, cannot be considered as representative even of urban jurisdictions within the four individual States, and do not provide a valid basis for comparison of Federal, State, and local record quality. Significance of Finding s The significance of a given level of record quality depends in part on the applicable legal/ regulatory framework and how specific criminal history record information is actually used. On the one hand, Federal law as expressed in the Crime Control Act of 1973 (carried forward by the Justice Systems Improvement Act of 1979) and the resultant Federal regulations (28 CFR 20) make clear that all dispositions of criminal charges should be reported, as noted earlier, and that all records should be complete, current, and accurate. The FBI operating procedures emphasize that agencies that enter records into Ident or NCIC have the responsibility to assure that information on individuals is kept complete, accurate, and current. The FBI helps to maintain the integrity of the NCIC files through automatic computer edits and purges, quality control checks, and periodic record validations by originating agencies. 7 Similar procedures are possible in Ident through use of the Automated Identification Division System (AIDS), but have not yet been implemented. Ident is also considering the use of a disposition followup form for arrest events more than a year old with no disposition reported, and a possible interconnection with State and local automated systems to speed up disposition reporting. 628 CFR j 20.37. See NCIC Operating Manual; also see statement of William A. Bayse of the FBI before the House Committee on the Judiciary, Subcommittee on Civil and Constitutional Rights, Oct. 22, 1981.

PAGE 104

Ch. 8Record Quality in Federal and state Criminal History Information Systems 95 In addition, in December 1981, the NCIC Advisory Policy Board (APB) created a subcommittee on NCIC record quality. This subcommittee is empowered to study and report back to the APB on possible new approaches to improving record quality. At the State level, as of mid-1981, 46 States had some statutory provision requiring the reporting of dispositions, although not all of them set out reporting deadlines and relatively few of these statutes contain sanctions to make them mandatory. These requirements in part reflect the importance attached to accurate and complete criminal history information in the protection of individual rights of privacy, due process, and equal protection of the laws. Despite these Federal and State requirements, disposition reporting is still far from complete in Ident and NCIC/CCH and in a significant number of States. In the OTA 50State survey, 14 of 41 States responding in 1979 and 13 of 47 States in 1982 indicated that disposition reporting to the State repository was less than 50 percent. In both 1979 and 1982, eight States indicated a reporting rate of less than 25 percent. The significance of the use of criminal history record information with record quality problems such as lack of disposition data depends on who is using the information and for what specific purpose. With few exceptions, Federal and State law authorizes the dissemination of criminal history informationwith or without dispositions, whether accurate and complete or not to the criminal justice community. Law enforcement and prosecuting agencies, in particular, note that it is recognized that criminal history records are frequently incomplete and/or inaccurate, but that these records are very useful as a pointer to the location of complete and accurate information. In a variety of situations, such as setting conditions for pretrial release, arrest-only records are useful to judicial officials. However, in criminal trial proceedings, the laws of SEARCH Group, Inc., fiends in State Security& Privacy Legislation, Sacramento, Calif., November 1981, p. 12. SEARCH, Security & Privacy, op. cit., pp. 9-10. criminal evidence in most jurisdictions preclude the admission or even consideration of uncertified criminal history records, and most certainly arrest-only records. On the other hand, Federal regulations permit dissemination of Ident and NCIC/CCH records without dispositions to Federal noncriminal justice agencies if authorized by Federal statute or Executive order. Dissemination is also permitted to State and local noncriminal justice agencies if authorized by Federal or State statutes and approved by the U.S. Attorney General. Dissemination of records without dispositions is prohibited only if the arrest charge is more than 1 year old and is not under active prosecution. l0 At the State level, as of mid-1981, 37 States authorized dissemination of arrest-only records to a variety of State and local noncriminal justice agencies (primarily for employment and licensing purposes), and 27 States authorized such dissemination to private sector organizations and individuals. In a large number of States, the disclosure of such records to private parties depends upon factors other than State law, such as local law, local agency policy, or the impact of the States public record or freedom of information law. 11 The wide dissemination of criminal history records with known record quality problems, especially missing or inaccurate disposition information, raises legitimate questions about: 1) the efficiency of law enforcement and criminal justice programs that use or rely on such records, notwithstanding their value as a pointer to more complete and accurate information; 2) the protection of constitutional rights (especially due process and equal protection of the laws) where such records are used in criminal justice decisionmaking; and 3) the protection of rights to privacy as well as due process and equal protection where such records are used for noncriminal justice purposes, as in employment and licensing decisions. These questions become even more important in light of many of the recommenda 0 28 CFR j 20.33. SEARCH, Security & Privacy, op. cit., p. 10.

PAGE 105

96 l An Assessment of Alternatives for a National Computerized Criminal History System tions of the Attorney Generals Task Force on convicted)l z which are intended to protect the Violent Crime (e.g., with respect to denial of public safety, but which depend in turn on bail to a person accused of a serious crime who high quality criminal history records. had previously committed, while in a pretrial release status, a serious crime for which he was *Attorney Generals Task Force, op. cit., p. xi.

PAGE 106

Chapter 9 State and Local Management of Criminal History Information Systems

PAGE 107

. -.. Contents Summary. o.... . . . . . . . . . . Locus of Authority . . . . . . . . . . . Arrest and Court Disposition Reporting . . . . . . . File Content. . Access, Review, and . . . . . . . . . . . Challenge Procedures . . . . . . Dissemination of Criminal History Information . . . . . Institutional Basis for Dissemination Applicability of Dissemination Policy Sealing and Purging . . . . Definition . . . . . Variation Among States . . Record Accuracy and Completeness Court Disposition Monitoring . . Transaction Logs and local Audits. Policy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . TABLES Table No. . . . . . . . ,..,, . . . . . . 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. Arrest and Court Disposition Reporting . . . Institutional Basis for Court Court Disposition Reporting: OTA Findings.. . . Disposition Reporting Comparison of GAO and . . . . . . . . . . . . . . Review, and Challenge of Criminal Criminal History Dissemination Policy Statutory Limitations on the Content of Criminal History Files . State Procedures for Access, History Records . . Institutional Basis for State Applicability of State Policies on Dissemination of Criminal Justice Information . . . . . . . . State Agencies That Have Conducted Record Quality Audits Criminal History Information Stored in State Repositories . . Procedures Used by State Agencies to Monitor Court Dispositions Information Contained in Dissemination Logs . . . . Procedures Used to Review Dissemination Logs . . . . . o f . . Page 99 100 100 102 102 103 103 103 103 103 104 104 104 105 Page 101 101 102 102 102 103 103 104 105 105 105

PAGE 108

Chapter 9 State and Local Managemen t of Criminal History Information System s Chapte r State and local agencies are the largest users of the National Crime Information Center (NCIC) and Identification Division (Ident) criminal history files. These files in turn largely rely on information generated by States and localities. Thus, it is important to assess the extent to which Federal and State statutes and regulations governing the management of criminal history information are in fact being implemented at the State and local levels. In 1979-80, OTA conducted a 50-State survey of management policies and practices (to which 48 States and 1 territory responded). A written questionnaire was sent by OTA to the State Governors and completed by State criminal record repository personnel. In 1982, OTA conducted a follow up telephone survey (to which 50 States and 1 territory responded) to determine if changes had occurred in key areas. ** Locus of Authority: In 40 States, there was a single State agency responsible for the development of a statewide privacy and security plan for criminal history information, but the nature of this authority appeared to be highly variable. C)ffice of Technology Assessment 50-State Survey conducted in 1978-80. Written questionnaires were sent to the Governors of all 50 States and Puerto Rico and the Mayor of the District of Columbia. As of March 1980, the final extended deadline, responses had been received from 48 States and Puerto Rico. For purposes of analysis, Puerto Rico was treated as a State. *see app, B for a list of the State officials who completed written questionnaires for the 1979 OTA 50-State survey. *The OTA followup survev was conducted by telephone dUring August 1982. See appu D for a list of State officials contacted. Summar y Arrest and Disposition Reporting: As of 1979, about 78 percent of arrests and 65 percent of court dispositions were reported to State repositories, on an average. The reporting rates of computerized States were significantly higher than States with manual criminal history systems. As of 1982, reporting rates had improved marginally to 82 percent for arrests and 66 percent for dispositions. File Content: In more than one-third of the States, there were no statutory limitations on criminal history file content. In another third, all offenses which are fingerprintable or result in incarceration were reported. Access, Review, and Challenge Procedures: While most States had these procedures, more than half of the States did not maintain data on the frequency of requests for access, review, and challenge of criminal history records. Data that were collected indicated infrequent requests. Dissemination of Criminal History Information: Almost three-quarters of the States had a dissemination policy that applied to all criminal justice and other agencies that use or maintain criminal justice information. In more than half the States, the policy was based on specific statutes. Sealing and Purging: State policies for sealing and purging vary significantly. Some States (e.g., New York) seal arrest events that do not result in conviction; other State systems maintain any police contact information as a permanent part of a criminal history record. 99

PAGE 109

100 l An Assessment of Alternatives for a National Computerized Criminal History System Record Accuracy and Completeness: As of 1982, about two-thirds of the States (35 out of 46 States responding) indicated that they did routinely employ a set of procedures to assure the accuracy of criminal history information. This represented a significant improvement over 1979, when only 23 out of 46 States employed such procedures. About three-quarters of the States had never conducted a record quality audit of either computerized or manual criminal history record systems, with no significant change between 1979 and 1982. Court Disposition Monitoring: Only 17 States in 1979 and 19 States in 1982 had automated procedures for monitoring court dispositions to help ensure record completeness. All but one of these States had computerized systems. Transaction Logs and Local Audits: Almost all States maintained logs of criminal history information disseminated by the State agency, although nearly two-thirds reviewed these transaction logs only when a specific abuse was indicated. Frequent systematic review was more prevalent in States with computerized systems. Only 12 States reported systematic audits of user agencies, and here the procedures varied widely. Locus of Authorit y In 48 States (out of 49 responding), as of 1979, there was a single State agency responsible for developing a statewide privacy and security plan for criminal history information, but the nature of this authority appeared to be highly variable. The authority was based on State statutes in about one-half (26) of the States; in another 20 the basis is executive policy (13) or order (7). The authority was unofficial or nonexistent in three States. Other structures, such as other State agencies and commissions on privacy and security, shared the authority with the designated agency in over half of the States. The responsible State agencies exercised considerable discretion in the development of regulations and management practices to implement title 28. 2 Title 28, Code of Federal Regulations, pt. 20, Criminal Justice Information Systems, subpt. B, State and Local Criminal History Record Information Systems. Arrest and Court Disposition Reportin g To a considerable extent, the accuracy and completeness of criminal history information depends on the ability of State CCH systems to assure the reporting of arrests by local police and the reporting of dispositions by courts. State managers indicated that as of 1982, 82 percent of reportable arrests and 66 percent of all court dispositions were reported to State repositories, as shown in table 19. This represented a marginal improvement over 1979 reporting levels. In 1979, only about half of the States had statutes to support an institutional basis for court disposition reporting; the other half operated by formal agreement or informally according to custom and tradition, as indicated in table 20. Between 1979 and 1982, the number of States with disposition reporting statutes increased from

PAGE 110

Ch. 9State and Loca/ Management of Criminal History Information Systems l 101 Table 19.Arrest and Court Disposition Reporting All States Computerized a Noncomputerized b Arrest reporting 1979 (N = 43) (N = 30) (N = 13) 78 3A 83.00/0 67.7 0/0 1982 (N = 47) (N = 32) (N = 15) 81 50/, 86.1 % 71 .80/0 Percent of arrest Number of States Distribution of arrest reporting 1979 1982 1982 025 2 1 1 2650 7 6 6 5175 4 5 5 76-100 30 31 35 43 47 All States Computerized Noncomputerized Court disposition reporting 1979 (N = 41) (N = 29) (N = 12) 64.80/o 70.9/ 49. 7% 1982 (N = 47) (N = 33) (N = 14) 66.40/0 70 60/0 56.3 % Percent of disposition reporting Number of States Distribution of court disposition reporting 1979 1982 1982 025 8 7 8 2650 6 3 5 5175 7 10 12 76-100 20 21 22 41 47 a computerized name Index and/or CCH file b Manual Index and file SOURCE Office of Technology Assessment 50-State Survey and 1982 followup 26 to 29. Both arrest and court disposition reporting were significantly higher for States with CCH systems (with an automated name index and/or CCH file) compared with States with manual systems. When compared with the results of a 1973 General Accounting Office study (based on 1970 Law Enforcement Assistance Administration data), the OTA results indicate significant improvement over the 1970-79 period as shown in table 21, but little improvement since 1979. Table 20 .Institutional Basis for Court Disposition Reporting 1979 Number of States 1982 Number of States A formal system mandated by statute. 26 (53.1 0 /0) 29 (59.20/o) A formal system by agreement with courts . . . 7 (14.3%) 6 (1 2.2/0) An informal system . . . . 6 (12.2/0) 5 (10.20/0) No system; depends on jurisdiction . 10 (20.40/o) 9 (18.40/o) Total . . . . . . . 49 (100.0%) 49 (1 OO.0%) SOURCE Office of Technology Assessment 50-State Survey and 1982 followup

PAGE 111

102 l An Assessment of Alternatives for a National Computerized Criminal History System Table 21 .Court Disposition Reporting: Comparison of GAO and OTA Findings 1973 1979 GAO study a OTA study b (1970 data) (1978-79 data) 1982 OTA study Disposition reporting rate N = 49 N = 41 N = 41 N = 47 Less than 65/0: Number of States . . 31.0 19.0 17.0 22.0 Percent. . . . . 63.30/o 46.30/o 41 .5% 46.80/o 65 to 90/0: Number of States . . 11.0 10.0 11.0 12.0 Percent . . . . . 22.40/o 24.40/o 26.80/o 25.50/o More than 90/0: Number of States . . 7.0 12.0 13.0 13.0 Percent. . . . . 14.3/0 29.30/o 31 .7% 27.70/, US Comptroller General. Development ofa FhWonwide Cmnmal Dafa Exchange Sys/emfVeed (o Determme Cosf and improve Report/rig, General Accounting Off Ice, January 1973, p 10 b OTA 50-State Survey Dlsposltion reporting rates provided by State repository of flclals C OTA 50. State Survey, 1982 followup SOURCE Off Ice of Technc,logy Assessment and General Accounting Office File Conten t A major finding of the OTA survey was the considerable variability in the nature of crimes reported to State systems. In more than a third of the States there were no statutory limitations on criminal history file content, and in another third all offenses that are fingerprinted or result in incarceration were reported (table 22). In some States, misdemeanors were considered to be fingerprintable offenses that create a criminal record, but in others the vast majority of misdemeanors were not so considTable 22.Statutory Limitations on the Content of Criminal History Files Statutory limitations on file content Number of States Felony or NCIC criterion felony only . . . . . 2 ( 4.1 0 /0) Felony and gross, indictable or serious misdemeanors only . 12 (24,50/o) For all adult offenses which are fingerprintable or result in incarceration . . . 17 (34.7%) No statutory limitations on criminal history file content . . . 18 (36.70/o) 49 (100/0) ered. Access Most States had review, challenge, SOURCE Office of Technology Assessment 50. State Survey Review, and Challenge Procedures procedures for access and =. appeal, and the like, as Table 23.State Procedures for Access, Review, outlined in table 23. There appeared to be no and Challenge of Criminal History Records significant differences between computerized and noncomputerized States in this regard. However, data on the frequency of use of these procedures, where collected, indicated infrequent use. During 1978, the number of access requests ranged from 110 in Delaware, to 44 in Oregon, 43 in Maine, 32 in Minnesota, to 12 in Florida, 2 in Arizona, 1 in Vermont, and none in Virginia. Number of States using procedures Procedures (N = 49) Individual access and review . . 45 Challenge . . . . . . 44 Appeal . . . . . . 43 Correcting information . . . 41 Informing public of right to access and review . . . . . 32 None . . . . . . . 4 SOURCE Office of Technology Assessment 50-State Survey

PAGE 112

Ch. 9State and Local Management of Criminal History Information Systems 103 Dissemination of Crimina l History Informatio n Regulating the dissemination of criminal history information is another strategy employed by the States to protect the privacy of individuals, while at the same time retaining the maximum amount of information deemed necessary for their own needs. Nearly all States include dissemination regulations in their statutes, but the degree of detail varies. Institutional Basis fo r Dissemination Polic y Over half of the 49 States responding indicated that State statutes with specific reference to criminal history or criminal justice information formed the bases for their dissemination policies, as shown in table 24. Applicability o f Dissemination Polic y Seventy-one percent of the States responded that their dissemination policy applied to all criminal justice and other agencies that use or maintain criminal justice information, as shown in table 25. The dissemination policy applied only to the central State Repository in 22 percent of the States responding. Table 24.institutional Basis for State Criminal History Dissemination Policy Number of States Institutional basis (total 49) State statutes with specific reference to criminal history or criminal justice information . . . . 27 State repository enabling legislation 16 Public or open records law . 8 State repository agency policy . 15 Executive order . . . . 4 Administrative procedure . . 12 Dont know . . . ... . 1 SOURCE Office of Technology Assessment-50-State Survey Table 25.Applicability of State Policies on Dissemination of Criminal Justice Information Central State repository only . 11 (22.40/o) Central repository and some local agencies . . 2 ( 4.1%) All criminal justice and other agencies that use or maintain criminal justice information . 35 (71 .4%) None . ... ... ... 1 ( 2.0%) 49 (1 OO%) SOURCE Off Ice of Technology Assessment 50 State Survey Sealing and Purgin g 3 An important aspect of criminal history information policy is-the ability of a system to: 1) purge or seal records for selected persons and/or offenses; 2) remove the punitive effects of a criminal history record for selected persons (removal of disqualifications); and 3) permit individuals under selected circumstances to freely state the nonexistence of a record. The sealing and purging capabilities within a system are important for the protection of in For a detailed discussion, see SEARCli Group, inc., Sealing and Purging of Criminaj Histor~ Information, Sacramento. Calif., April ,1981. Number of States dividual rights, as well as for the efficient management of large record files. Definitio n State statutes on sealing and purging reveal a rather confusing variety of terminology used as well as the type of information that is sealed or purged. Terms like deleted, annulled, returned to the individual, and expunged are used, sometimes interchangeably with purging and sealing. Where the meaning is clear, purging is generally defined as taking

PAGE 113

104 An Assessment of Alternatives for a National Computerized Criminal History System place when records are physically destroyed or returned to the individual; sealing is defined as taking place when they are not destroyed, but are not accessible to the public at large, or perhaps even to the criminal justice community. Variation Among State s States vary widely in terms of statutory and management purge and seal policies. Some States seal arrest events that do not result in conviction (e.g., New York), whereas other State systems maintain any police contact inRecord Accurac y As noted in the research on record quality in State and Federal systems, the level of record quality varies enormously from one State to another. In the 50-State survey, as of 1982, about two-thirds of the States (35 out of 46 responding), both computerized and noncomputerized, indicated that they did routinely employ a set of formal procedures (generally known as quality control checks or validity checks on input data) to assure the accuracy of criminal history information. This represented a significant improvement over 1979, when only 23 out of 46 employed such procedures on a routine basis. The 12 additional States with quality control checks are all either computerized (with an automated name index and/or CCH file) or in the process of computerizing. While most State CCH repositories now have procedures to assure the reliability of information as a permanent part of a criminal history record. As of mid-1981, 35 States had statutes or regulations on purging nonconviction information, and 24 States had laws on purging conviction information. Statutes or regulations on sealing nonconviction information had been enacted in 20 States, and on sealing conviction information in 22 States. 4 SEARCH Group, Inc., Trends in State fkxx.wity and Fkivacy l~gislation, Sacramento, Calif., November 1981, p. 5. and Completenes s formation put into the system, as of 1979 about three-quarters had never conducted an audit of the quality and validity of information stored in CCH repositories, as indicated in table 26. A 1982 followup indicated no significant change. States implementing a record quality audit since 1979 were offset by States cutting out or drastically reducing their existing audit function, primarily due to budget and staff reductions. Table 26.State Agencies That Have Conducted Record Quality Audits of Criminal History Information Stored in State Repositories Number of States Conducted quality audit . . . 13 (26.50/o) Never conducted quality audit . 36 (73.5/0) 49 (1000/0) SOURCE Office of Technology Assessment 50-State Survey, and 1982 followup Court Disposition Monitorin g The most significant record quality problem there were no statutes to support a formal sysin State systems, which is reflected in Federal tern of court disposition reporting. Seventeen NCIC and Ident files, is record incompleteness of the States in 1979 and 19 States in 1982 inthe failure to capture court disposition infordicated that they had automated procedures mation. for the routine review of disposition completeAs shown in table 20, in about two-fifths of ness, as shown in table 27. With the exception the States responding to the survey, as of 1982 of one, all of these had computerized systems.

PAGE 114

Ch. 9State and Local Management of Criminal History Information Systems 105 Table 27. Procedures Used by State Agencies to Monitor Court Dispositions 1979 Number of States 1982 Number of States Automated review of file . . . 17 (34.7%) 19 (38.80/o) Manual review of file . . . 8 (16.30/o) 11 (22,40/o) Sometimes inquire of courts before dissemination . . . . . 5 (10.2%) 4 ( 8.20/o) No review of delinquent dispositions 18 (36,80/o) 14 (28.60/o) Dont know . . . . . . 1 ( 2.0%) 1 ( 2.00/0) 49 (100%) 49 (100/0) SOURCE Off Ice of Technology Assessment 50. State Survey and 1982 followup Transaction Logs and Local Audit s One intent of Federal and State regulations is to ensure that the flow of CCH information can be accounted forwho received what information, what was the purpose, what was the type of information, and what happened to the information after its use. Implementing this intent of the regulations requires formal transaction logging procedures and audits of local users to ensure that the procedures are being followed. Only two of the 49 responding States reported that as of 1979 they did not maintain dissemination logs, and one other indicated that it was in the process of developing a dissemination log. Logs of most States contain information on the name or identification number of the requesting agency and the type of information disseminated, as shown in table 28. Logs in about two-thirds of the States also contain information on the purpose of the request, the requestors terminal identification number, and the name or identification number of the person requesting information. While nearly all States maintained logs of criminal history information disseminated by the State agency (as of 1982 only one State did not maintain a log), nearly two-thirds reported that they reviewed transaction logs only when a specific abuse was indicated, as shown in table 29. Frequent systematic monitoring of transaction logs was more prevTable 28.information Contained in Dissemination Logs Number of States Type of information (N == 49) Name or I.D. of requesting agency 46 Purpose of request ., ... ... 34 Requesters terminal code I.D. . 32 Type of information disseminated ... 43 User agreement or authority base code, ... ... ... 13 Name or I.D. of person requesting information. . . 32 SOURCE Off Ice of Technology Assessment 50-State Survey Table 29. Procedures Used to Review Dissemination Logs Number of States Frequent, systematic monitoring of user activity . 12 (24.50/o) Annual monitoring of user activity 3 ( 6.1 /o) Review of logs generally only when a specific abuse indicated 29 (59.20/o) None . . . . . 6 (12.2/0) 49 (100.0%) SOURCE Office of Technology Assessment 50 State Survey alent in States with computerized systems than in those with manual systems. In the 12 States that reported systematic audits of user agencies, these procedures varied from occasional visits by State audit teams to, in some cases, the completion of questionnaires by local officials testifying to their compliance with State and Federal regulations.

PAGE 115

Chapter 1 0 Major Structural Alternatives for a National Computerized Criminal History System

PAGE 116

Contents Page Chapter Summary. . . . . . . . . . . . 109 National Repository . . . . . . . . . . 109 Single-State/Multi-State . . . . . . . . . . 109 Interstate Identification Index Pilot and Phase 1 Tests . . . 110 National Index . . . . . . . . . . . . 110 Regional and Ask-the-Network Systems . . . . . . 110 National Repository . . . . . . . . . . . 111 Single-State/Multi-State. . . . . . . . . . . 112 Interstate Identification Index Pilot Phase I Tests . . . . 114 National Index . . . . . . . . . . . . 118 Regional and Ask-the-Network Systems . . . . . . . 119 Regional CCH Systems. . . . . . . . . . . 119 Ask-the-Network System . . . . . . . . . 120 FIGURES Figure No. Page 7. National Repository CCH Altematives . . . . . . 112 8. Single-State/Multi-State CCH Altemative With Message Switching. 113 9. Interstate Identification Index Pilot Test as Proposed in 1980.. . 115 10. Interstate Identification Index Pilot Test as Conducted in 1981 . 116 11. National Index CCH Alternative With No Message Switching . 119 12. Decentralized Ask-the-Network CCH Altemative With a National Switcher . . . . . . . . . . . . . 121

PAGE 117

Chapter 10 Major Structural Alternatives for a National Computerized Criminal History Syste m Chapte r Over the last 12 years, a wide range of structural alternatives have been proposed for a national computerized criminal history (CCH) system. These alternatives can be grouped into four categories: 1) national repository; 2) single-State/multi-State; 3) national index; and 4) regional and ask-the-network systems. Each category has a number of possible variations. From a technological systems perspective, three significant changes have occurred since the debate over CCH began. First, advances in computer and communication technology have reached the point where both centralized and decentralized system structures are possible. Second, many States and localities and several Federal agencies have developed their own CCH capability, and 49 of the 50 States now have their own criminal history record repositories. Third, the Federal Bureau of Investigations (FBI) Identification Division (Ident) has made progress in automating its own operations through the Automated Identification Division System (AIDS) program. All of the CCH alternatives discussed below assume that Ident (or its equivalent) will continue to provide a national fingerprint identification capability. National Repositor y Both Ident and NCICICCH are currently structured as national repositories with records of single-State, multi-State, and Federal offenders for criterion offenses. Ident is, indeed, fully functioning as a national reposiSummar y tory since all 50 States submit fingerprints. NCIC/CCH is operating much like a Federal repository with national access rather than a national repository, since all current Federal offenders are included but only eight States are contributing records. A national CCH repository could evolve from either the AIDS file of Ident or the CCH file of NCIC. In actual practice, the repository would likely draw on elements of both. Single-State/Multi-Stat e The original FBI plan was eventually to implement the single-State/multi-State alternative with the switching of messages through the NCIC computer for both the record inquiry and the response. States would maintain single-State offender records. NCIC/CCH would include records of multi-State and Federal offenders, plus an index of single-State offender records. Inquiries against the index resulting in a hit would be routed through the NCIC computer and over the NCIC communication lines from the requesting State or agency to the originating State. The record of interest would be transmitted to the requesting State via the NCIC network. Some alternatives involving message switching have raised questions about the impact on Federal-State relations and the potential for monitoring and surveillance use. The FBI has argued that message switching would provide a legitimate service to the States that would improve efficiency and provide a faster response time. Nevertheless, Congress has continued to prohibit NCIC/CCH message switching. 109

PAGE 118

110 An Assessment of Alternatives for a National Computerized Criminal History System Interstate Identification Inde x Pilot and Phase 1 Test s In early 1980, the FBI advised Congress of an Interstate Identification Index (III) pilot project with Florida to demonstrate the functional equivalent of the single-State/multiState alternative with no message switching, viewed as the first step toward implementing III. However, owing to problems with handling widely varying message formats and in verifying requests from the other States, a plan was developed for routing all inquiries to Florida through NCIC/CCH with records provided via NLETS. In December 1980, this inquiry referral plan was endorsed by the NCIC Advisory Policy Board and the NLETS Board of Directors. Thus, the pilot test conducted during July through September 1981 (as well as the Phase 1 test carried out in spring 1982) involved a partial message switching technique known as automatic inquiry referral. National Inde x The current III development plan calls for a national index. In a national index system, States would maintain both single-State and multi-State offender records. The NCIC/CCH file would then include only Federal offender records, plus a national index (sometimes known as a pure pointer index, a national criminal identification name index, and, more recently, III) of single-State and multi-State records. If implemented without any NCIC/ CCH message switching, States or agencies making an inquiry would be advised only if the subject were listed in the index as having a record and, if so, in which State repository. The requesting agency would then obtain the record directly via NLETS or other means. The national index alternative could also be implemented with partial or complete message switching via NCIC/CCH. A national index would avoid the expense of duplicating records at both State and Federal levels and would preserve State control over in-State records. Regional an d Ask-the-Network System s In the 1979 OTA survey, several States listed decentralized regional systems between contiguous States as a secondary preference and a better-than-nothing alternative. However, most States contended that regional systems were infeasible or impractical. An analysis of NLETS and III pilot test traffic patterns indicates that Florida is receiving messages from distant States more often than from contiguous States. In a completely decentralized national askthe-network system, there would be no national index or repository. Instead, each State could poll any or all of the other 49 States when seeking CCH information. One option would tie all 50 States and the FBI together on a computerized party line. A more likely option would be the use of a national switcher similar to the one operated by NLETS in Phoenix, Ariz. With its technology upgrade now complete, NLETS is operating at about 7 to 10 percent of capacity and could handle a substantial increase in CCH-related message traffic. However, the potential for use of NLETS in an askthe-network mode must be tempered by the experience with ROIR (reply only if record) messages. Here, inquiring States looking for a record would send messages to all other States (or a large number). NLETS found that many States began to ignore messages when the probability of a hit was low and the effort (and cost) of checking out all inquiries was high. Also, the FBI and various State officials believe that an ask-the-network system would not be cost effective and would be harder to secure against unauthorized access.

PAGE 119

Ch. 10Major Structural Alternatives for a National Computerized Criminal History System 111 Nationa l Both Ident and NCIC/CCH are currently structured as national repositories. That is, the criminal history files are designed to include full records of all offenderssingle-State, multi-State, and Federal-for all criterion offenses.* However, NCIC/CCH was intended by the FBI to serve as a full record national repository only until the single-State/multiState concept could be implemented. In practice, Ident is fully functioning as a national repository since all 50 States submit fingerprints. The NCIC/CCH file falls far short, however, since only eight States currently maintain criminal history records in CCH. All Federal offenders are included in NCIC/CCH, and 49 of the 50 States can directly access the CCH file. All NCIC/CCH entries require an FBI identification number, which in turn must be based on positive fingerprint identification. Ident conducts the fingerprint identification and assigns the identification number. Given the high rate of Federal agency participation in NCIC/CCH and the low rate of State participation, NCIC/CCH is operating much like a Federal repository with national access rather than as a national repository. About two-thirds of the non-Federal CCH traffic is with the eight fully participating States. NCIC/CCH differs from Ident in three other major ways. First, the NCIC/CCH file uses the NCIC communication network for receiving and sending messages and can respond to inquiries from the States in a matter of seconds. Ident must depend primarily on the mails. Even if AIDS were able to substantially reduce the internal turnaround time for Ident, the total response time would still be measured in days (by mail) or hours (with facsimile transmission) rather than seconds, since Ident Includes serious and)or significant offenses. Excludes the offenses of drunkenness, vagrancy, disturbing the peace, curfew violation, loitering, false fire alarm, nonspecific charges of suspicion or investigation, and traffic violations (other than manslaughter, driving under the influence of drugs or liquor, and hit and run). See 28 CFR $20.32. Repositor y conducts fingerprint NCIC/CCH involves only name searches. Indeed, searches whereas online computerized all records in the NCIC/CCH file must be based on positive fingerprint identification. Another difference between NCIC/CCH and Ident is in record format. Ident rap sheets and the automated AIDS rap sheets present criminal history information in chronological order, limiting contents to the offenders name, Federal and State identification numbers, arrest dates, charges, and dispositions (with dates). The NCIC/CCH format includes additional personal descriptor information (height, weight, identifying marks, etc.), presents charges and dispositions in summary as well as chronological order, and makes provision for additional judicial and custody information plus supplemental comments. Thus, the utility of a CCH record is presumed to be greater than that of an Ident/AIDS record, but the costs and difficulty associated with keeping a CCH record up to date are also higher. A third difference is in the area of privacy and security requirements. While both Ident and NCIC/CCH have the same statutory and regulatory frameworks for the use and dissemination of criminal history records, NCIC/CCH has developed a much more stringent set of operating procedures to protect the privacy and security of criminal history records, in part because CCH is an on-line file. It is possible for a national CCH repository to evolve from either the AIDS file of Ident or the CCH file of NCIC. In actual practice, the repository would likely draw on elements of both, as shown in figure 7. In theory, a national repository would not require message switching. Since all States and agencies would enter criminal history records into the repository and update these records on a continuous basis, confirmation of hits (a match between an inquiry and a record) with the originating States or agencies would not be necessary. In practice, unless disposition reporting were virtually instantaneous, confirmation via NLETS

PAGE 120

112 An Assessment of Alternatives for a National Computerized Criminal History System or by some other means would be needed to are now), as illustrated in figure 7, there would ensure record accuracy and completeness. As be no need for the national CCH repository to long as the confirmations were carried out by conduct message switching. the States and agencies themselves (as they Figure 7. National Repository CCH Alternative n Requesting agency Criminal history record f i lea Computer b I r P Iines c Request for confirmation of record Confirmation reply aeon [al~ all ~lngle.sta~e, ~ It ,. State, and FedeTal ~ffender~ Could us,e Al DSj data base when f u I Iy automated, and A IDS or CCH record fOrMaf Or some corn t)l nat Ion bcould use NCICor AIDS computer cCould use NCIC comm unlcat[on Ilnes SOURCE Off Ice of Technology Assessment Single-State/Multi-Stat e While the NCIC/CCH file currently serves native whereby States would maintain singleas a national repository for the records of eight State offender records. A central repository States plus all Federal offenders, this arrange would include records of multi-State and Fedment was originally viewed by the FBI as eral offenders plus an index (composed of name transitional. The plan was eventually to imand identifiers only) of all single-State offender plement the single-State/multi-State alterrecords contained within the State repositor-

PAGE 121

Ch. 10 Major Structural Alternatives for a National Computerized Criminal History System l 113 .ies. When the NC IC/CCH program began 12 years ago, very few States had an in-State CCH capability for their own records. Thus, NCIC/CCH initially was to maintain singleState offender records (as well as multi-State and Federal), but would return single-State records to the States as they developed their own CCH systems. The functional equivalent of this concept was first tested in a 1981 pilot project with the State of Florida. 1 Approximately two-thirds of the 25 States that now have their own on-line CCH systems are not For a discussion of the results, see FBI, Interstate ZclentificaCion Index Background and Findings for Juibv-September 1981 Phase 1 Pilot Project, Dec. 4, 1981. Figure 8.Single-State/Multi. State CCH CCH file a ? v I .- .. presently contributing records to NCIC/ CCH.* Under the single-State/multi-State alternative proposed by the FBI in 1970, inquiries against III resulting in a hit would be routed through the NCIC computer and over the NCIC communication lines from the requesting State or agency to the State where the record originated. The record of interest would then be sent back to the requesting State via the NCIC computer for both the record inquiry and the response, as illustrated in figure 8. *As of Auwst 1981, ~Ven of the eight StateS fly Participating in NC ICICCH had their own on-line State CCH file. Eighteen other States with on-line CCH files were not participating. Alternative With Message Switching r ) Requesting agency Record inquiry k Record provided NCIC NCIC computer lines 4. Record provided If a hit on index, Index b then record requested u Originating agency a For mu I t I Slate and Fed era I offender records b F or slng{e State offender records located In State rePosltorles SOURCE Office of Technology Assessment

PAGE 122

114 l An Assessment of Alternatives for a National Computerized Criminal History System .... In 1973, the FBI proposed to have NCIC assume all law enforcement message switching (not just NCIC/CCH traffic), including messages sent over NLETS. This alternative, known as single-State/multi-State with full message switching, has generated policy questions about the impact on Federal-State relations and the potential for monitoring and surveillance use. 2 In the face of opposition, the For a detailed review of the message switching controversy, see Donald A. Marchand, et al., A History and Background Assessment of tbe National Crime Information ilmi%r and Computerized Criminal History Program, Bureau of Governmental Research and Service, University of South Carolina, June 1979, sec. IV, pp. 122-167. See also related discussion in ch. 5 for further discussion. The FBI prefers to use the term limited message switching to mean full message switching of NCIC-related messages, and full message switching to refer to full message switching of all inter-State crimin al justice messages. FBI proposed to limit its message switching to NCIC/CCH-related traffic, and in 1975 circulated a limited message switching implementation plan. The FBI has argued that message switching is within FBI authority, would provide a legitimate service to the States, and would provide CCH records located in State systems faster and more efficiently. Nevertheless, Congress has continued to prohibit NCIC/CCH message switching. In 1979 and 1980, Congress conditioned approval of the NCIC technology upgrade (for the frontend processor and mainframe computer) on the strict prohibition of any message switching applications. *See ch. 5 for further discussion, Interstate Identification Inde x Pilot and Phase 1 Test s In early 1980, the FBI officially advised Congress of a III pilot project to demonstrate the functional equivalent of the single-State/ multi-State alternative with no message switching, viewed as the first step toward implementing 111. 3 The plan was for NCIC/CCH to return all single-State offender records to selected State repositories and establish an index to these records in NCIC/CCH. Thus, when a request would come in for one of these records, the NCIC/CCH index would indicate that a CCH record existed in a particular State and that the requesting agency should contact that State directly (via telephone, mail, or teletype), as illustrated in figure 9. It was anticipated that most agencies would use NLETS, which would perform the message switching function for both inquiries and record responses. Florida was selected as the pilot State. However, early in the planning stage, Florida con.. 3 See identical letters dated Jan. 7, 1980, from the FBI Director to the Chairman of the Senate Judiciary Committee and the Chairman of the House Judiciary Subcommittee on Civil and Constitutional Rights. eluded that it would encounter a number of problems in handling the incoming record inquiries from agencies in the other States because the formats of the requests would vary widely and verification of agency authorization would be difficult. Therefore, a plan was developed for routing all inquiries to Florida through NCIC/CCH. In this way, NCIC/CCH would use a consistent message format and verify agency requests. This, in effect, would give NCIC/CCH a partial message switching role, in that inquiries (messages) eliciting hits on the index would be switched through the NCIC computer and over the NCIC communication lines to Florida, which in turn would provide the CCH record to the requesting agency via mail and/or NLETS, as shown in figure 10. The single-State/multi-State pilot test with partial message switching (known as automatic inquiry referral or AIR) was completed during July through September 1981. The test provided useful data on the number of inquiries, hits, and records provided, response time, and perceived value of the records ulti-

PAGE 123

Ch. 10Major Structural Alternatives for a National Computerized Crimina/ History System 115 Figure 9. Interstate Identification Index Pilot Test as Proposed in 1980 Index b / NCIC computer a For mu III State and Federal offender records b For ~,ngle State offender records located n State rePOst~res c V I a te Ie p h one rn al I or fe let y pe e q N L E TS SOURCE Office of Technology Assessment mately provided. Phase 1 of the III development plan extended the pilot test to include five additional StatesMichigan, North Carolina, South Carolina, Texas, and Virginia-and was conducted in February and March 1982. The FBI asserted that AIR does not involve message switching since the inquiries from requesting agencies are reformatted by NCIC. Some information is deleted and other information added before referring the inquiry to the State of record in the case of a hit (or a match between an inquiry and an index entry indicating a single-State record). In addition, Record inquiry NCIC communication Record lines request c Record provided c the FBI noted that both the NCIC Advisory Policy Board and the NLETS Board of Directors had approved the use of AIR, and that the appropriate congressional committees were advised in advance of its use in the pilot project. The FBI concluded that AIR notification is not message switching as defined by the DOJ (appropriation) Authorization Act, inasmuch as the notification message to be sent to See identical letters dated Mar, 2, 1981, from the FBI Director to the Chairman of the Senate Judiciary Committee and the Chairman of the House Judiciary Subcommittee on Civil and Constitutional Rights.

PAGE 124

116 An Assessment of Alternatives for a National Computerized Criminal History System Figure IO. Interstate Identification Index Pilot Test as Conducted in 1981 ( n Requesting CCH file a Index b aF or mu It I. State and Federal offender records b F or ~lng[e. state offender records located In Slate rePOslf Orles cV(a telephone mail or teletype, e g ,NLETS SOURCE Office of Technology Assessment Florida (and, by inference, to the other States in the III Phase I test) will be created from the index record and transmitted in a set format. The inquiry will not be retransmitted or switched to Florida. As defined in the act, message switching is the technique of receiving a message, storing it in a computer until the proper outgoing line is available, and then retransmitting, with no direct connection between the incoming and outgoing lines. 6 Record inquiry NCIC Record provided c lines If a hit on Index, then record requested w Ibid., p. 2. Quoted in FBI, 111 Background and Findings, op. cit., p. 63. The House Rqort 96-628, dated Nov. 16, 1979, emphasizes that the conferees definition of message switching is that used by the Office of Technology Assessment. During the 3-month III pilot test, about 973,000 CCH inquiries were received by NCIC. Of that total, 11,415 (or about 1.2 percent) resulted in a match (or hit) between the subject of the inquiry and a Florida single-State record. 7 For these matches, NCIC notified the inquiring agency of a hit and forwarded an AIR message s to Florida over the NCIC communication lines. Florida then provided a summary CCH record to the requesting agency via NLETS. A full record was provided by mail if requested. During the pilot test, agencies Ibid., p. 144. Ibid., p. 143.

PAGE 125

.. Ch. 10Major Structural Alternatives for a National Computerized Criminal History System l 11 7 from 34 States (including Florida), 3 metropolitan agencies, and several Federal agencies (e.g., U.S. Customs, U.S. Postal Service) made record requests that resulted in hits on Florida single-State records. 9 During the 2-month III Phase 1 test, 10,934 CCH inquiries resulted in a hit. Agencies from 39 States made requests that resulted in hits on single-State records of one or more of the six States with entries in the index. l0 Examination of sample message formats indicates that the AIR notification message does include some information that is different from that contained in the initial inquiry message. For example, AIR may include an out-of-State identification number that was not known by the inquiring agency but that was added by NCIC based on the index record match. However, some information is likely to be the same in both the inquiry and AIR, such as the name of the subject and the identifying number of the inquiring agency .11 Thus, AIR does involve the switching of some key information from the inquiring agency to the State of record. FBI officials have argued that only housekeeping data, such as originating case identifying number, purpose code, and mailing address, are taken from the inquiry message and included in the AIR message. According to the FBI, technically all key information, such as the name of the subject and other descriptive information, is taken from the 111 record, even though such information may be a part of the inquiry message. In the pilot test, the inquiries were switched from criminal justice agencies in 33 States, the District of Columbia, and the Federal Government through NCIC to the State of Florida. During the 111 Phase 1 test, inquiries were switched from inquiring agencies to any of the six States participating. However, since only some of the information in the inquiry message is actually switched, and since the record itself is transmitted via NLETS or the mail, AIR is properly considered a form of partial message switching. AIR is clearly a change from the 1980 FBI III proposal that involved no message switching. The justification for AIR advanced by the FBI and the III Subcommittee of the NCIC Advisory Policy Board included the ability of NCIC to check inquiries to make sure that the requesting agency identifier, control terminal line, and purpose code were properly authorized. Also, NCIC already had developed standard inquiry message formats that could be used during the pilot test. Finally, the use of AIR would eliminate the need for inquiring agencies to send two messagesone to NCIC/ CCH and, if a hit occurs, a second message via NLETS to Florida to request the record. The NLETS Board of Directors had initially declined to support AIR, in part because of anticipated concerns over message switching. At that point the NCIC staff recommended against pursuing AIR further. 12 However, on the recommendation of its III Subcommittee, the NCIC Advisory Policy Board endorsed AIR in December 1980, but only if it was (also) endorsed by the NLETS Board of Directors and subsequently presented to the appropriate congressional representatives for their understanding and concurrence. In the event AIR proved to be unacceptable, the NCIC Board endorsed a pure pointer index with no NCIC message switching as a fullback. The NLETS Board then reversed itself and endorsed AIR on December 17, 1980. On March 2, 1981, the FBI Director advised the House and Senate Judiciary Committees by letter of the plan to use AIR. I bid., pp. 158-159. ONCIC Advisory Policy Board, Interstate Identification lrJdex Phase I Test: Report of III Evaluation Cbmmittee, June 1982, pp. 10, 12, 13. ] Based on comparison of sample messages in Ibid., pp. 120-124. FBI, minutes of the Dec. 10-11, 1980, meeting of the NCIC Advisory Policy Board, pp. 48-50. At its Oct. 22-23, 1980 meeting, the NLETS Board voted 4 to 3 against supporting AIR. Ibid., p. 50. ,f I

PAGE 126

118 An Assessment of Alternatives for a National Computerized Criminal History System .. --. . National Inde x In its 1980 proposal, the FBI indicated that the pilot project, if successful, might be extended to include the return to the States of multi-State as well as single-State offender records. The current III long-range development plan calls for the NCIC/CCH file to include only the records of Federal offenders, plus a national index 4 of all single-State and multiState offender records. A National Fingerprint File (NFF) is considered to be an integral part of III. The NFF would contain no arrest or disposition data, would perform the technical fingerprint search and assign FBI identification numbers, and would be predicated on single-source submission policies. States or agencies making an inquiry would receive either a hit or a no-hit response. If a hit response were received (indicating that the subject individual is listed in the national index as having a record in one or more State repositories), the requesting agency would also be provided with the name of the State (or States) holding the records. The requesting agency would contact the State repositories directly to obtain the records, as shown in figure 11. In this CCH alternative, NCIC/CCH message switching would not be required. This alternative is sometimes referred to as a pure pointer index because all the index does is point to the location of a record. The national index also could be implemented with message switching. Nationwide implementation of phases 2 and 3 of the III development plan would be the equivalent of a national index with partial message switching. With partial switching, such as the AIR technique used in the III pilot test, NCIC/ CCH would route inquiries through the NCIC computer to the States holding records. These Also known as a national criminal identification name file, and more recently as an Interstate Identification Index. See SEARCH Group, Inc., A Framework for Constructing an lmproved National Criminal History System (Sacramento, Calif.: SEARCH Group, Inc., April 1978), p.6, and Essential Elements and Actions for Implementing A Nationwide Criminal History Program (Sacramento, Calif.: SEARCH Group, Inc., February 1979), p.7. States would provide records directly to the requesting State or agency. With complete message switching, both inquiries and records would be routed through NCIC/CCH. In effect, NCIC/CCH would query all States for which a hit is indicated (a record is held on the subject individual), collect all the records from the various States into a consolidated record, and provide it to the requesting agency. In a 1979 survey, OTA found that State repository personnel favored the national index CCH alternative (the single-State/multi-State alternative was a distant second choice). s A national index would avoid the expense of duplicating records at both State and Federal levels. In addition, a national index would preserve State control over in-State records. A national index might have a somewhat slower total response time than the singleState/multi-State or national repository alternatives because in the case of a hit either the uiring State or the index would subsequentinq ly have to contact the State(s) of record. There is a legitimate question as to how fast States would respond to out-of-State requests. Some States are not computerized (23 do not have CCH files, and 16 of these do not have even an automated name index); a few have no immediate plans to computerize. Even if computerized, out-of-State requests might be given low priority. This has not been the case to date with either the III pilot test or Phase 1 development. Florida (the pilot test State) and five other States participating in Phase 1 are among the more advanced computerized States, and have given both high priority and quick turnaround to out-of-State requests for records. l6 of 42 States responding, State repository personnel in 24 States favored the national index, 11 favored the single-State multi-State, 1 each favored the national repository, a decentralized system, and a regional system, and 4 indicated no preference. Steven W. Hays, et al., An Assessment of the Uses of Information in in National Crime Information Center and Computerized Criminal History Program, Bureau of Governmental Research and Service, University of South Carolina, October 1979, Sec. IV, pp. 178-179.

PAGE 127

Ch. 10Major Structural Alternatives for a National Computerized Criminal History System 119 Figure 11. National Index CCH Alternative With No Message Switching n Requesting agency t CCH Record inquiry file a l Computed re t I National Index b r Iecord ques t c Record provided c / + / Originating \ u agency a For Federal offender records b For single State and multi State offender records c Via telephone mall or teletype e g NLETS d could use NCIC or AIDS computer. e could use NCIC communication Iines SOURCE Office of Technology Assessment Regional and Ask-the-Network System s Regional CCH System s In the 1979 OTA survey of State repositories, 16 several States listed regional CCH systems as a secondary preference, but most contended that regional systems would be infeasible or impractical. The States that believed regional systems to be feasible viewed them primarily as a better-than-nothing alter+ native. Few appeared willing to endorse regional systems enthusiastically. However, several noted that regional systems inevitably would develop if Congress does not decide on some national CCH alternative. Other States perceived regional systems as a possible transitional strategy until a more long-term solution is found. Except for attitudinal data, few definitive 16 Ibid., Sec. IV, pp. 176-177. indicators were found to support the feasibili-

PAGE 128

120 An Assessment of Alternatives for a National Computerized Criminal History System ty of regional systems. Conversely, officials in seven States regarded regional criminal history files as feasible for most informational needs. NLETS traffic logs indicate that criminal history traffic between States does not conform to regional patterns. For example, Florida communicates most frequently with Midwestern and Western States. 17 These trends were confirmed by the results of the III pilot test. During the test period, excluding Florida intrastate traffic, almost threequarters of the hits were on inquiries from the Midwest and West. 18 Regional CCH systems are feasible from a technical perspective; however, the potential problems could be significant. Without formal coordination, jurisdictions on the borders between regions or adjacent to several regions might find themselves participating in several regional systems at a significantly higher total cost. If common standards for message formats and the like were lacking, automated exchange of criminal history records could prove to be difficult, if not impossible. The magnitude of the problem is illustrated by the decision in the single-State/multi-State pilot project to have all message inquiries pass through the FBI to ensure a common message format, based in part on the conclusion that it would be too difficult (and costly) to accommodate widely varying message formats. Ask-the-Network Syste m Many of these technical problems could be overcome if all States agreed and were equipped to participate in a completely decentralized ask-the-network national CCH system. States would retain single-state offender records and the FBI would retain Federal offender records, as in the national index alternative. In the ask-the-network version, however, there would be no index. Instead, each State could poll any or all of the other 49 States plus the FBI when seeking CCH information. Ibid., p. 181. ISFBI, ~~~ Bac&ound and Finciings, op. cit., pp. 158-159. There are several technical options for an ask-the-network design. One option would tie all 50 States and the FBI together on what would be a computerized party line. Messages could be sent to everyone on the line. Another technical option would involve the use of a national switcher or several interconnected regional switchers, as illustrated in figure 12. For example, State-to-State message traffic could be routed through the upgraded NLETS switcher in Phoenix, Ariz., and State-to-FBI traffic could be routed to Washington, D. C., through the NCIC network. Alternatively, a switcher located in Phoenix (or elsewhere) could handle both State-to-State and State-toFBI traffic. For all practical purposes, NLETS presently offers this capability. As of September 1981, NLETS was operating at about 7 to 10 percent of capacity and thus could handle a substantial increase in CCH-related message traffic. The NLETS response time (to switch a message from the sending State to the receiving State) is now less than 5 seconds. Also, NLETS users can send messages to any desired combination of States; for example, all Western States, all States contiguous with Colorado, or all 50 States plus the FBI (an all points bulletin). The potential for use of NLETS must be tempered by experience with ROIR messages, where inquiring States send messages to all other States (or a large number). Only those States with a record on the subject individual need reply. This is similar to an ask-thenetwork capability. However, NLETS found that many States began to ignore the messages, especially where the probability of a hit was very low. In many cases the effort (and cost) of checking out all inquiries apparently does not justify the results. This is particularly true for smaller States, those that are not yet computerized, and those that, while computerized, still maintain a significant number of manual records (and consequently have to check both manual and computerized files).

PAGE 129

Ch. 10Major Structural Alternatives for a National Computerized Criminal History System l 121 Figure 12. Decentralized Ask-the-Network CCH Alternative With a National Switcher v CCH NCIC Federal computer offender L If a hit, State B If a hit, NCIC provides record provides record National switcher computer State C l Record inquiry provides record from State A [ Records from NCIC, States B and C SOURCE: Off!ce of Technology Assessment A major problem is that, based on 1979 OTA record quality research, a high percentage (about 75 percent for Ident)* of multiState offenders had arrests in at least one noncontiguous State; and about 43 percent of multi-State offenders had arrests in three or more States.** Thus, it appears that in an askthe-network system, all States and the FBI would have to be polled every time in order to make sure arrests were not missed. However, the inquiry-to-hit ratio would then be very low. As noted above, under similar cir *Of the 168 Ident records with verifiable arrest events (See ch. 8), 51 records showed arrests in multiple States, and 38 of the 51 records showed arrests in at least one noncontiguous State. **Of the 51 Ident multi-state offender records, 22 showed arrests in 3 or more States. -n State CCH f i Ie (B) -5 State CCH fil e (c) cumstances NLETS found that manv States began to ignore the inquiries. Also, the FBI and various State criminal justice officials believe that an ask-the-network approach would not be cost effective, due to the increased communications and processing requirements, and would be harder to secure against unauthorized access. In addition, the FBI has pointed out that the preparation and mailing of fingerprint cards to all 50 States would be costly and time-consuming and that, furthermore, several State identification bureaus do not have the capability to conduct fingerprint searches. Nonetheless, ask-thenetwork systems are used successfully in the defense intelligence community and in the private sector, and their potential use in a national CCH system is an area of possible further research.

PAGE 130

Chapter 1 1 Possible Impacts of a National CCH System on the Criminal Justice Process

PAGE 131

Contents Chapter Summary. . . . . . . . . . . . Criminal Justice Process. . . . . . . . . . Impact on the Criminal Justice Process . . . . . . . Police Use..... . . . . . .....4... . . . Prosecutorial Use.. . . . . . . . . . . . Judicial Use . . . . . . . . . . . . Defense Use . . . . . . . . . . . . Probation Use . . . . . . . . . . . . Correctional and Parole Use. . . . . . . . . . Page 125 125 127 128 130 131 132 133 133

PAGE 132

Chapter 11 Possible Impacts of a National CCH System on the Criminal Justice Process Chapter Summar y During the 12-year debate over a national computerized criminal history (CCH) system, much of the attention has focused on the possible impacts of a national system in five key areas: criminal justice process, employment and licensing decisions, minority groups, federalism, and monitoring or surveillance potential. These areas are discussed in general terms in this chapter and the next. Criminal Justice Proces s The primary purpose of a national CCH system would be to improve the functioning of the criminal justice process. There is no question that criminal history information is used throughout the criminal justice process. The impacts of a national CCH system are more difficult to assess, due in part to the absence of generally accepted measures of effectiveness. Very few criminal justice agencies systematically keep track of how CCH information actually contributes to arrests, property recoveries, charging decisions, successful investigations, and the like. Nonetheless, many Federal, State, and local law enforcement and criminal history record repository officials believe that a national CCH system would make a substantial contribution. However, some local and State criminal justice officials (especially district attorneys, judicial officials, and public defenders) believe that, to be useful, a national CCH system would have to be able to provide information that is more accurate, complete, and timely than is generally available from existing Federal and State criminal history record systems. Police Use: When the police are investigating a reported crime, they sometimes use criminal history records to search for characteristics of past offenders that might connect them to the present crime. When the police are patrolling, looking for suspicious circumstances or individuals, they frequently use criminal justice information in deciding how to handle situations that arise; e.g., whether to interrogate, detain, issue a summons, or make an arrest. However, such information comes primarily from hot files (wanted persons, stolen property) rather than criminal history files. Nonetheless, in the 1979 OTA 50-State survey, 37 States indicated that on-duty law enforcement officers can gain access to criminal history information in both State and local files through local police patrol and inquiry systems. Since patrol decisions often must be made quickly, a national CCH system could make criminal history records more readily available, thus increasing their use. After an arrest, police make or participate in decisions about whether to release or how long to hold the suspect, whether to fingerprint, and the level of charges to be placed. Each of the decisions clearly affects the creation of a criminal history record, and conversely, criminal history records (and thus a national CCH system) may potentially influence these decisions. Prosecutorial Use: District attorneys use criminal history information in arraignment and bail hearings, plea bargaining, formal 125

PAGE 133

126 An Assessment of Alternatives for a National Computerized Criminal History System charging, trial, and sentencing, as well as for special pretrial release and career crime programs. The frequent lack of court disposition information limits the usefulness of State and Federal criminal history records to district attorneys in cases where subjects have records from outside of the local jurisdiction. In many jurisdictions, judges will not consider an arrestonly record as indicative of criminal propensity, or as a factor in considering whether or not a subject might jump bail. Speedy arraignment rules adopted in most States point to the need for rapid, accurate, certifiable rap sheet or criminal history information. The impact of a national CCH system could be particularly significant in pretrial release and bail decisions. If accurate and complete, such CCH records could help prosecutors and judges better balance the need to protect the public from harm by defendants out on bail, versus the need to minimize the detention of defendants on charges for which they have not been tried and convicted under due process of the law. Such a system might also promote the more consistent use of CCH records in charging decisions, and allow prosecutors to direct police resources to the need for additional investigation in cases involving repeat serious or violent offenders. Judicial Use: Criminal court judges use criminal history information in bail hearings, trial proceedings, and sentencing. Research suggests that a national CCH system probably would have limited impact on judicial sentencing behavior, since other factors (circumstances of the crime, police behavior, community expectations) play such a large role, but could significantly enhance the quality of pretrial decisions. However, in States with mandatory sentencing guidelines for offenders with prior convictions, information from a national CCH system could significantly affect even sentencing decisions. Public Defender and Defense Attorney Use: Public defenders and criminal defense attorneys use criminal history information to support the credibility of their clients and witnesses, to interpret for the court their clients circumstances where the client has a past record, and to challenge the credibility of prosecution witnesses. Public defenders are critical of existing Federal, State, and even local CCH systems because of serious record quality problems, their exclusion from equal access to these systems, and what they believe to be the discriminatory impacts of these systems, primarily on minority groups. Public defenders give limited support to a national CCH system if it provides equal access to defense interests, is accurate and up to date, and is purged regularly. Probation Use: The most common use of criminal history information by probation officers is in the preparation of presentence investigation reports. Judges use these to determine sentences suited to offenders, and they are subsequently used by courts and the corrections departments in assigning offenders to appropriate institutions. A national CCH system would appear to be helpful in preparing presentence reports (and in pretrial services) if it were based on accurate and complete records that could be obtained quickly and easily. Correctional and Parole Use: Correctional officials have noted that criminal history is an important input to decisions on initial levels of supervision and security. To the extent that correctional decisions rely on presentence reports, a national CCH system could make a difference if, as a result, these reports were more accurate and complete. With respect to parole decisions, the nature and seriousness of any prior record can be important. However, the impact of a national CCH system would most likely be limited, since an inmates behavior in prison and the seriousness of the current offense can carry as much or more weight than prior criminal history.

PAGE 134

Ch. 11Possible Impacts of a National CCH System on the Criminal Justice Process l 127 Impact on the Criminal Justice Process The impacts of a national CCH system are difficult to assess. A 1980 survey by the Missouri State Highway Patron and an August 1981 survey by the National Crime Information Center (NCIC) 2 found that historically it has been difficult to describe or document the effectiveness of computerized real-time criminal justice information systems. While perhaps half of the agencies responding indicated that they measure the number of hits on a system (i.e., matches between an inquiry and a record on file), very few keep track of how hit information contributes to a law enforcement action (e.g., arrest, summons, recovery, apprehension, successful investigation). None of the agencies responding keeps track systematically of how such information contributes to other criminal justice actions such as prosecutions, setting of bail, sentencing, and the like. The NCIC Advisory Policy Board has established a subcommittee to study and develop recommendations on how best to measure the effectiveness of NCIC. These surveys and others conducted by OTA, the Department of Justice (DOJ), 3 and the House Judiciary Subcommittee on Civil and Constitutional Rights 4 have found strong support for computerized hot files among Federal, State, and local law enforcement officials. Of the dozens of State and local law enforcement officials surveyed, not one seriously 1 Robert J. Bradley, State-Level Criminal Justice Network Systems Effectiveness Survey, prepared by the Information Systems Division, Missouri State Highway Patrol, September 1980; summarizes the results of a brief questionnaire sent to 11 States. NCIC Staff Paper on Statistical Measurement of NCIC Benefits prepared for the Dec. 9-10, 1981, meeting of the NCIC Advisory Policy Board, Topic 410; summarizes the results of a questionnaire sent to 97 Federal, State, and local NCIC users. See U.S. Department of Justice, Representative Viewpoints of State Criminal Justice Officials Regarding the Need for a Nationwide Interchange Fac&ty, Mar. 6, 1978, reprinted in U.S. Congress, Office of Technology Assessment, A Preliminary Assessment of the National Crime Information Center and the Computerized Criminal History System, Washington, D. C., December 1978, p. 69. The Subcommittee Chairman sent a brief questionnaire on NCIC to the chiefs of police of 36 metropolitan areas. For a summary of the results, see May 18, 1981, letter to the FBI Director from the Chairman, Subcommittee on Civil and Constitutional Rights, House Committee on the Judiciary. questioned the need for, or benefits of, the NCIC hot files. Indeed, many noted that law enforcement personnel simply could not do without Federal hot files, and that the interstate transportation of stolen properties and interstate mobility of wanted persons necessitates a national system. Likewise, many Federal, State, and local law enforcement and criminal history record repository officials support the concept of a national CCH system, although there is some disagreement over the specifics. However, some local and State criminal justice decisionmakers (especially district attorneys, judicial officials, and public defenders) emphasized that to be useful such a system needs to provide information that is more accurate, complete, and timely than is generally available from existing Federal and State criminal history record systems. Some believe that a national CCH system should be limited to certain kinds of records (e.g., felony convictions) and purged at periodic intervals. On the other hand, some officials, especially in law enforcement, emphasize the importance of CCH records, however incomplete they may be, as a pointer to sources of more complete information. Also, no hit (or no record) information may, in some situations, be just as useful to police as hit (or record) information. A survey conducted in September 1981 by the Florida Department of Law Enforcement, based on 269 responses from 529 criminal justice agencies, found that the Florida State CCH summary record information met the needs of about 85 percent of law enforcement agencies responding, 64 percent of prosecuting attorneys responding, and 49 percent of probation and parole agencies responding. The survey also found that Florida State CCH full record information met the needs of about 91 percent of law enforcement agencies, 96 percent of prosecuting attorneys, and 72 percent of probation and parole agencies responding. 5 See Federal Bureau of Investigation, Interstate Identification Index: Background and Findings for Jul.y-Septemher 1981 Phase I Pilot Project, Dec. 4, 1981, p.89.

PAGE 135

128 An Assessment of Alternatives for a National Computerized Criminal History System The most common intended uses of criminal history records were for criminal investigations and booking and intake by law enforcement agencies, criminal investigations and bail and bond determinations by prosecuting attorneys, and pretrial intervention and presentence investigations by probation and parole agencies. State officials interviewed in the 1978 DOJ survey, which included the State of Florida, also make the point that the equal treatment of offenders is in part dependent upon the equal availability of appropriate and relevant information at all stages of the criminal justice process. Police Us e When the police are investigating a reported crime, they sometimes use criminal history records to search for characteristics of past offenders that might connect them to the present crime. Prior research indicates that relatively few crimes are solved as a result of investigation alone. 8 One study found that the vast majority of case clearances result from patrol capture at the crime scene, complete identification by victims or witnesses, or provision of uniquely linking evidence (such as license plate numbers) by victims and witnesses. A small proportion of case clearances were found to result from matching latent fingerprints to existing fingerprint files, matching offense modus operandi (MO) with those of offenders on file, using information tips, or having victims view mug shots. One area that seems directly relevant to a national CCH system is the matching of a crime MO, physical descriptions, or other evidence with information in existing criminal history files. In the local jurisdictions surveyed by OTA in 1979, police used local files primarily and State files to a lesser extent. In investigating serious and violent crimes, on Ibid., p~87. 7DOJ, Representative Viewpoints, Ibid., p. 69. 8Charles E. Silberman, Criminal Violence, Criminal Justice (New York: Random House, 1979), pp. 217-219. Peter W. Greenwood, et al., The Criminal Investigative Process (Lexington, Mass.: D.C. Heath and Co., 1977), pp. 125 and 135. the other hand, the need for rapid retrieval of criminal history information from State and Federal systems was perceived as more important. When the police are patrolling, looking for suspicious circumstances or individuals, they frequently use criminal justice information in deciding how to handle situations that arise, e.g., whether to interrogate, detain, issue a summons, or make an arrest. The Co remission on Criminal Justice Standards and Goals concisely described the police need for information while on patrol:l 10 In any citizen contact, the officer should know if the citizen is wanted by police, is in possession of any stolen property or is otherwise involved with criminal activity that might indicate the persons present intentions or behavior upon police contact The officer should have knowledge about the contact that serves either to protect the officer or increase his chances of success. This is one area of police operations where computer and communication capabilities are almost essential. To affect patrol decisions response times must be rapid. Radio calls to the local police station or to a central State repository for manual searches of criminal record files frequently would be too slow to be very useful, especially in heavily populated jurisdictions. As a consequence, computerized recordkeeping and communication systems have been used to increase the information available to patrol officers, particularly with respect to wanted persons, stolen vehicle and other stolen property information, and, to a lesser extent, criminal history information. OTA found that police on patrol make relatively little use of criminal history information, but rely quite heavily on hot files. However, in the 1979 OTA 50-State survey, 37 States indicated that on-duty law enforcement officers can gain access to criminal history information in the State as well as the local file through local police patrol and inquiry systems. National Advisory Commission on Criminal Justice Standards and Goals, Criminal Justice System (Washington, D. C.: U.S. Department of Justice, 1973), p. 39.

PAGE 136

Ch. 11Possible Impacts of a National CCH System on the Criminal Justice Process l 129 Relatively little is known about the extent to which police patrol decisions are based on criminal history information when it is available and used. An observational study of arrest practices in three States by the American Bar Foundation found that prior criminal record history did influence police arrest decisions, and noted that courts have held that police indeed may use a past record as one of several factors in making legal arrests. 1] An experimental study of situations where the probable cause to arrest was ambiguous concluded that a prior criminal history record did affect the decision to arrest. ]2 A prior criminal history record is also a factor in police decisions on whether to use an alternative to arrest. Growing concerns about the number of petty cases clogging the courts and the stigma of arrest have stimulated the use of a summons or field citation in place of arrest. Commonly used in traffic offenses and some misdemeanors, the field citation orders a person to appear in court on a given day for violating a statute. A summons serves the same purpose, but is used for all types of charges. For example, in 1977, a summons was issued to about 3 percent of persons charged with murder, 14 percent of persons charged with aggravated assault, 40 percent of persons charged with fraud, and 61 percent of persons charged with violating liquor laws. 13 A persons prior arrests and convictions time considered in deciding whether to make an arrest or issue a summons. Some police departments require officers to use a point system that takes into account information concerning identification, employment, residence, and family ties, as well as criminal history. Since the decision to summon or arrest often must Wayne R. LaFave, Arres& The Decision to Take A Suspect IrIto Custody (Boston, Mass.: Little, Brown and Co., 1965), pp. 150-354. R. C. Smith, et al., Background Information: Does It Affect the Misdemeanor Arrest? Journal of Police Science and Administration, vol. 4, March 1976, pp. 111-113. Lynne Eickholt Cooper, et al., An Assessment of the social Impacts of the National Crime Information Center and Cbnputerizeci Criminal History Program, Bureau of Governmental Research and Service, University of South Carolina, October 1979, sec. II, pp. 102-103. be made quickly, a national CCH system could make criminal history records more readily available, thus increasing their use. After an arrest, police make or participate in decisions about whether to release or how long to hold the suspect, whether to fingerprint, and the level of charges to be placed. Each of these decisions clearly affects the creation of a criminal history record, and conversely, criminal history records may potentially influence the decisions. Police do not have exclusive control over these decisions; prosecutors may also be involved. The division of authority between police and prosecutors varies with each jurisdiction. For example, in some areas charges are placed against a suspect by the police that are not reviewed by the prosecutor until the preliminary hearing. In other areas, police book suspects on suspicion, and prosecutors place formal charges. Police may decide to release an individual for a variety of reasonsfurther investigation reveals evidence is lacking, witnesses refuse to cooperate, the offense was minor, or the arrest was made for reasons other than prosecution (e.g., to protect the person from harm). Plausibly, knowledge of criminal history could influence a police officers decision to release a suspect, especially if the offense is minor. On the other hand, a criminal history, especially one related to the current offense, may give police additional evidence on which to charge a suspect. The level or seriousness of charges may also be affected by criminal history records. In some States, persons arrested for a crime of which they were previously convicted may face a more serious charge that carries a stiffer penalty. In order to provide adequate evidence to charge, police may decide to hold arrested individuals if they are suspected of offenses requiring further investigation. A criminal history is likely to affect police suspicions, and therefore will influence the decision to hold an individual and whether to hold without booking or to proceed immediately to booking and fingerprinting, even if the arrestee is subsequently released.

PAGE 137

130 l An Assessment of Alternatives for a National Computerized Criminal History System Finally, criminal history records are used in career criminal programs to identify arrestees whose cases will receive priority for investigation and case preparation by police and prosecutors. Police may make an early determination that an arrestee qualifies as a career criminal (dangerous repeat offender) and intensify investigatory efforts to provide adequate evidence for the highest possible charge. Prosecutorial Us e Prosecuting attorneys use criminal history information in arraignment and bail hearings, for plea bargaining, formal charging, trial, and sentencing, as well as for special pretrial release and career crime programs. Given the predominantly local nature of crime in the urban areas investigated by OTA, prosecuting attorneys generally opted for the development of comprehensive local and State criminal history systems that identify a large percentage of local criminals who have records. However, prosecuting attorneys also supported a national CCH system if it would provide information that is more complete, accurate, and timely than is presently available. Aside from problems of verifying out-ofState records, prosecuting attorneys also pointed out that in many jurisdictions judges do not consider that a record of arrest necessarily indicates either a criminal propensity or a tendency to jump bail. Here, prosecuting attorneys stressed that the frequent lack of court disposition information on State and Federal criminal history records was a major impediment to using them in cases where subjects have records from outside of the local jurisdiction where they are being processed. Record quality and the ease with which criminal history records can be interpreted are even more important in specialized programs that are developing in local areas, such as prior felon, career crime, and violent felon programs. These programs assign special prosecutors to subjects who have prior felony convictions. .. ..--. 14 Ibid., sec. 11, pp. 111-150. The lack of conviction information on many criminal history records disseminated by existing Federal systems, and the use of valuable time and resources to verify out-of-State felony arrests, can limit prosecutorial use of outof-State information in these specialized programs. Speedy arraignment rules adopted in most States point to the need for rapid, accurate, certifiable rap sheet or criminal history information. As a matter of law, the use of arrestonly information in bail decisions under some circumstances has been found unconstitutional by the courts. In the case of Tatum v. Rogers, the court ruled in 1979 that:15 Plaintiffs are clearly and systematically being deprived of due process in violation of the Fourteenth Amendment ..., and of the right to effective assistance of counsel as guaranteed by the Sixth Amendment, whenever rap sheets containing erroneous, ambiguous, or incomplete data with respect to prior arrests and dispositions are submitted to courts at arraignment sessions for use in connection with bail determination. The Eighth Amendment right to reasonable bail is also thus denied. It is important to note that the court found a violation of constitutional rights only when arrest information without dispositions was used, when disposition information was otherwise available, and when the incomplete information was used in setting bail. 16 Thus, a national CCH system could have a significant impact on pretrial release and bail decisions by district attorneys and judges.* A 1978 study in the District of Columbia confirmed that, with all other factors held constant, release conditions were made more stringent for arrestees whose criminal histories were more extensive. 17 A 1979 study con15TatUm ~gers, 75 CiV. 2s2 (U.S. District court, South District of New York), Findings of Fact and Conclusions of Law, p. 20. e Ibid., p. 24. *In most cases, judges make the actual pretrial release and bail decisions, but prosecutors play a significant role when making recommendations. Jeffrey A. Roth and Paul B. Wice, Pretrial Release and ll4isconduct irI the District of Columbia Executive Summary (Washington, D. C.: Institute of Law and Social Research, 1978).

PAGE 138

Ch. 11Possible Impacts of a National CCH System on the Criminal Justice Process l 131 ducted for the Tatm v. Rogers case in New York showed similar results. This research also highlighted the frequent lack of disposition data in early stages of the criminal justice process and the reliance of prosecutors (and judges) on incomplete rap sheets listing only arrests. 18 While the value of conviction data in predicting pretrial rearrest (for another crime committed while out on bail or on personal recognizance) has not been established empirically, l9 the majority of States have legislation or rules requiring judges to consider prior convictions in determining pretrial release conditions. Thus, a national CCH system, if accurate and complete, presumably could assist prosecutors and judges in better balancing the need to protect the public from harm by defendants while out on bail, versus the need to minimize the detention of defendants on charges for which they have not been tried and convicted under due process law. With respect to charging and plea bargaining, the impact of a national CCH system is less clear. Evidence suggests that the criminal history record of an arrestee can affect the prosecutors decisions to bring or drop charges, the level and number of charges, and whether to negotiate at trial for lower charges through plea bargaining. To some extent, prosecutors can delay a decision to dismiss a case or to place final charges until they have received criminal history information from existing files, whether manual or computerized. The time factor is not as critical as it is in pretrial release decisions. However, some anecdotal evidence suggests that prosecutors, under the crush of heavy caseloads, may make quick decisions to dismiss or reduce charges without considering a ]~~a~U~ ~ ger s, Op. cit., Affidavit of Richard Faust. Roth, op. cit., p. x; Wayne Thomas, Bail Reform in America (Berkeley, Calif.: University of California Press, 1976), pp. 234-240. Vera Institute of Justice, Felony Arrests: Their Prosecution and Disposition in New York Citys C@rts (New York: Vera Institute of Justice, 1977); Arthur Rosett and Donald R. Cressey, Justice by C.lmsent: Plea Bargains in the American Courthouse (Philadelphia, Pa.: J.B. Lippincott and Co., 1976), ch. 5. defendants criminal history. To the degree that this occurs, the rapid response time of a national CCH system may promote the more consistent use of such records in charging decisions, thereby allowing prosecutors to direct police resources to the need for additional investigation in cases involving repeat serious or violent offenders. Judicial Us e Criminal court judges use criminal history information in bail hearings, trial proceedings, and sentencing. Judges do not perceive their role as simply to judge the accused, but also to understand and consider the behavior of the police and the prosecution in order to understand the circumstances of the accused. Criminal history information is one input to judicial decisions. Some judicial researchers believe that the quality of justice administered in criminal courts could be significantly enhanced if relevant and reliable criminal history information were to be available to key decisionmakers at the courthouseparticularly the judge and the district attorneywithin a short time after arrest. Critical decisions about charge and pretrial custody are typically made very shortly after arrest, and hard information about an arrestees prior record can be of great value. However, these researchers also point out that such information is seldom available from existing CCH systems at the present time and that the lack of reliable information about dispositions is a major problem in many criminal history records systems. 2 With respect to sentencing, there are several realities that limit the role of criminal history information. For example: 23 If there is no agreement on what causes crime, there can be no agreement on how zlHer~fi s. Miller, et al., Plea Bargaining iII the United S~teS (Washington, D. C.: U.S. Government Printing Office, 1978), pp. 72-73. Sept. 15, 1981, letter tQ OTA from Barry Mahoney, Research Director, Institute for Court Management. zgAlexader B. Sfith and Harriet Pollack, Crimina] Justice: An Overview (2d cd.) (New York: HoIt, Rinehart, & Winston, 1980).

PAGE 139

132 l An Assessment of Alternatives for a National Computerized Criminal History System criminals should be handled. If criminals are sick they should be treated, but if they are bad they should be punished. Judges, juries, probation, parole, and corrections officials do not handle all cases in a uniform manner but vacillate, sometimes unpredictably, between these two polar positions, One cause of sentence disparity, thus, is the lack of a consistent theory of crime causation and punishment. In the absence of a consistent theory of punishment, other factors play a large role in judicial sentencing, such as the expectations of a particular community or the personality and philosophy of the judge. Sentencing decisions are also affected by the overloaded conditions that exist in many urban courts, prisons, and other detention facilities. Nevertheless, criminal history information clearly has an influence on sentencing. For example, a study of sentencing practices in the District of Columbia found that decisions to incarcerate were influenced most strongly by the prior criminal history of the offender, the severity of the current offense, and the philosophy of the judge. Sentence length was influenced mainly by the seriousness of the offense, community ties, type of plea, and prior convictions (to a lesser extent) .24 Many States now have mandatory sentencing guidelines (especially for offenders with prior records). Defense Us e Public defenders and criminal defense attorneys use criminal history information, where available, to support the credibility of their clients and witnesses, to interpret any past record of their clients to the court, and to challenge the credibility of the prosecution witnesses Public defenders are critical of existing Federal, State, and even local CCH systems because of serious record quality problems, the exclusion of defense interests from equal ac24 Terence Durgworth, An Empirical Assessment of Sentencing Practices in the Superior Court of the District of Columbia (Washington, D. C.: Institute for Law and Social Research, 1978), pp. VI-7 to VI-25. cess to these systems, and what they believe are discriminatory impacts of these systems, particularly on minority groups. Public defenders are concerned that arrest-only information is used by prosecutors to characterize their clients unfairly, and that such information is frequently included in the presentence investigation report obtained from a probation department, which also influences the courts behavior. In addition, public defenders are unhappy that they are excluded from access to most local, State, and Federal criminal history systems. Information from these systems can be used by the prosecution to attack the credibility of defense witnesses without allowing public defenders the same opportunity with respect to prosecution witnesses. Public defenders believe that the creation of arrest records discriminates against minority groups in their localities. Defenders believe that local police are more likely to stop, search, detain, and arrest members of minority groups. Defense interests believe they can exercise a greater influence over the use of information in local CCH systems through both legal and political means, and therefore are more supportive of local systems than of State and Federal systems. In general, public defenders give limited support to a national CCH system that provides equal access to defense interests, is accurate and up to date, and is purged regularly. In comments to OTA, the National Legal Aid and Defender Association (NLADA) noted that one of the major problems public defenders have with a CCH system is its lack of ability to quickly and accurately expunge incorrect information. Further, there need to be legislative guidelines to insure such modifications can be made. These guidelines would also be useful in insuring public defenders equal access to this information. The only way to insure such access is to require prosecutorial employees to turn over, within 48 hours, all CCH information (or lack thereof) on the accused. Defenders also need, according to the NLADA, a means to get into the system to check on records of their own and government

PAGE 140

Ch. 11Possible Impacts of a National CCH System on the Criminal Justice Process l 133 -. witnesses without the government bein g aware of these checks. It is unrealistic in most jurisdictions to merely suggest that defenders be given access. Legislation must mandat e such access. 5 Probation Us e Probation officers use criminal history information in specialized pretrial services programs, such as pretrial release programs However, the most common use of such information is in the preparation of presentence investigation reports. These reports are mad e available to the court for the purpose of evaluating the character of persons brought before it. They are used by judges in arriving at a sentence suited to an offender, and are subsequently used by courts and by corrections departments in assigning offenders to appropriate institutions Despite the intention of State legislation arrest-only information may influence the presentence investigation report and judicial sentencing. This influence stems from the concern of the probation profession to include all the information that may be helpful in explaining the criminal conduct of an offender. For example, a presentence report may cover significant police contacts which could include arrests without dispositions Given the incomplete disposition reporting in State and Federal record systems, probation agencies must spend considerable resources to verify records from such systems. There is a tendency to ignore the verification of Federal records because of the difficulty involved. It appears that a national CCH system could be useful in pretrial services (principally in bail hearings, as noted earlier) where accurate and reliable information is needed within 72 hours of arrest. Time is not as important in the preparation of presentencing reports Here the major problem is incomplete disposition data and the shortage of resources (time and money) necessary for verification. A na25 Mar. 16, 1982, letter to OTA from the Deputy Director, Defender Division, National Legal Aid & Defender Association. tional CCH system would appear to offer an advantage if based on accurate and complete records, and if those records could be obtained quickly and easily. Correctional and Parole Us e The criminal history of an offender is considered significant in determining initial custody rating (level of supervision needed) and institutional placement (e.g., maximum v. medium security). Correctional officials have noted that criminal history information is more important to these decisions than the results of inmate evaluation and testing. Career criminals are more likely to be placed under closer supervision in more secure facilities. However, once in prison the behavior of an inmate largely determines the correctional program, e.g., assignment to educational, work, and rehabilitation activities. Before an offender is imprisoned, criminal justice officials have ample time to retrieve criminal history information. Correctional authorities have long relied on detailed presentence reports, which include criminal history information, in making their decisions. There is no reason to believe that a national CCH system alone would alter this approach. A national CCH system could make a difference if, as a result, presentence reports were more accurate and complete. With respect to parole decisions (to release an inmate from prison, subject to supervision by a parole officer), criminal history is one of many factors considered. One study found that, in the majority of cases decided by the U.S. Parole Commission, parole decisions could be predicted by the seriousness of the current offense, the nature and seriousness of any prior record, and the offenders conduct in prison. 26 Thus, there is reason to believe that a national CCH system would have at least some impact on parole decisions. However, any Leslie T. Wilkins, et al., sentencing Guidelines: Structuring Judicial Discretion, final report (Albany, N. Y.: Criminal Justice Research Center, 1976), pp. 13-19. j -J. :

PAGE 141

134 An Assessment of Alternatives for a National Computerized Criminal History System major changes in the way in which criminal persons with extensive criminal histories history information is used in parole decisions should be allowed parole, and whether the will likely result from resolution of other degree of supervision after release should deissues. These include whether the parole funcpend on their criminal histories. tion should continue at all and, if so, whether

PAGE 142

Chapter 1 2 Other Impact Areas Relevant to a National CCH System: Licensure, Employment and Minority Groups, Federalism, Monitoring or Surveillance Potential, and Constitutional Rights

PAGE 143

Contents Page Chapter Summary. . . . . . . . . . . . 137 Employment and Licensing Decisions . . . . . . . 137 Minority Groups . . . . . . . . . . . 137 Federalism . . . . . . . . . . . . 138 Surveillance Potential. . . . . . . . . . . 138 Constitutional Rights . . . . . . . . . . 138 Impact unemployment and Licensing Decisions. . . . . . 139 Impact on Minority Groups . . . . . . . . . 141 Impact on Federalism . . . . . . . . . . 142 Impact on Monitoring or Surveillance Potential. . . . . . 145 Impact on Constitutional Rights . . . . . . . . 147 TABLE Table No. Page Justice Requests to State Criminal History Repositories 140 30.Noncriminal

PAGE 144

Chapter 12 Other Impact Areas Relevant to a National CCH System : Employment and Licensure, Minority Groups, Federalism Monitoring or Surveillance Potential, and Constitutional Rights Chapter Summar y Employment and Licensin g Decision s Criminal records may be used to screen individuals out of positions where they could pose a threat to other citizens or coworkers or present an excessive risk to the protection of valuable assets. However, limiting job opportunities may hinder the rehabilitation of former offenders who may become dependent on public welfare or return to crime if suitable employment is unavailable. Federal, State, and local governments require criminal history checks or character evaluations (which frequently include record checks) for literally millions of public sector jobs or publicly licensed private sector jobs. The private sector also frequently seeks criminal history information in making employment decisions. A national computerized criminal history (CCH) system might further increase the use of Federal and State criminal history files for noncriminal justice purposes. There is no doubt that the use of criminal history information affects employment and licensing decisions. Even a record of arrest and acquittal will often work to the disadvantage of the applicant. A problem here is that a noncriminal justice decisionmaker is more likely to misinterpret a record, especially when criminal history records contain inaccurate, incomplete, or ambiguous information. Also, except in particular cases such as repeat violent offenders, the ability of criminal history records to predict future employment behavior is a matter of debate. For some occupations, the law is quite clear about what kinds of criminal conduct are disqualifying. However, in most cases substantial discretion is left to licensing boards and employers. In the OTA 50State survey, noncriminal justice use of criminal history records accounted for about onefifth of total use, and several States reported that noncriminal justice use already constituted more than 40 percent of total use. Finally, a national CCH system could involve up to 28 to 30 percent of all persons in the labor force, many with arrest records showing no arrests for serious crime, arrests and convictions for minor crime only, or arrests that were disposed in favor of the arrestee. Minority Group s Some minority groups have a higher probability of police contact and account for a disproportionate percentage of arrest statistics. For example, the percentage of blacks with arrest records has been estimated at 30 percent nationwide and over 50 percent in some cities. In States like California, blacks are more likely 137

PAGE 145

138 l An Assessment of Alternatives for a National Computerized Criminal History System than are whites to be arrested, to have the arrest reported to the State repository, but then to be released without formal charging. When used in employment decisions, for example, arrest-only criminal history information can have a discriminatory effect. Indeed, the courts have found that a policy of refusing employment to blacks with an arrest record without convictions had a racially discriminatory impact because blacks are arrested substantially more frequently than whites in proportion to their numbers (Gregory v. Litton Systems, 1970). In this context, any discriminatory impacts from the use of national CCH information would depend on whether and under what conditions noncriminal justice access is permitted. Federalis m Many of the proposed alternatives for a national CCH system would encounter difficulties resulting from the historic constitutional division of powers and duties in the U.S. Federal system. State governments have basic jurisdiction over law enforcement and criminal justice within their borders. At the same time, the Federal Government has a legitimate role in the enforcement of Federal criminal law and prosecution of Federal offenders, both intrastate and interstate, and in assisting with the apprehension of criminal offenders who cross State and/or national borders. To the extent that crime is perceived as a national problem, the Federal Government has a defined role in providing voluntary support to State and local law enforcement activities. A national CCH system could be used to circumvent State laws, especially with respect to system access. Given the wide variation among State laws and regulations, any national standards included in a CCH system could easily conflict with the standards of at least some States. In addition, with Federal funding for State CCH development now ended, the States and localities would have to bear most of the cost of any national CCH system. Surveillance Potentia l The flagging of criminal recordsboth hot files and criminal history filesis a common surveillance practice and an accepted law enforcement tool. Placing a flagon a file helps law enforcement personnel keep track of the location and activity of a suspect, apprehend wanted persons, and recover stolen property. Concern has focused on the possible use of a CCH system by Federal agenciesand particularly the Federal Bureau of Investigation (FBI)for surveillance of the lawful activities of individual citizens or organizations. The basis for this concern is largely the well-documented tendency of the FBI and other Federal agencies, in the late 1960s and early 1970s, to expand intelligence investigations into the realm of political surveillance. FBI officials have repeatedly stated that they will not permit the National Crime Information Center (NCIC) or the Identification Division (Ident) to be used for such purposes, and that a national CCH system would represent little, if any, danger to law-abiding citizens. Constitutional Right s The enactment of national legislation that includes statutory protections and mandates specific accountability measures (especially outside audit) was found to be very important in protecting constitutional rights across the board. First and fourth amendment rights could be further protected through tight restrictions (or a prohibition) on noncriminal justice access and strong and independent policy control. Mandatory record quality standards, established by statute and backed up by the necessary funding and technical assistance to ensure implementation (and outside audit to ensure compliance), appear to be the most effective mechanism for protecting fifth, sixth, eighth, and 14th amendment rights.

PAGE 146

. Ch. 12Other Impact Areas Relevant to a National CCH System l 139 Impact on Employment an d Licensing Decision s Criminal history information is used in employment and licensing decisions ostensibly to protect the public or the employer from harm. Criminal records may be used to screen individuals out of positions where they might cause harm to other citizens or coworkers or present an excessive risk to the protection of valuable assets (e.g., money, securities, precious jewelry, and other property). However, limiting job opportunities on the basis of a criminal record in effect involves an additional punishment for crime, that is, a civil disability in addition to the punishment administered by the court. This civil disability may in turn hinder the rehabilitation of offenders and prevent them from becoming useful and productive members of society, even if they want to do so and are otherwise capable. Former offenders who cannot find suitable employment may become dependent on public welfare or return to crime. Federal and State legislatures must balance these considerations when requiring criminal history checks or evaluations of good moral character as conditions of employment or licensing. Such requirements may result in employment or a license being denied to individuals with specified criminal offenses. For example, the Federal Government requires a criminal history record check for all new employees, and many States have adopted similar requirements. No recent systematic research has been done on the number of occupations in the United States that require a criminal history background check. A study conducted in 1974 by the American Bar Association identified 1,948 separate statutory provisions that affected the licensing of persons with an arrest or conviction record, averaging 39 provisions per State. 1 James E. Hunt, James E. Bowers, and Neal Miller, Laws, Licenses and the Offenders Right to Work: A Study of State Laws Restricting the Occupational Licensing of Former Offenders (Washington, D. C.: American Bar Association, 1974). At that time, an estimated 7 million people worked in licensed occupations. This is consistent with OTAS findings in selected States. For example, in California in 1979,47 different licensing boards were authorized to use State criminal history record information for screening applicants. As it stands now, dissemination of Federal criminal history records is permitted to officials of any State or local government for purposes of employment and licensing, if authorized by State statute and approved by the Attorney General. Dissemination of State and local criminal history records is governed by a plethora of widely varying State laws, Executive orders, local ordinances, court orders, and judicial rulings.* Private sector use of criminal history records in employment decisions is even more difficult to document. Research conducted in 1976 found that between 40 and 80 percent of private sector employers seek criminal history information, frequently as part of information requested on employment application forms. 2 With the exception of federally insured or chartered banking institutions and the securities industry, Federal law prohibits the dissemination of Federal criminal history records to private employers.** But in a majority of States, private organizations can lawfully obtain conviction information, and frequently arrest information as well, from State criminal history record files. A 1981 SEARCH Group study found that, in most States, local police may lawfully release to private employers whatever arrest or conviction data they choose from local files. *% chs. 6, 7, 8, and 9. Neal Miller, Employment Barriers to the Employment of Persons With Records of Arrest or Convictions, A Review and Anafysis (Washington, D. C.: U.S. Department of Labor, 1979), pp. 20-23. *** chs. 6 and 7. SEARCH Group, Inc., Privacy and the Private Employer, September 1981 draft, p. 33. Ibid., pp. 34-35; see also ch. 7.

PAGE 147

140 l An Assessment of Alternatives for a National Computerized Criminal History System . A potential problem is that a national CCH system might further increase the use of Federal and State criminal history files for noncriminal justice purposes in ways that might be detrimental to former offenders seeking legitimate employment, without necessarily improving the protection of the public and employers. The problem has several dimensions. First, there is no doubt that the use of criminal history information affects employment and licensing decisions. The results of research, case studies of employers, surveys of employer attitudes, as well as the experience of Federal and State parole officers, all suggest that any formal contact between an individual and the criminal justice process will influence the employers decisions on job applicants. A record of arrest and conviction will have the greatest influence, but even a record of arrest and acquittal will frequently work to the disadvantage of the applicant. The problem is that criminal history records are designed for use by those who are familiar with the criminal justice process and who understand the limitations of a record. At best, a criminal history record provides a snapshot or series of snapshots of a persons contact with the criminal justice process at various points in time. Much of the contextual and background information necessary to properly interpret the record is not included. A record is more likely to be misinterpreted when used by someone outside the criminal justice system, particularly when criminal history records contain inaccurate, incomplete, or ambiguous information. For some occupations, the law is quite clear about what kinds of criminal conduct are disqualifying. In most cases, however, substantial discretion is left to licensing boards and public employers (as well as to private employers) in weighing the applicant criminal record along with all other factors. See Lynne Eickholt Cooper, et al., An Assessment of the Social Impacts of the National Crime information Center and Computerized Criminal History Program, Bureau of Governmental Research and Service, University of South Carolina, October 1979, sec. III, pp. 213-268. Second, little concrete evidence exists to support the thesis that criminal history records have predictive value with respect to employment (i.e., can accurately predict future employment behavior), except in particular cases such as repeat violent offenders. G Other factors such as education, prior work experience, length of time in the community, and personal references may be more predictive. Thus, there is the added risk that individuals with criminal records will be denied employment solely because they have a record, not because of a determination, based on all of the facts available, that they represent an unacceptable risk to the prospective employer. On the other hand, the high recidivism rates suggest that once a person is arrested or convicted, he or she is much more likely to be convicted of a subsequent crime within a few years than those without a prior criminal record. Whether or not this is relevant to or predictive of employment behavior is a matter of debate. Third, despite the limited ability of criminal history records alone to predict future employment behavior, noncriminal justice use has already reached significant levels. In the OTA 50-State survey, noncriminal justice use of criminal history records accounted for about one-fifth of total use, and several States reported that such use already constitutes more than 40 percent of total use, as shown in table 30. Given the politics of State legislatures, in. Ibid. Table 30.Noncriminal Justice Requests to State Criminal History Repositories (as a percent of total requests) Number of Percent States Year Overall average . . . 19.9/0 37 1979 18.70/o 37 1982 17.80/o 45 1982 Percent of noncriminal Number of States justice requests 1979 1982 1982 Distribution of percent noncriminal justice requests 09.90/0 . . . . . . 15 16 18 10-19.90/0 . . . . . . 11 10 13 20-39.90/o . . . . . . 4 4 7 40+ % . . . . . . . 7 7 7 37 45 SOURCE: Office of Technology Assessment 50-State Survey and 1982 followup

PAGE 148

Ch. 12Other Impact Areas Relevant to a National CCH System l 141 terest groups can exert strong leverage to gain access for noncriminal justice purposes. At least 14 States have recently enacted (since 1979) or have pending State legislation or regulations that further broaden noncriminal justice access. Some States are now charging fees (e.g., $6 to $14 per record in New York, about $7 in California, $5 in West Virginia and Nebraska, $3 in Maine, and $2 in Florida) which may serve to restrain noncriminal justice use, although this has not been the case in States like Florida. Between 1979 and 1982, 10 States reported an increase in noncriminal justice use, 5 States reported a decrease, and 22 States indicated that such use remained approximately the same (as a percent of total use). Fourth, a national CCH system could involve a sizable proportion (perhaps 25 to 30 percent) of all persons in the overall labor force. After a careful review of existing research, OTA estimated that as of 1979 about 36 million living citizens had criminal history records held by Federal, State, and/or local repositories. 7 Of these, OTA estimated that about 26 million persons were in the labor force (representing, conservatively, 28 to 30 percent of the total labor force), and thus were potentially exposed to employment disqualifications because of an arrest record. Of the 36 million, OTA estimated that 35 percent had no arrests for serious crime and one arrest for a minor crime, and that 24 percent had more than one minor arrest but no major arrests. The remaining 41 percent (15 million persons) had at least one arrest for a serious crimes See Cooper, et al., op. cit., sec. I, pp. 51-84 and especially p. 83. Ibid., p. 83. Impact on Minority Group s The implications of a national CCH system for employment and licensing decisions involving minority groups, particularly blacks, is a subject of debate. Certain minority group members do have a higher probability of police contact, and account for a disproportionate percentage of arrest records. Professional recordkeepers (CCH system managers) acknowledge this reality, but point out that criminal history record systems simply reflect data created by local agencies. Various studies have estimated the percentage of blacks with arrest records as ranging from 30 percent nationwide to over 50 percent in certain cities such as Philadelphia. The 30 percent nationwide estimate used FBI Uniform Crime Report data as a baseline, and corrected for double counting arising from multiple arrests. The Philadelphia study found that one-half of all black males in the sample had already been arrested at least once, as an adult, by the time they were 30 years old. 9 As Neal Miller, A Study of the Number of Persons With Records of Arrestor Conviction in the Labor Force (Washington, D. C.: U.S. Department of Labor, 1979). of February 21, 1980, blacks accounted for about 29 percent of all records in the NCIC/ CCH file, ]o which is almost triple the percentage of blacks in the total U.S. population. Some further insight can be obtained by looking at the disposition of adult felony arrests in California where information was available to OTA. In 1981, although blacks accounted for 30.6 percent of arrests and about 7.7 percent of Californias total population, they accounted for 37.7 percent of the law enforcement releases. In other words, blacks are disproportionately released after arrest without being formally charged. Likewise, blacks account for a disproportionate number (38.7 percent) of complaints denied. Whites, on the other hand, account for 44.2 percent of the arrests but only 33.1 percent of the law enforcement releases and 36.2 percent of the complaints denied. 11 Minutes of the Dec. 10-11, 1980, NCIC Advisory Policy Board meeting, Topic #13, NCIC Race Categories and Codes Used in the Wanted Persons, Missing Persons, and CCH Files, P. 38. State of California Department of Justice, Bureau of Criminal Statistics, fiirniml ~~s~jce I+ofile, 1981.

PAGE 149

142 An Assessment of Alternatives for a National Computerized Criminal History System Thus, blacks in California are more likely to have arrests that result in law enforcement releases and complaints denied. A law enforce ment release occurs when police detain and arrest a person, obtain fingerprints, report the arrest to the State record system, but subsequently release the person and do not present the case to the district attorney. A complaint is denied when the police arrest and present a person to the district attorney, but the district attorney decides not to prosecute the case. Releases and complaints denied may occur for a variety of reasons, such as insufficient evidence, refusal of the victim to prosecute, lack of probable cause, unavailable witness, or illegal search. Despite the dismissal of an arrest by the law enforcement agency or the denial of a complaint by the district attorney, the arrest event is recorded in the State criminal history record system. The evidence in California seems to indicate that blacks are more likely to be arrested, to have that information reported to the State repository, and then be released without any formal charges being presented. In this instance, the California recordkeeping system is operating precisely as it was designed to operate; it merely records and retains information concerning a police contact when it is submitted by the police, even if the arrest information does not lead to a formal charge or a court disposition. Some States have adopted stringent expungement rules regarding arrest-only information. For example, in New York State, arrest events that do not lead to conviction are sealed and the fingerImpact o n Federalism-the balance of authority and power between Federal, State, and local governmentshas been a central issue in the debate over a national CCH system. Because of the decentralized nature of the U.S. criminal justice process and because the generation and use of criminal history information occurs mostly at the State and local levels prints taken with such events are returned to the originating agency. The New York law states that it is the responsibility of the arresting agency, the prosecutor, or the judge to inform the State repository. In California, felony arrests that result in detention only are retained in the State criminal history record for 5 years, and felony arrests that otherwise do not result in a conviction are retained for 7 years. As discussed earlier, a criminal arrest record, even without convictions, can have an adverse impact on employment and licensing decisions. Indeed, the courts have found that a policy of refusing employment to blacks with an arrest record without convictions had a racially discriminatory impact because blacks are arrested substantially more frequently than whites in proportion to their numbers. 12 Similar judicial reasoning has been extended to black applicants refused employment due to criminal convictions where the offense does not significantly bear upon the particular job requirements. 13 In this context, any discriminatory impacts from the use of national CCH information would depend on whether, and under what conditions, noncriminal justice access is permitted. z American Civil Liberties Union, Employment Discrimination Against Persons with Criminal Records, draft report, 1977, p. 11, and Gregory v. Litton Systems, 316 F. Supp. 401 (C.D. Cal. 1970), modified on other grounds, 472 F. 2d 631 (9th Cir. 1972). Green v. Missouri Pacific RR 523 F. 2d 1290 (8th Cir. 1975), at 1298. Federalis m of government, most States seek a primary role in any national CCH system. State governments have basic jurisdiction over law enforcement and criminal justice within their borders under their constitutionally reserved powers, and many have been reluctant to share this jurisdiction with the Federal Government except with respect to Federal offenders. Most States have appreciated other kinds

PAGE 150

Ch. 12Other Impact Areas Relevant to a National CCH System 143 of support from the Federal Government, such as the FBI fingerprint identification services and the Law Enforcement Assistance Administration funding for State CCH system development, as long as this support was provided on a voluntary basis and the States retained control over the operation and use of their own criminal history record systems. The Federal Government has a legitimate interest in the enforcement of Federal criminal law; in the prosecution of Federal offenders, whether intrastate or interstate; and in assisting with the apprehension of interstate and international criminal offenders who cross State and/or national borders. To the extent that crime is perceived as a national problem deserving national attention, the Federal Government also has a defined role in the provision of voluntary support to State and local law enforcement activities. Many of the proposed alternatives for a national CCH system encounter difficulties resulting from the historic constitutional division of powers and duties in our Federal system. Since the standards of the States vary so widely, any national standards for a CCH system could easily conflict with those of at least some States. Massachusetts and Florida illustrate the potential conflict with respect to system access. Both States have developed their own CCH systems, both have given considerable attention to the question of who should have access to these systems, and each has adopted a very different approach. Massachusetts has passed a Criminal Offender Record Information Act which defines the classes of agencies and individuals that are eligible for access. Criminal justice agencies may receive criminal history records after being certified by a Criminal History Systems Board. Noncriminal justice agencies granted access by statute must also be certified to receive criminal history records to carry out statutory duties. Other agencies and individuals may be certified to receive criminal history record information only if the public interest in disseminating such information to these parties clearly outweighs the interest in privacy and security. 14 Such agencies and individuals must be certified by both the Criminal History Systems Board and the Security and Privacy Council. On the other hand, Florida has passed a Public Records Statute that makes its CCH records generally available for access by noncriminal justice agencies and private citizens. 15 For a $2 fee, a number of private firms located in Florida have CCH record access that would be denied in many other States. Massachusetts officials have stated in the past that they would not contribute State records to a national CCH file unless they could retain control over access to and dissemination of these records. Massachusetts has been particularly concerned about indirect access to criminal history records by agencies or individuals not authorized to receive such records directly. As a hypothetical example, a private firm with branch offices in both State A and State B could conceivably be denied access to CCH records in State A and permitted access in State B. If State B CCH records were included in a national CCH file, the private firm could circumvent State A law by gaining access to State A records via a request from its State B branch office to the national CCH file processed through the State B CCH system. Information denied to the private firm in State A therefore could be obtained indirectly through access in other States to a national CCH file. While Florida now requires that the access policy of the donor State (State holding the record) be respected, most States do not require that donor State policy be followed. l6 The problem of secondary dissemination can also occur with Federal agencies. For example, Massachusetts State and local police agencies provide arrest records to the FBI, which is authorized to receive such information for law enforcement purposes. Under Federal regulations, the FBI may disseminate arrest in Massachusetts General Laws, ch. 6, sec. 172. 16 Florida Statute 119. Access is restricted for records sealed pursuant to F,S. 893.14 (first offense possession of drugs) and F.S. 901.33 (first offenders who are acquitted or released). SEARCH Group, Inc., The Interstate Exchange of Criminal History Records, Sacramento, Calif., May 1981, pp. 5-7.

PAGE 151

144 An Assessment of Alternatives for a National Computerized Criminal History System formation in its files to other Federal agencies authorized by executive order, such as the Office of Personnel Management (OPM) pursuant to Executive Order No. 10450. 17 However under Massachusetts State law, OPM is authorized access only to Massachusetts offender records pert aining to Federal Riot Ac t 18 convictions within the last 15 years. Thus, what OPM cannot get from Massachusetts directly, it can get through the FBI indirectly. As a consequence, Massachusetts has declined to provide records to NCIC/CCH and has severely curtailed fingerprint submissions to Ident. As discussed in chapter 9, in addition to different policies on access to and dissemination of criminal history information, States also vary widely in their statutes and regulations on file content, access, review and challenge procedures, sealing and purging, record accuracy and completeness, court disposition monitoring, and transaction logs and local audits. Even States with similar policies may show considerable variation in the level of resources and management devoted to enforcement. They also differ in their definition of crime; a felony in one State may not be so considered in another It was thought by some that the develop ment of Federal regulations in the area of criminal history information systems (title 28, Code of Federal Regulations, pt. 20) would provide the answer to overcoming many of the difficulties of sharing criminal history information among the States and between the States and the Federal Government. Despite a dramatic increase in State statutes and regulations, many States have experienced a number of problems in implementing the Federal regulations, including insufficient resources, confusion in interpreting the regulations, and the lack of a State legislative mandate. In the OTA 50-State survey, 30 out of 46 States indi1720 CFR $ 20.33(2). 185 Usc j 7313. cated that insufficient funds was the principal constraint to fully implementing annual audits, unique record tracking numbers, 90-day disposition reporting, and the like. Fourteen States indicated that a lack of statutory or policy mandate was the principal constraint. l9 States that are willing to fully implement the Federal regulations face a significant problem in obtaining sufficient funds. Throughout the 1970s, it was Federal Government policy to support the development of State CCH systems and the implementation of the Federal regulations. However, this funding has now ended. In any event, the major portion of the long-term costs of a national CCH system would be the operating costs incurred by participating State and local criminal justice agencies, rather than development costs. Until recently, some of these costs could be recovered through Federal block grants, but this avenue of Federal support has also been phased out. Thus, at present, the States and localities would have to bear most of the cost of any national CCH system. The only exceptions would be the direct cost of any federally operated facilities (such as Ident and NCIC), and the costs of Federal agencies participating in the system. The difficulty of finding new money or reducing other expenses to pay for a national CCH could discourage State participation Some States in the past have criticized what they believe to be the excessive cost of national CCH alternatives that call for substantial duplication of records at State and Federal levels. In any case, financing could be particularly difficult for States with less well-developed CCH systems, less need for a national CCH system (e.g., relatively low levels of interstate criminal movement), and/or less ability to pay (e.g., smaller, poorer States). Wfice of Technology Assessment 50-State Survey conducted in 1979-80.

PAGE 152

Ch. 12Other Impact Areas Relevant to a National CCH System l 145 Impact on Monitoring o r Surveillance Potentia l Several alternatives for a national CCH systemparticularly those involving large central files and/or extensive message switching have generated concern about their possible use for monitoring or surveillance. The flagging of criminal records is a common monitoring or surveillance practice and an accepted law enforcement tool. Placing a flagon a file helps law enforcement personnel keep track of the location and activity of a suspect, apprehend wanted persons, or recover stolen property whenever there is a police contact. At the State level, both manual and automated files are used for flagging. This practice differs from State to State. The most frequent application seems to be for parole violators and wanted persons. Others include flagging for modus operandi, for individuals with a history of violent acts, and for vehicle files. Several States indicated that flagging is easier with automated systems 2 0 At the Federal level, the flagging of records in Ident is usually through use of a wantedflash-cancellation notice for persons with an outstanding arrest warrant (wanted notice) or persons placed on probation or parole (flash notice). A cancellation notice is posted when the person is no longer wanted or under supervision 21 With respect to NCIC, since hot files are flags by definition, all wanted and missing persons and stolen property records included in NCIC represent flags to law enforcement and criminal justice users One concern expressed about a nationa l CCH system focuses on its possible use for indiscriminate Government monitoring or surveillance of individual citizens or groups of citizens. For instance, a national CCH system could be used to run a criminal history record .-.-_ Based on interviews with State criminal records repository personnel. *Letter from Conrad S. Banner, Identification Division, FBI, to Marcia MacNaughton, OTA, U.S. Congress, dated July 26, 1979. check similar to the relatively common police practice of running a warrant check on motorists stopped for traffic violations. Here, the drivers name and perhaps license number are checked against the local, and sometimes State and Federal, files. Also, the vehicle license plate number and description might be checked against the stolen vehicle files. Even this practice has been challenged in the courts, especially where the police detain individuals and conduct routine checks not based on reasonable suspicion. The courts have found that a detention of an individual which is reasonable at its inception may exceed constitutional bounds when extended beyond wha t is reasonably necessary under the circumstances. 22 If the computer check is conducted during a detention, and the detaining officer has no reason to consult the computer other than curiosity, the resulting conviction may be overturned 23 Indiscriminate criminal history checks on individuals would appear to be even more likely than hot file checks to conflict with the constitutional protections provided by the fourth, fifth, and 14th amendments Concern about a CCH system also has focused on its possible use by Federal agencies and particularly the FBIfor monitoring or surveillance of the lawful activities of individual citizens or organizations. This concern is based largely on the well-documented tendency of the FBI and other Federal agencies, in the late 1960s and early 1970s, to expand intelligence investigations into the realm of political surveillance.24 ZzPeople v. Harris, 1975, 15 Cal. 3D, 384, 390. g Pennsylvania v. Jones, U.S. Sup. Ct., No. 77-958 (Mar. 27, 1978). U.S. Congress, Senate Select Committee to Study Governmental Operations with Respect to Intelligence Activities, Inte} Ligence Activities and the Rights of Americans, Book II, 94th Cong., 2d sess., Apr. 26, 1976, p. 4. See also Seth Rosenfeld, The Berkeley Files: 17 Years of FBI Surveillance in Berkeley, The Daily Californian, May 28 and June 4, 1982.

PAGE 153

146 l An Assessment of Alternatives for a National Computerized Criminal History System For example, the FBIs COINTELPRO and COMINFIL programs were designed to disrupt groups and to neutralize individuals who were considered threats to domestic security, such as civil rights and anti-Vietnam War leaders. Among the tactics employed was the use of criminal arrest records to impede the political careers of individuals who the FBI deemed to be threats. 25 Other intelligence programs resulted in widespread invasions of the privacy of American citizens through the use of surveillance strategies ranging from mail covers and openings to wiretapping and surreptitious entry. 26 Also during the early 1970s, the FBI made very limited use of NCIC for intelligence purposes which, although law enforcement in nature, had not been authorized by Congress. This was revealed in 1975 during hearings before the Subcommittee on Constitutional Rights of the U.S. Senate Committee on the Judiciary. On July 15, 1975, the Subcommittee Chairman charged that the FBI was using NCIC to keep track of individuals that might be of interest to the FBI for whatever purposes, including possibly political reasons The FBI conceded that a pilot flagging program using NCIC had been operational from April 1971 to February 1974, but did not exceed 4,700 active flags. The FBI never advised either State and local officials, or apparently congressional officials, about the flagging program because it was experimental in nature. According to the FBI, the flagging practices were confined to national security intelligence investigations and to the tracking of Selective Service delinquents, top jewel thieves, and bank robbery suspects. The projects objective was to enable law enforcement agencies to locate, through the NCIC, individuals being sought for law enforcement purposes who did not meet the criteria for inclusion in the NCIC wanted person file. 27 In other words, NCIC was being used to track Ibid., p. 10. Ibid., p. 38. zT~t@r from H~o]d R. Tyler, Jr., Deputy Attorney Gener~~ U.S. Department of Justice, to Senator John Tunney, Chairman, Subcommittee on Constitutional Rights, Committee on the Judiciary, U.S. Senate, Oct. 29, 1975. individuals who had not been formally charged with a crime and did not have an outstanding warrant for a Federal offense or other extraditable felony or serious misdemeanor offense. Since that time, the FBI has rejected all requests or proposals for intelligence use of NCIC.* During the course of the OTA study, FBI officials have repeatedly stated to Congress and to OTA that they will not permit Ident or NCIC to be used for unauthorized purposes of any kind. 28 FBI officials believe that a national CCH system would not have any significant surveillance potential since surveillance, by definition, means a close watch over someone. FBI officials assert that a system such as NCIC which depends primarily on chance contacts with law enforcement officers does not meet this definition and certainly represents little, if any, danger to law-abiding citizens. Strong and independent policy control over a national CCH system and tight restrictions on noncriminal justice access, coupled with outside audit and explicit statutory guidelines for operations, would help protect against the possibility-however remotethat a national CCH system could be used at some point in the future in violation of first amendment or other constitutional rights. In comments to OTA, various criminal justice officials have suggested a statutory prohibition on intelligence use of the Interstate Identification Index or any other national CCH system. On the other hand, some State officials have noted that there may be legitimate intelligence and surveillance applications of a national CCH system, and that these possibilities should not be abandoned solely because of their sensitivity. A final concern involves the intercomection of FBI criminal record systems with other Federal information systems that might collectively constitute a de facto national data bank with even greater monitoring and surveillance potential. For example, the already xA~ of septem~r 1982, the Department of Justice ~d FBI had approved but not yet implemented a U.S. Secret Service proposal to establish an NCIC file on persons judged to represent a potential threat to protectees. 28 See testimony of William Bayse, FBI, before the Oct. 22, 1981, hearing of the Subcommittee on Civil and Constitutional Rights, House Judiciary Committee.

PAGE 154

Ch. 12Other Impact Areas Relevanf to a National CCH System l 147 authorized interconnection of NCIC with the Treasury Enforcement Communication System and the Justice Telecommunication System means that the NCIC data base is accessible to dozens of Federal agencies, including the Internal Revenue Service (IRS), Bureau of Customs, and Immigration and Naturalization Service, among others.* Although these agencies are subject to Federal law and regulations and NCIC operating procedures, the actual use of NCIC data by Federal agencies does not ap*see Ch. 4. pear to be subject to outside audit or oversight. Some agencies surveyed by OTA have developed their own detailed procedures. For example, IRS Criminal Investigative Division procedures require that NCIC be queried when evaluating possible tax fraud. While such Federal agency use is entirely legal, the interconnection of networks and information systems in effect extends the overall surveillance potential. *See ch. 6. Impact on Constitutional Right s The enactment of national legislation could provide explicit guidelines for the operation and use of a national CCH system and include statutory protections against the use of CCH information in ways that might violate constitutional rights. Such legislation could mandate specific accountability alternatives such as access, review, and challenge procedures, criminal penalties, and privacy standards. Comprehensive legislation would help ensure a major and continuing role for Congress in the development and oversight of a national CCH system. SEARCH Group, Inc., among others, has concluded that enactment of Federal legislation may be the single most important factor in developing a national index or any other national CCH system 29 SEARC H Group believes that in addition to protecting constitutional rights 3 0 legislation is necessary in order to: 1) provide a clear mandate for a national CCH system; 2) establish a strong national commitment in terms of political and financial support; and 3) specify which organi2 SEARCH Group, Inc., Essential Elements and Actions for Implementing a Nationwide Criminal History Program, Sacramento, Calif., February 1979, p. 4. 30 SEARCH Group, Inc., Standards for Security and Privacy of Criminal Justice Information, Sacramento, Calif., January 1978, pp. 18-19. This report covers many other areas, such as sealing and purging standards, that might be covered in comprc+ hensive legislation. zations or entities shall have policy and/or management responsibility .31 In the absence of comprehensive Federal legislation, a national CCH system could be established through user agreements among the 50 States or by an interstate compact. In the former arrangement, each State would hav e to execute user agreements with all othe r States. To create an interstate compact, each State legislature would have to ratify the compact which would then be signed by the Governor. Congress would need to enact legislation consenting to the compact, followed by the signature of the President. 32 Whether or not establishing user agreements or an interstat e compact would be less cumbersome and more feasible than enactment of comprehensiv e Federal legislation is an open question. However, it seems likely that the legislative route would provide stronger and more direct protection of constitutional rights Outside audit is another accountabilit y measure found to be very important in protecting constitutional rights and was recom 3SEARCH Group, Inc., Essential Elements and Actions for Implementing a Nationwide Chlninal History Program, Sacramento, Calif., February 1979, p. 4. SEARCH Group, Inc., The Ieasibfity of an Interstate Gbmpact for Exchanging Criminal History Information, Sacramento, Calif., April 1980, pp. 3-4.

PAGE 155

148 l An Assessment of Alternatives for a National Computerized Criminal History System mended by SEARCH Group. 33 Outside audit would be necessary to ensure that a national CCH system was being operated and used for authorized purposes and in accordance with any guidelines established by Congress. The General Accounting Office and/or an independent national board could conduct audits of Federal and State agencies, and State CCH repositories (or possibly independent State boards) could conduct audits of user agencies within the State. The first amendment provides that Congress shall make no laws abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for redress of grievances. First amendment rights could be violated to the extent a national CCH system was used to monitor the lawful and peaceful activities or associations of citizens or to discourage such activities or associations through the dissemination of criminal history information. The dissemination of arrest-only information for noncriminal justice purposes could violate an individuals freedom of speech and association. 34 Strong and independent policy control over a national CCH system and tight restrictions (or a prohibition) on noncriminal justice access, coupled with outside audit and compre hensive legislation, would help minimize the possibility that a national CCH system could be used at some point in the future in violation of first amendment rights. The right to privacy is embodied principally in the fourth amendment, which guarantees the right of the people to be secure in their persons, houses, papers, and effects, against unreasonable searches and seizures. The courts have generally taken a middle ground here in recognizing that the individuals fundamental right to privacy and the potential economic and personal harm that results if his arrest becomes known to employIbid., p. 49. *SW ch. 14 for further discussion. 34 See Schware v. Board of Bar Examiners of New Mexico, 353 U.S. 252, 1957, where the court held that a State cannot refuse to permit a law graduate to take a bar exam where such refusal was based on arrests for offenses for which the applicant was not tried or convicted. ers, credit agencies, or even neighbors must be balanced against the importance of arrest records to law enforcement officials. 36 Tight restrictions on noncriminal justice access and mandatory quality standards (especially with respect to disposition reporting) for those records that are disseminated outside the criminal justice community would help minimize the possibility that information from a national CCH system could be used in violation of individual privacy. The fifth amendment provides that no person shall be deprived of life, liberty, or property, without due process of law. The 14th amendment extends this due process protection to the States. The presumption of innocence has been construed as falling within the concept of liberty. Furthermore, the courts have held, under certain circumstances, that the dissemination and/or use of incomplete and/or inaccurate arrest and conviction records violates due process. 37 This is especially the case when the complete and accurate criminal history information was otherwise available but was not used as a basis for criminal justice decisions. The fifth amendment guarantees the right to a speedy and public trial, by an impartial jury ... and to have the assistance of counsel for his defense, among other criminal process rights. The eighth amendment provides that no person shall be subjected to excessive bail or fines, nor cruel and unusual punishments. In the leading case of Tatum v. Rogers (1979), the court found that the use of rap sheets containing erroneous, ambiguous, or incomplete data with respect to prior arrests and dispositions in setting bail constituted violation of the 14th (due process), sixth (right to effective assistance of counsel), and eighth (right to reasonable bail) amendments. Mandatory record 3S% State v. Pinkney, 290 NE2d 923,924 (C.D. Ohio, 1972). a See Ibid.; Houston Chronicle Publishing Cbmpany v. City of Houston, 531 SW2d 177, 187 (Court of Civil Appeals of Texas, 14 Dist. 1975); William Rehnquist, Is an Expanded Right of Privacy Consistent With Effective Law Enforcement, Kansas Law Review, vol. 23, fall 1974, pp. 1-22; and Menard v. Mitchell 430 F.2d 486 (D.C. Cir. 1970). see ch. 6 for discussion of Tarkzm v. Saxbe (1974) and Tatum v. Rogers (1979). *SW further discussion in chs. 6 ad 11

PAGE 156

. Ch. 12Other Impact Areas Relevant to a National CCH System l 149 quality standards established by statute and backed up by the necessary funding and technical assistance to ensure implementation (and outside audit to ensure compliance) appear to be the most effective ways to minimize the possibility that national CCH informatio n could lead to fifth, sixth, eighth, and 14th amendment violations The 14th amendment also guarantees th e equal protection of the laws to any person within the jurisdiction of a given State. It is also arguable that any act by the Federa l Government which would in effect be a denial of equal protection of law would constitute a deprivation of liberty prohibited by the fifth amendment due process clause. 38 Thus, for example, it can be argued that the use of incomplete criminal history records (especially when lacking information on dispositions that have already occurred) violates equal protection of the law by merging the distinction between innocence and guilt. a William A. Ratter, C2msitutiona] Law, Gilbert Law Summaries, Gardena, Calif., 1970, p. 87. See I?oMng v. Sharp, 347 U.S. 497, 1954. SEARCH, Standards for Security and Privacy, op. cit., p. 5.

PAGE 157

Chapter 13 Congressional Policy Considerations, Part I

PAGE 158

-Contents Page Chapter Summary. . . . . . . . . . . . 153 Action or No Action . . . . . . . . . . 153 Further Study . . . . . . . . . . . . 153 Strengthen State/Local CCH Systems . . . . . . . 154 Select a National CCH System . . . . . . . . 154 Action or No Action.. . . . . . . . . . . 155 Further Study.. . . . . . . . . . . . . 157 Local Level Use . . . . . . . . . . . 157 Use of Out-of-State Information . . . . . . . . 157 Record Quality.. . . . . . . . . . . . 157 costs . . . . . . . . . . . . . . 157 Strengthen State/Local CCH Systems . . . . . . . 158 Improving Management of Already Existing CCH Systems. . . 158 Developing New Systems . . . . . . . . . 158 Clarifying Title 28 Privacy and Security Regulations . . . 159 Improving Court Disposition Reporting. . . . . . . 159 Range of Possible National CCH System Structures . . . . 160 Select a National CCH System . . . . . . . . 160 Possible Impacts on the Criminal Justice Process . . . . . 161 Importance of Complete, Timely, and Verifiable Information . . 161 Improving Record Quality-Difficult But Possible . . . . 162 Shifting Preference on system Structure . . . . . . 164 TABLES Table No. Page 31. Comparison of Possible National CCH System Alternatives . . 160 32. Shifting Preferences of Federal and State Crimin al Justice Officials for a National CCH System. ........ . . . . . . 164

PAGE 159

Chapter 13 Congressional Policy Considerations, Part I Chapter Summar y A wide range of policy options are available to Congress relative to a national computerized criminal history (CCH) system. The choice of an option or set of options will depend in large part on congressional judgments regarding the importance of a national CCH system relative to other priorities, and which national CCH system alternative offers the most acceptable balanc e verse impacts Actio n Arguments for between beneficial and ador No Actio n No Action. Criminal history lnformatlon is only one element in the criminal justice process. There are competing priorities for congressional attention and funding, such as expanded prison facilities or revision of the criminal code. Also, it might be argued that criminal justice decisionmakers are aware of the current deficiencies in criminal history information and take them into account, and that effective control over a national CCH system would be difficult to achieve. The benefits of a national CCH system are further constrained by the local nature of most crime, and the diversity and constitutional prerogatives of the States Arguments for Action. The National Law Enforcement Telecommunications Syste m (NLETS) and the National Crime Information Center (NCIC) hot files are two examples of successful State cooperation in national systems. The lack of Federal direction has been a major barrier to a national CCH system even though a Federal role could be justified on several grounds. Many believe that a clear decision on the future of NCIC/CCH is needed and that this requires major congressiona l involvement Perhaps the strongest argument for congressional action is that criminal history information is vital to and used at virtually every stage of the criminal justice process, and that several surveys have identified a need for more timely and complete criminal history information. This need becomes even more important in view of recent and proposed criminal justice reforms that place greater reliance on criminal history information. Some argue that the time is ripe for a decision on a national CCH system owing to new alternatives now feasible with advanced computer and communication technology; the dramatic progress by many States in automating their own systems and enacting State statutes on criminal justice information systems; the progress of the Identification Division (Ident) in automating its own operations through the Automated Identification Division System (AIDS) program; and the early results of the Interstate Identification Index (III) pilot test and Phase I development. Further Stud y Additional study could be carried out on the use of criminal history information at the local level, the use of out-of-State criminal history information, and record quality. For example, while the OTA record quality research documented problems with record quality. Congress may wish to commission additional rec ord quality studies to provide further statistical confirmation. OTA did not attempt to make an independently verified estimate of the total cost of a 153

PAGE 160

154 An Assessment of Alternatives for a National Computerized Criminal History System national CCH system. Congress may wish to commission a detailed cost analysis of any system alternatives selected for serious consideration. Strengthen State/Loca l CCH System s Congress may wish to focus initially or solely on actions to strengthen State and local CCH systems or to strengthen such systems in parallel with the development of a national CCH system. OTA research identified four possible priorities for strengthening State and local CCH systems. The first is funding and technical assistance to improve the management of already existing CCH systems. A second is funding and technical assistance for the development of automated systems in States that are still using a manual system. Should Congress wish to pursue these two priorities, budget action will be required since CCH-related grants previously provided through the Law Enforcement Assistance Administration (LEAA) were eliminated in fiscal year 1981. A third priority is the clarification of title 28 regulations on privacy and security of criminal history information systems. These regulations have now been in effect for about 7 years. The time may be ripe for Congress to initiate a review of title 28. The results of this review could be reflected in a title 28 revision to be promulgated through normal rulemaking procedures or in comprehensive legislation, should Congress choose to move in this direction. A fourth priority is the improvement of court disposition reporting. Some c riminal justice decisionmakers see this as a key prerequisite for a national CCH system. Given the constitutional separation of powers, Congress is limited in what it can do to encourage reform and reorganization in the judicial branch. However, Congress could authorize financial and technical support to the States for use in developing court disposition reporting systems, and could commission a survey of the current status and needs of the judiciary with respect to disposition reporting systems. Select a National CCH Syste m Should Congress wish to select a national CCH system, OTA found that three of the many possible alternative systems could most readily evolve from current systems: 1) a national CCH repository on serious offenders; 2) a single-State/multi-State CCH system with partial message switching; and 3) a national index (III) with partial or no message switching. To the extent that a national CCH system provides more complete, timely, and verifiable (based on positive identification) information than is presently available, the system would improve the functioning of the criminal justice process. The most significant improvements are likely to be in the areas of criminal investigation, police booking and intake, pretrial release and bail decisions, and presentence investigation reports. Better information is also important to sentencing, correctional, and, to a lesser extent, probation and parole decisions. Available evidence indicates that improvement in criminal identification and criminal history exchange by the Federal Government and most States is facilitated through computerization. Response times of 5 to 10 hours for fingerprint identification checks and a few seconds to several hours for criminal history checks are technically feasible. However, actual response times will depend on the priority assigned by the States and Federal Government, staffing and resource (including computer) limitations, and policies on record dissemination and use. While computerization can improve response time, improvements in record quality are more difficult to achieve. Untimely or incomplete disposition reporting is a significant problem in many States. Nonetheless, the experience of three States with online CCH files California, Minnesota, and North Carolina shows that improvement is possible. Available evidence indicates that strengthening State

PAGE 161

Ch. 13Congressional Policy Considerations, Part 1 l 155 and local criminal history systems and court disposition reporting systems is a prerequisite to improving record quality, regardless of the national CCH system structure. Despite movement among many State and Federal criminal justice officials toward agreeAction o r Congress always has the option of taking no action. Criminal history information is only one element of the criminal justice process. There are, of course, competing priorities for congressional attention and funding. These include, for example, additional trained police investigators, new or expanded prison facilities, better local community support (e.g., neighborhood watch programs), or revision of the criminal code. It might be argued, furthermore, that currently available criminal history information, however imperfect (e.g., incomplete with respect to disposition reporting), is good enough since most criminal justice decisionmakers are aware of these deficiencies and take them into account. The potential benefits of a national CCH system are further constrained by the local nature of most crime. It also might be argued that effective control over the contents and use of any national CCH system would be difficult to achieve, given the very large number of potential users and access points. Some technology experts believe that the time for a decision is not ripe because new technologies, such as small, lowcost computers with flexible programing capability and improved data security techniques, will soon offer alternatives that are inherently easier to control. Finally, some argue that since the States, under the constitution, have basic jurisdiction over law enforcement and criminal justice within their borders, any efforts to implement a national CCH system should be undertaken by the States, not by Congress or the Federal Government in general. There is also the concern that perhaps the States are too diverse ment on a national index (III) concept (along with a national fingerprint repository), questions raised with respect to policy control, noncriminal justice access, and record quality, among other issue areas discussed in the following chapter, have yet to be resolved. No Actio n in their criminal justice laws and practices for a national CCH system ever to be feasible. On the other hand, NLETS is an example of successful State cooperation in the operation of a national communication system. Also, the success of the NCIC hot files is evidence that Federal, State, and local law enforcement agencies can work together effectively. Some State and local criminal justice officials claim that the major barrier to a national CCH system has been the lack of clear direction at the Federal level, although in the last 2 years there has been some movement toward a consensus on the national index (III) as a system structure. A Federal role in a national CCH system could be justified to the extent that the system assisted in the enforcement of Federal criminal law and prosecution of Federal offenders, whether intrastate or interstate, and in the apprehension of criminal offenders who cross State and/or national borders. To the extent that crime is perceived as a national problem, the Federal Government could define its role in a national CCH system as simply another way to provide voluntary support to State and local law enforcement and criminal justice activities. From a legal standpoint, a Federal role could be based on the criminal record information needs of Federal agencies as established by various Federal statutes and executive orders, implementation of title 28 regulations for State and local criminal justice information systems that have used Federal funding,** the interstate commerce clause of the U.S. Constitution, and the constitutional pro*See ch. 6. **sW chg. 5 and 6.

PAGE 162

156 An Assessment of Alternatives for a National Computerized Criminal History System visions (including the first, fourth, fifth, sixth, eighth, ninth, and 14th amendments) guaranteeing individual rights of privacy and due process. Perhaps the strongest argument for congressional action is that criminal history information is vital to and used at virtually every stage of the criminal justice process, and that surveys conducted by OTA, the U.S. Department of Justice (DOJ), and the Florida Department of Law Enforcement identified a need for more timely and complete criminal history information.* This need becomes even more important in view of State bail and sentencing reforms that place greater reliance on criminal history information, and the many recommendations of the U.S. Attorney Generals Task Force on Violent Crime that involve criminal history records. For example, the task force recommended action to establish an NCIC file on firearms violators, to require a mandatory criminal records check on handgun purchasers, and to deny bail to persons with prior convictions for serious crimes committed while in pretrial release status. 2 Incomplete disposition data is perceived as a major problem in Federal and some State criminal record systems, and this was confirmed by the OTA record quality research.** An additional problem with Ident is the slow response time, due in large part to the manual processing of fingerprint cards. These problems were noted by the Attorney Generals Task Force that recommended, among other things, swift completion of the Ident automation program. 3 Another argument in favor of action is the 12-year debate over the NCIC/CCH program. See SEARCH-Group, inc., S~azI&r&for Security and Privacy of Criminal Justice Information, Sacramento, Calif., January 1978, p. 18. *See ch. 11. Attorney Generals Task Force on Violent Crime, Final Re por~ U.S. Department of Justice, Aug. 17, 1981, pp. 13,29,50. **See ch. 8 Task Force, Final fipor~ op. cit., pp. 11-12, 18-19, 67-69. A number of criminal justice decisionmakers interviewed by OTA and DOJ noted the substantial time and money already expended on the NCIC/CCH debate with no real resolution of several underlying issues. Many believe that a clear decision on the future of NCIC/ CCH is needed, and that this decision would require a major, if not dominant, congressional involvement. Finally, four key trends are cited to support congressional action. First, computer and communication technology has advanced to the point where both centralized and decentralized system structures are possible. Computer hardware is less expensive than either the software or the communication links, and decentralized systems can be cost effective.* Second, many States and localities and several Federal agencies have developed their own CCH capability.** At the State level, CCH systems (when compared to manual) obtain relatively higher arrest and court disposition reporting rates and have shown significantly greater improvement.*** Many States have made dramatic progress in automating their own systems and in enacting State statutes and regulations on criminal justice information systems.**** Third, Ident has made progress in automating its own operations through AIDS. Ident could become part of a national CCH system and could compete with, duplicate, and/or supersede the existing NCIC/CCH file. Congressional action may be necessary to determine the most cost-effective role for Ident/AIDS and NCIC/CCH as separate systems and, perhaps, as part of a national CCH system. Fourth, the early results of the III pilot test and Phase I development suggest that it is operationally as well as technically feasible. Some point to these developments as evidence that this is a propitious time for a decision on a national CCH system. *see Chs. 5 and 10. **See ch. 4. ***See ch. 9. ****See ch. 6. TSee ch. 4.

PAGE 163

Ch. 13Congressional Policy Considerations, Part 1 l 157 Further Stud y In the course of the OTA research, a number of areas for possible further study were identified, including the following: Local Level Us e One possible area is the use of criminal history information at the local level. The flow of criminal history information in the United States resembles an enormous pyramid. OTA was able to identify the flow of information at the Federal and State levels in a fairly systematic way through analysis of Ident and NCIC/CCH use statistics, interviews with Federal user agencies, the comprehensive 50-State survey and telephone or written interviews with State repository personnel in 44 States (including site visits to 5 States). However, at the local level, given the vast size of the user population, OTA had to depend on the results of limited interviewing and comments received on the draft report. Should Congress seek additional information on the needs and perspectives of local criminal justice decisionmakers, a written survey could be used to cover a large number of local areas, as was done at the State level with the OTA 50-State survey. Use of Out-of-Stat e Informatio n A second area of possible further study is the use of out-of-State criminal history information. A major justification for a national CCH system is the perceived need to exchange out-of-State criminal history information. However, OTA found that the usefulness of out-of-State criminal history information can be limited due to difficulties in interpreting the records, variations from State to State in the definition of many criminal offenses, wide discrepancies in local police behavior and criminal justice practices, and the substantial time and effort required to validate incomplete and/or ambiguous records. It could be advantageous to determine more precisely the nature and ex4 For suggestions on a research methodology, see working paper B, sec. VII, pp. 236-240. tent of these and other constraining factors and how they might best be overcome. Record Qualit y A third area in need of further study is record quality. The results of the OTA record quality research documented various record quality problems. Congress may wish to commission additional studies to provide statistical confirmation of present levels of record quality. In order to establish the exact imprecations of record quality, a sample of cases could be followed through a single jurisdiction for a period of time, with close attention given to how the records are used and interpreted. cost s A fourth area needing further study is costs. As noted in chapter 5, quantifying the costs of a national CCH system is very difficul t mainly due to the absence of systematic cost data at the State and local levels. OTA did not attempt to make an independently verified estimate of the total costs. Congress may wish to commission a detailed cost analysis of any system alternatives selected for serious consideration Whether or not further study in some or all of the areas discussed above is necessary at this time is a judgment best left to Congress. The U.S. Senate was concerned enough about the lack of information and analysis to have included a new study effort in the DOJ Appropriation Authorization Acts for fiscal year s 1981 and 1982. Provisions of these acts called for the Attorney General, after consultation with the Committees on the Judiciary of the Senate and the House of Representatives, to arrange for an appropriate independent entity to prepare a report and recommendation on a national CCH system. 5 This OTA repor t should provide one basis for determinin g whether a further comprehensive study i s needed See sec. 113 of the Department of Justice Appropriations Act of 1981 and Amendment 612 to the Department of Justice Appropriations Act of 1982 (Congressional Record Nov. 12, 1981, Pp. S13290-91.

PAGE 164

158 l An Assessment of Alternatives for a National Computerized Criminal History System Strengthen State/Local CCH System s Congress may wish to focus initially or solely on actions to strengthen State and local CCH systems or to strengthen such systems in parallel with the development of a national CCH system. In effect, the latter was the option exercised by Congress during the 1970s. In the absence of a clear-cut congressional decision on a national CCH system, Congress did authorize and fund both NCIC/CCH and Ident/AIDS at the national level (through the Federal Bureau of Investigation (FBI) budget), and the development and strengthening of State and local CCH systems (through the LEAA budget).* Partially as a result, the States have made substantial progress and are in a much better position technically to support a national CCH system than they were 12 years ago. Nonetheless, further improvements in many States and localities appear to be necessary. The record quality problems of Ident and NCIC/CCH reflect in part underlying deficiencies at the State and local levels. OTA research identified the following possible priorities for strengthening State and local CCH systems: Improving Management o f Already Existin g CCH System s States vary widely in their management practices.** Federal funding and technical assistance could be focused in those States most in need of improved procedures to assure the accuracy and completeness of criminal history information, to conduct audits of local users, to maintain and periodically review transaction logs, and to train CCH system employees and users. Developing New System s Funding and technical assistance could be provided for the development of automated *See ch. 5. **See ch. 9. systems in States that are still using a manual system. The purpose would be to help each State determine what kind of system would be best suited to its needs. Some States simply may not be able to justify a completely automated system. For example, several of the smaller States believe that an automated name index would be sufficient to permit their effective participation in a national system, and that computerizing the full records would not be necessary or cost effective. Should Congress wish to pursue these two priorities, budget action would be required since CCH-related grants previously provided through LEAA were eliminated in fiscal year 1981. The termination of the CCH funding pro gram resulted more from severe criticism leveled at LEAA as a whole than from specific problems with CCH-related grants.* Indeed, it can be argued that LEAA CCH-related grants made a significant contribution to the relatively rapid development of State CCH systems during the last 12 years.** Congress may wish to fund this grant program again, establish a new categorical grant program with a focus on the priorities outlined above, 6 or assign a priority to CCH system management and development as part of a larger block grant for State and local law enforcement. Congress could authorize States to fund their own CCH technical assistance out of Federal categorical or block grant money and/or retain a small technical assistance team in DOJ. In the past, LEAA has provided both direct technical assistance and funding for technical assistance efforts by SEARCH Group, Inc., and others. With the demise of *D~g the ]a~ 1970s LEAA was criticized for, among other things, subsidizing the operating budgets of local police departments, underwriting the purchase of expensive and unnecessary equipment, and in general spending tax dollars in ways that did not result in significant progress in the war against crime. **See chs. 4, 5, and 6. The Attorney Generals Task Force on Violent Crime has recommended legislation to authorize and fund a new categorical grant program. See Final Report, op. cit., pp. 67-69.

PAGE 165

Ch. 13Congressional Policy Considerations, Part / l 159 LEAA, Congress may wish to assign responsibility for technical assistance to the Attorney General, leaving to him the decision about which DOJ entity should carry out this function. Alternatively, Congress could explicitly provide that a technical assistance team shall be located within, for example, an existing DOJ office, the FBI, or perhaps a new Bureau of Criminal Justice Information (or the equivalent), should one be established to operate a national CCH system. Clarifying Title 28 Privacy and Security Regulations Title 28 regulations have now been in effect for about 7 years. Many States and localities have accumulated several years of experience in attempting to interpret and implement the regulations. The time may be ripe for Congress to initiate a review of title 28. One aspect of the review might be to commission a survey of the implementation of title 28 by States as of 1982. The results could then be compared with the OTA survey conducted in 1979* to see to what extent States are still making progress toward its full implementation. The results of this review could be reflected either in a title 28 revision to be promulgated by DOJ through normal rulemaking procedures, or in comprehensive legislation should Congress choose to move in this direction. Presumably the results would also serve as a valuable input to setting priorities for any funding and technical assistance that may be authorized by Congress. Improving Court Dispositio n Reportin g The OTA research has documented incomplete disposition reporting at the State and Federal levels.** Some criminal justice deciCFR 20; see ch. 6. *See ch. 9. **sW chs. 8 and 9. sionmakers see an improvement in court disposition reporting as a key prerequisite for a national CCH system. The improvement of court disposition reporting is closely tied to court reorganization. Minnesota is one of the States that have made the most progress in court reform. The State has implemented a State Judicial Information System that is achieving close to 100 percent court reporting. The system will track clients through the entire State court system, automatically record court dispositions, and feed them directly into the State repository. According to a Minnesota court official: 8 In Minnesota weve got three things going for us: (1) money; (2) the power of the (State) Supreme Court backed up by the funding and political support of the State legislature; and (3) the technical know-how to put together an adequate system. Given the constitutional separation of powers, Congress is limited in what it can do to encourage reform and reorganization in the judicial branch. However, as in the past, Congress could again authorize financial support and technical assistance to the States for use in developing court disposition reporting systems.* In addition, Congress could itself commission, or direct the Attorney General to commission, a survey of the current status and needs of the judiciary with respect to disposition reporting systems. The survey could attempt to identify on a State-by-State basis exactly what additional judicial information systems development would be necessary to support a national CCH system. Interview in 1979 with a Minnesota court official; reaffirmed in 1982. *During the 1970s, a portion of LEAA grant funds was Pro vided to the States for development of offender-based transaction systems intended to include court disposition reporting and provide input to CCH systems. See a related study by the National Center for States Courts, A Review of OBTS and CCH Program Requirements in the Judiciary, Williamsburg, Va., 1979.

PAGE 166

160 An Assessment of Alternatives for a National Computerized Criminal History System Range of Possible Nationa l CCH System Structure s Select a National CCH Syste m Should Congress wish to select a national CCH system, the five summarized in table 31 are representative of the entire range of possible structures.* All five could evolve from current systems. For example, full development of the AIDS file of Ident would constitute a national CCH repository when hooked up to NCIC (or other) communication lines to permit nationwide electronic access. The repository would include records on the roughly 21 million persons with arrests for serious offenses. If the repository was limited to records on violent and very serious offenders (FBI index crimes), then full development of AIDS with NCIC interconnection would be the equivalent of a national full record CCH system with a file size of about 9 million records. On the other hand, nationwide implementation of Phase I of the III development plan 10 would constitute a single-State/multi-State *see ch. 10 for detailed discussion. For a discussion of the Interstate Identification Index (111) development plan, see Federal Bureau of Investigation, 111 Background and Findings for July-Sept. 1981 Phase I Pilot l?rojec~ Dec. 4, 1981, pp. 111-135. Phase I of the plan involves decentralization of single-State offender records of six of the eight States currently fully participating in NC IC/CCH. Ibid., p. 116. CCH system with partial message switching. The index could also be implemented with no message switching. The central file would include records on Federal and multi-State serious offenders (about 6 million) and index entries (names and personal identifiers only; no criminal history information) on single-State serious offenders (about 15 million). Nationwide implementation of the second and third phases of the III development plan 11 would be the equivalent of a national index with partial message switching. The central file would include only records on Federal offenders (0.5 million), plus index entries on singleand multi-State offenders (20.5 million). Finally, an ask-the-network CCH system could result if both the AIDS and III programs were not implemented due to some combination of technical, fiscal, administrative, and/or political factors. In this case, an alternative would be to use only the existing NLETS network (or a privately offered network) to exchange criminal history information between and among the States and the Federal Government. Phases 11 and III of the plan would include index entries for single-State offender records from States not currently participating in NCICICCH, and decentralization of multi-State offender records of the fully participating States. Ibid. Table 31 .Comparison of Possible National CCH System Alternatives Single-State/ National full multi-State CCH National Index (Ill) National CCH record CCH system system with partial with partial or no Ask-the-network repository (FBI index crimes) message switching message switching system Central file size: Records . . 21 million 8.6 million 6 million 0.5 million 0.5 million Index entries . 15 million 20.5 million Central file content: Records . . All serious offenders Violent and very Federal and Federal offenders Federal (Federal, singleand serious offenders multi-State serious offenders multi-State) offenders Index entries . Single-State serious Singleand offenders multi-State serious offenders SOURCE: Office of Technology Assessment

PAGE 167

Ch. 13Congressional Policy Considerations, Part I l 161 Since 49 of the 50 States maintain a State driving forces behind efforts to keep the reidentification bureau and 45 of 50 States reordkeeping function decentralized so that duquire fingerprint submissions to this bureau plication between the Federal and State Govon arrest, 12 records on almost all offenders will ernments is minimal. OTA did not conduct the continue to be maintained by the States. detailed cost studies needed to provide specific Therefore, any records maintained in a nationcost estimates for the various alternatives. al repository will incur extra operational costs to the Federal Government for storing the recFrom an operational point of view, OTA found that three alternative systems could orals and to the States for updating the records. Cost control has thus been one of the most readily evolve from current systems: 1) a national repository; 2) a single-State/multi. State system with partial message switching; 12FB 1, ~cl c staff paper prepared for the NOV. 3-A, 1981 i meeting of the NCIC Advisory Policy Board Subcommittee on and 3) a national index with partial or no mesthe Interstate identification Index, Topic #6, p. 2. sage switching. Possible Impacts on th e Criminal Justice Process Importance of Complete Timely, and Verifiabl e Informatio n The results of OTA research indicate that the more complete, timely, and verifiable (i.e., backed up by positive identification) the CCH information, the more useful it would be. While even incomplete information has some value as a pointer to the sources of additional information, many criminal justice decisionmakers noted the problems caused when criminal history records lack important data (e.g., on dispositions and charges), arrive too late to be useful, and/or are not based on positive identification (i.e., fingerprints or State or Federal identification numbers which are in turn based on a fingerprint check). Available evidence indicates that improvement in criminal identification and criminal history information exchange by the Federal Government and most States is facilitated through computerization. During the period 1970-79, States with CCH record systems achieved significant increases in disposition reporting, while States with manual systems showed very little improvement. The operating experience of the Ident AIDS program and *See ch. 9. several State identification bureaus has documented the much shorter turnaround time possible with automated as compared to manual systems. A recently completed comprehensive Jet Propulsion Laboratory (JPL) study of AIDS concluded that full automation could reduce the overall Ident processing time for fingerprint checks from about 36 work days to about 3 hours. 14 In addition, the cost per fingerprint search would drop by about 50 percent.15 Turnaround time could be further improved through the use of high-quality facsimile electronic transmission. Two States, New York and Illinois, already make relatively extensive use of this technology. l6 The JPL study suggested that the five States with the largest volumes of fingerprint card submissions (and collectively accounting for about half of all submissions) could have facsimile interconneclsln~mation~ ASSmiatiOn for Identification, FUrJc~iOnd ~ quirement.s and Systems Development Plan for State Identification Bureaus: Executive Summary of FinaYngs and Recommendations, Utica, N. Y., October 1980, pp. 9-11. 14Jet propulsion Laboratory, FBI Ftigt?rprtit ~dentification Automation Study: AIDS 111 Evaluation llepor~ California Institute of Technology, Pasadena, Nov. 15, 1980, pp. 1-2 and 1-3; prepared for the U.S. Department of Justice, Federal Bureau of Investigation. K Ibid. International Association for Identification, State Identification Bureaus, op. cit., pp. 8-9. !.4

PAGE 168

162 An Assessment of Alternatives for a National Computerized Criminal History System tion with AIDS. JPL projects a 3-hour internal processing time for AIDS if fully implemented, and 8 hours for the residual manual files. 17 Thus, use of facsimile transmission could conceivably further reduce the total response time for fingerprint checks to the range of 5 to 10 hours. New York State responds to fingerprint inquiries submitted via facsimile within an average of 1 hour and 50 minutes, and within 3 hours 90 percent of the time. Response time for criminal history record checks could be even faster. In theory, response time for a full record CCH system would approach the 5-second or less range of the NCIC hot files. However, in order for the records disseminated to be complete and accurate the States would have to update the records in the central file on an almost continuous basis. One reason many States support a national index system is that they are unable and/or unwilling to update full records maintained in a central national CCH repository. Response time for a national index CCH alternative would likely be in the range of several hours or less. However, this will vary depending on the capability of individual States. The III pilot test with the State of Florida has demonstrated that response times of less than an hour are possible. In the 3-month pilot test (July-September 1981), the response time was less than 1 hour 86 percent of the time and under 15 minutes 64 percent of the time.18 However, Florida is fully committed to the III concept, has an advanced online State CCH system, and operates under a State open records law that simplifies record dissemination decisions. Thus, the average response time for all States could be longer. Nonetheless, for the 38 States with an online CCH file or automated name index (collectively representing about 95 percent of all criminal history record activity as measured by fingerprint card submissions), a response time Of several hours or less seems technically feasible. Actual response time will also depend on the priority Jet Propulsion Laboratory, ~111 ~tigerprti~ Identification, op. cit., p. 1-2. ISFBI, ~11 llackaou~d and Findings, op. cit., P 162. *% table 5, ch. 4. assigned by participating States, staffing and resource (including computer) limitations, and State policies on record dissemination and use. In the III Phase I test (February-April 1982), the response time was less than 15 minutes 85 percent of the time and under 1 minute 48 percent of the time. 19 Thus, it appears that the III response time could approach the response time achieved by individual States with on-line CCH files. In the OTA 50-State survey, for example, California, Colorado, Delaware, Minnesota, New Jersey, Ohio, and South Carolina reported CCH response times of, respectively, 5, 3, 15, 20, 10, 9, and 5 seconds. Response time was considerably longer for States with manual files. For example, New Hampshire, New Mexico, Pennsylvania, West Virginia, and Wyoming reported manual response times of, respectively, 4, 10, 3, 14, and 2 days. 2o The FBI and OTA used slightly different definitions of response time. OTA defined response time as the total time from receipt of a request for criminal history information by the State repository to receipt of the CCH record (response) by the requesting agency. The FBI defined response time as the time from inquiry by the requesting agency to the time of receipt of the summary CCH record by the requesting agency. Improving Record Quality Difficult But Possibl e While computerization can improve the response time of fingerprint and criminal record checks, improvements in record quality are more difficult to achieve. This is because high record quality depends on timely (and accurate) submissions from a large number of criminal justice agencies. Court disposition reporting is particularly important. The OTA 50State survey found that the average record update time for courts was about 64 days compared with 20 days for law enforcement agencies, 47 days for prosecutors, 24 days for proFB1, III Preliminary Findings for February-March 1982 Test, April 1982, p. 17. 2)OTA 50-State Survey conducted in 1979-80.

PAGE 169

Ch. 13Congressional Policy Considerations, Part 1 163 bation/parole agencies, and 16 days for correctional agencies. 21 Update time refers to the total time that elapses from the occurrence of a criminal history event (e.g., arrest or court disposition) to the updating of the subjects criminal history record. The average update time for the courts ranged as high as 1 year. Many States did not know what the average time was. Nonetheless, the experience of three States with online CCH filesCalifornia, Minnesota, and North Carolina-shows that improvement is possible. In California, partly as a result of field audits and local training efforts, the statewide felony disposition reporting rate has increased from 66.6 percent in 1978 to 70.8 percent in 1980. The statewide superior court disposition reporting rate has increased from 69.1 percent in 1978 to 79.3 percent in 1980. 22 In Minnesota, a State Judicial Information System was implemented on July 1, 1980. This, coupled with mandatory reporting forms and followup from the office of the State Supreme Court administrator, has resulted in essentially 100 percent final disposition reporting. 23 Roughly 80 percent of all final dispositions are reported within 1 to 2 days after the disposition occurs, and almost all are reported within 5 days. The dispositions are checked for accuracy and completeness and then forwarded to the State CCH repository. However, due in part to problems in establishing positive identification, the actual disposition level in the State CCH file is about 85 percent and has remained essentially constant over the last 3 years. 24 The State repository can update records only if the update information is based on positive identification. In North Carolina, the State identification bureau criminal history file was automated in 1976. Since that time, the disposition reportibid. Based on data in Feb. 2, 1982, letter and enclosures from Fred H. Wynbrandt, Assistant Director, Criminal Identification and information Branch, California Department of Justice. Feb. 12, 1982, conversation with James Rebo, office of the Minnesota State Supreme court Administrator, 24 Feb. 11, 1982, conversation with Clayton Mellem, Criminal Justice information System, Bureau of Criminal Apprehension, Minnesota State Department of Public Safety. ing rate has increased from 26 percent in 1975-76 to 48 percent in 1978-79, and is projected at 61 percent for 1981-82. As of December 31, 1981, the North Carolina Police Information Network showed a court disposition for about 75 percent of the 206,683 arrest events included in the system at that time. This improvement reflects considerable effort to update records, including the use of interns in the three largest North Carolina court districts (Wake, Guilford, and Mecklenberg) to assist in the submission of dispositions. 25 On January 1, 1982, a new State law went into effect requiring submission of felony arrest fingerprints and dispositions to the State identification bureau. 26 This will presumably contribute to further improvement in disposition reporting. Thus, available evidence indicates that strengthening State/local criminal history systems and court disposition reporting systems is a prerequisite to further improving record quality, regardless of the national CCH system structure. Particularly important are efforts to: 1) upgrade court administration; 2) establish standardized (and perhaps even codified*) court reporting procedures; 3) improve the coordination between judicial and other criminal justice agencies (especially law enforcement) responsible for timely record update actions; 4) strengthen field audits of reporting procedures and record quality; and 5) increase funding and technical assistance to implement computer-based systems where appropriate. To the extent that a national CCH system provides more complete, timely, and verifiable information than is presently available, the system would improve the functioning of the criminal justice process. Based on the surveys and research reviewed in chapter 11, the most Based on data in Jan. 29, 1982, letter from WiUiarn C. Corley, Director, Police Information Network, State of North Carolina Department of Justice. s General Assembly of North Carolina, House Bill 118, An Act to Require the Reporting of Complete and Accurate Criminal Histories to the State Bureau of Investigation, July 8, 1981. *Minnesota has promulgated a standard criminal complaint form which must be used by prosecuting attorneys and court clerks.

PAGE 170

164 l An Assessment of Alternatives for a National Computerized Criminal History System significant improvements are likely to be in the areas of criminal investigations, police booking and intake, pretrial release and bail decisions, and presentence investigation reports. Since criminal history information is used throughout the criminal justice process, better information is also important to sentencing, correctional, and, to a lesser extent, probation and parole decisions. Shifting Preferences on System Structur e In general, over the last 3 to 4 years many State and Federal criminal justice officials have shifted their support from the singleState/multi-State alternative to the national index (or III) concept. This shift is illustrated in table 32. In the 1982 OTA follow-up survey, officials from about two-thirds of the States indicated a clear preference for the III concept, with officials from most of the other States either actively considering III or seeking further information on which to base a decision. However, many States, even some of those strongly supporting III, noted a variety of implementation problems that might preclude their participation, in some cases for years. Many of these officials also support the concept of a National Fingerprint File (NFF), considered to be an integral part of III and which would be limited to fingerprint cards and related personal descriptors on each criminal offender. The NFF would contain no arrest or disposition data. It would perform the technical fingerprint search to establish positive identification or nonidentification based on fingerprint cards received from State identification bureaus or Federal agencies. It would also assign FBI identification numbers, and could enter identification data into III. The NFF concept is predicated on single-source submission policies. That is, only one agency per State would be authorized to submit fingerprint cards, and submission of only one fingerprint card per subject per State would be permitted. 27 OTA surveyed the States with respect to single-source fingerprint card submission and found that, as of August 1982, 18 States had implemented single-source submission (compared to 17 in a September 1981 FBI survey) and 4 more States had scheduled a late 1982 implementation for a total of 22 States. Officials from about one third of the other States indicated that implementing single-source submission could be difficult due to a potential work overload, staff and funding shortages, local agency resistance, and/or privacy concerns. Nonetheless, despite the movement towards agreement on III and NFF as the basic national CCH system concept, questions raised with respect to, for example, policy control, noncriminal justice access, record quality, and system accountability, have yet to be resolved. These and other issues are discussed in the next chapter. 27 See SEARCH Group, Inc., Essential Elements and ActiorIs for Implementing a Nationwide Criminal History Program, Sacramento, Calif., February 1979; and NCIC Advisory Policy Board, A Proposed Chcept for a Decentralized Criminal History Record System, Apr. 12, 1978. See also NCIC Advisory Policy Board, Phase I Test Interstate Identification Index: Report of the III Evaluation Ckmnnittee, June 1982, p. 4. Table 32.Shifting Preferences of Federal and State Criminal Justice Officials for a National CCH System System alternative Single-State/multi-State National Index (Ill) FBI . . . . . . . . . (throughout 1970s) (since 1980) NCIC Advisory Policy Board . . . . (early mid-1970s) (since 1978) States (DOJ survey of 10 States) . . . (1978) States (OTA survey a ) . . . . . 11 States (1979) 24 States (1979) SEARCH Group, Inc.. . . . . . (1970-78) (1979 to present) NLETS, Inc. . . . . . . . (1981) U.S. Attorney Generals Task Force on Violent Crime. . . . . . . (1981) aT~O states preferred ask.the.network; one preferred a national repository; and four indicated no preference. Forty-two States responding. SOURCE: Office of Technology Assessment, Federal Bureau of Investigation, Department of Justice, SEARCH Group, Inc and National Law Enforcement Telecommunications System

PAGE 171

Chapter 1 4 Congressional Policy Considerations, Part II

PAGE 172

. . Contents Page Chapter Summary. . . . . . . . . . . . 167 Policy Control . . . . . . . . . . . . 167 File Size and Content . . . . .. ~. . . . . . 167 Record Quality. . . . . . . . . . . . 167 Non criminal Justice Access . . . . . . . . . 168 Oversight and Audit . . . . . . . . . . 168 Public Participation. . . . . . . . . . . 168 Comprehensive Legislation . . . . . . . . . 168 III Development Plan . . . . . . . . . . 168 AIDS/CCH Consolidation. . . . . . . . . . 168 Private Carrier Role in a National CCH System.. . . . . 169 Policy Control . . . . . . . . . . . . 169 File Size and Content . . . . . . . . . . . 173 Record Quality . . . . . . . . . . . . 174 Noncriminal Justice Access . . . . . . . . . 176 Oversight and Audit. . . . . . . . . . . . 177 Public Participation . . . . . . . . . . . 179 Comprehensive Legislation . . . . . . . . . 180 Access, Review, and Challenge Procedures . . . . . . 182 Criminal Penalties . . . . . . . . . . . 182 Privacy Standards. . . . . . . . . . . . 182 Technical Accountability . . . . . . . . . 183 Uniform Crime Codes and Record Formats . . . . . . 184 III Development Plan . . . . . . . . . . 184 AIDSICCH Consolidation Plan . . . . . . . . . 186 Private Carrier Rolein NationaI CCH System . . . . . 188 TABLES Table No. Page 33. 1979 Preferences of State Repository Personnel for CCH System Structure and Policy Control . . . . . . . . 172 34. File Size Options for a National CCH System Based on Composition of Ident File . . . . . . . . . . . . 173 35.Subject Areas Relevant to Comprehensive Legislation on Criminal Justice Information Systems . . . . . . . . 181 36. Possible File Structure of a Reorganized National Crime Information Center. . . . . . . . . . . . . . 187

PAGE 173

Chapter 14 Congressional Policy Considerations, Part II Chapte r As noted in chapter 13, the emerging consensus among Federal and State law enforcement and criminal history record repository officials supports the national index concept known as the Interstate Identification Index (III). However, full implementation of III (or any other national computerized criminal history (CCH) system) would require resolution of a number of issues that warrant congressional attention to ensure that beneficial impacts are maximized and adverse impacts are controlled or minimized. Policy Contro l Considerable debate has focused on which agency or organization should have direct policy control over a national CCH system. Suggestions include a consortium of States, a broadened and strengthened National Crime Information Center (NCIC) Advisory Policy Board (APB), an independent board, and/or the FBI. For example, a broadened and strengthened APB could include greater representation from the prosecutorial, judicial, correctional, and public defender sectors of the criminal justice community that at present, and could include an advise and consent role, at least with respect to State and local participation in a national system. There are many other possibilities, but the key issue is how to devise a mechanism that will effectively represent the interests of the diverse users of a national system, and afford them a strong and possibly controlling policy role. File Size and Content Under the III concept, the national index would include only names and identifying inSummar y formation (e.g., height, weight, social security number, and State and Federal criminal identification numbers). Proposals have been made to limit the index to entries on violent or very serious offenders, that is, for crimes included in the Federal Bureau of Investigation (FBI) crime index. However, this would exclude entries for drug, weapons, drunk driving, and other offenses generally considered to be serious but not included in the FBI crime index. At the other extreme, a totally unrestricted index could include entries on as many as 36 million persons. Other national index issues include the need for policies on limited retention periods for some entries, and on the handling of juvenile offender records. Record Qualit y With a national index, the FBI would no longer maintain non-Federal records, and the problems of record quality in Ident and NCIC/ CCH would be reduced. However, the quality of records maintained by the States, as well as the quality of any index based on those records, would still be a matter of concern. Record quality could be strengthened by tightening the disposition reporting requirements and/or requiring confirmation of records lacking disposition data with the originating agency prior to any dissemination. In the opinion of some, the latter requirement would be costly and impractical. The progress made by many States in recent years indicates that improved disposition reporting is possible, but continued record quality improvement would require a significant further commitment measured in manpower, dollars, and system improvements at the State and local levels. 167

PAGE 174

168 l An Assessment of Alternatives for a National Computerized Criminal History System Noncriminal Justice Acces s Significant noncriminal justice use of Federal and State criminal history record systems, coupled with widely varying State statutes defining authorized users and State policies on sealing and purging, has generated concern about control of access to criminal history records. Noncriminal justice access to a national index could be prohibited, although this would conflict with many Federal and State laws. Nonc riminal justice access could be permitted, but only under stronger Federal guidelines than presently exist. A dual index could be established, one for criminal justice use and a second for noncriminal justice use, perhaps with the latter based on disposition or conviction information only. Even under the status quo, access to a national index would require complicated safeguards (which are technically feasible with a computer-based system) to be consistent with the wide variety of existing State laws and regulations, and would require some means to resolve conflicts among State laws, and between Federal statutes and Executive orders and State laws. Oversight and Audi t The purposes of new oversight mechanisms would be to help assure Congress, the public, and others that a national index (or any other national CCH system) is operating within the boundaries of law and regulation, and to help identify any problems that may emerge. Oversight is closely linked to system audit. Several possibilities have been suggested. First, Congress could require an annual management report on the operation of a national CCH system. Second, Congress could require periodic audits of Federal and State CCH files to help ensure compliance with whatever system standards may be established. To keep costs down, the audits would presumably be conducted by sampling Federal and State files on a rotating and perhaps unannounced schedule. Any Federal audit authority, whether by the General Accounting Office (GAO) or some other body, would appear to require new Federal legislation and/or regulations. Public Participation NCICS APB is the only direct avenue of public participation in the governance of the existing NCIC/CCH system. However, at present the APB does not include representation from the general public or from public defenders. Public defenders feel strongly that they should be represented on any policy board established for a national CCH system and that defense interests should have access to that system. The experience of Alameda County, Calif., where public defenders are considered to be part of the criminal justice community, has been that public participation in oversight can help ensure accountability of criminal justice record systems and can be beneficial in terms of system performance. Comprehensive Legislatio n Legislation represents one of the strongest measures to provide Federal direction and ensure accountability and control. It could provide explicit authority for a national index or other national CCH system, and include statutory guidelines for its operation and use. In addition to the areas discussed above, legislation could establish access, review, and challenge procedures; criminal penalties; privacy standards; funding for computer-based user audits and court disposition monitoring procedures; and uniform crime codes and criminal history record formats. III Development Plan In order to develop important additional data from the III test now underway, Congress may wish to consider whether the plan should be revised so that: 1) some or all of the participating States can be tested with no NCIC message switching as well as with partial message switching (known as automatic inquiry referral); and 2) record quality research can be conducted. AIDS/CCH Consolidatio n At present, the Ident/Automated Identification Division System (AIDS) and NCIC/CCH

PAGE 175

Ch. 14Congressional Policy Considerations, Part II l 16 9 files duplicate each other to a significant and growing extent. Any AIDS/CCH consolidation is likely to have a significant impact on the cost of FBI criminal history and identification services and could be an integral part of a national CCH system. Congress may wish to request the preparation of several alternative consolidation plans, including the possible creation of a new National Criminal Information and Identification Division of the FBI, which would combine Ident, NCIC, and related activities. Congress may also wish to examine the pros and cons of shifting management of a national CCH system to a new bureau within the Department of Justice (DOJ) or elsewhere. Private Carrier Role in a National CCH Syste m Congress may wish to review the role of private communication carriers in a national CCH system. Privately offered nationwide data communication networks using satellite as well as landline transmission, and providing security measures such as data encryption, may offer significant benefits over lines currently used by NCIC and the National Law Enforcement Telecommunications System (NLETS). Policy Contro l Considerable debate has focused on which agency or organization should have direct policy control over a national index (or other national CCH system). There is general agreement that any national CCH system would require some degree of policy control to ensure that compatible message formats, operating protocols, and the like would be used. This would be essential even for an ask-the-network system resembling NLETS. The policy control requirements would be correspondingly greater for systems with a centralized file or index. Over the years, various proposals have suggested that a national CCH system be controlled by the Law Enforcement Assistance Administration (LEAA), by a single State, by a consortium of States, by NCIC, by Ident, by a new division within the FBI, by a new bureau within DOJ, or by a new independent Federal agency or board. For example, SEARCH Group, Inc., has in the past advocated that policy control be vested in a consortium of States. SEARCH believed that a policy advisory board, similar to the current NCIC/APB, would not be sufficient even if broadened to give greater representation to the States. At the same time, SEARCH supported the role of the FBI in creating and maintaining III, and also recognizes that Federal agencies would be participating as well. In light of the constitutional and practical difficulties of State control over Federal agencies, SEARCH has recommended a policy advisory role for the States with respect to Federal participation in a national CCH system, but has maintained that States must have policy control over State participation in such a system. What is necessary is a consortium of all the States whose members are responsible to the Governors. 1 Another policy control alternative is to strengthen and possibly broaden the NCIC/ APB. Officially, the role of NCIC/APB is limited to reviewing NCIC issues and making appropriate recommendations to the FBI Director. 2 In practice, at least on some issues, NCIC/APB has had a significant influence on FBI decisions. Nonetheless, as early as 1978 NCIC/APB has sought to modify its charter to include a formal advise and consent relaISee SEARCH Group, Inc., A Framework for Constructing an Improved National Criminal History System, Sacramento, Calif., April 1978; and SEARCH Group, Inc., Essential Ele ments and Actions for Implementing a IVationwide Crirninaf History Program, Sacramento, Calif., February 1979. 2 FB 1, Bylaws for the NCIC Advisory Pohcy Board and Regional Working Groups, Dec. 11, 1980, p. 2.

PAGE 176

170 An Assessment of Alternatives for a National Computerized Criminal History System -- .. tionship to the FBI Director on matters relating to NCIC and especially 111. 3 NCIC/APB in the past has agreed with SEARCH that policy control over State participation in III should be vested directly in the States, but apparently has disagreed on what should be the instrument of control. NCIC/APB is currently composed of 20 elected and 6 appointed members. The 20 elected members all represent State and local law enforcement State (9 from State police or patrols and 7 from State bureaus of identification or the equivalent) and 4 local (all city or county chiefs of police). The six appointed members are designated by the FBI Director and must include two members each from the judicial, prosecutorial, and correctional sectors. Thus, at present, NCIC/APB is composed of about three-quarters law enforcement and one-quarter other criminal justice representatives. If a more even balance between law enforcement and other criminal justice sectors is desired, the size of NCIC/APB could be expanded, for example to 40 with 20 members elected from law enforcement and 20 from other criminal justice sectors. Alternatively, the current size could be maintained, but with 13 law enforcement members and 13 other criminal justice members. A 1978 DOJ survey of 10 States found that many, although not all State officials, were critical of the composition of the NCIC/APB. Indeed, some State law enforcement officials acknowledged that CCH was of primary interest to prosecutors, judges, correction officials, etc., and was of only limited interest to law enforcement agencies per se. 4 Yet another alternative is to have two policy control groups-NCIC/APB and an independent board (or a consortium of States). NCIC/ APB could have an advise and consent role with respect to the NCIC hot files that are 3 Minutes of the Apr. 12, 1978, meeting of NCICS APB, A Proposed Concept for a Decentralized Criminal History Record System, p. 19. U.S. Department of Justice, Representative Viewpoints of State Criminal Justice Officials Regarding the Need for a Nationwide Criminal Justice Information Interchange Facility, March 6, 1978, reprinted in U.S. Congress, Office of Technology Assessment, A Preliminary Assessment of the National Crime Information Cknter and Computerized Criminal History l?r~ gram, Washington, D. C., December 1978, p. 71. used primarily by law enforcement agencies, and an independent board could have a similar role with respect to a national index or other national CCH system. The role of NCIC/APB or an independent board could be extended to include a national fingerprint repository, which is generally viewed as an integral component of the national index (III) concept. Various representatives of the judicial and prosecutorial communities have forcefully argued that any viable national CCH system must have their active involvement and cooperation, since judges and prosecutors would be major users of the system and central to achieving improved disposition reporting. This may require a much larger role (and representation) for judges and prosecutors on the NCIC/APB and/or an independent policy board than has historically been the case at either the Federal or State levels. The public defender community also believes that it has a legitimate stake in any national CCH system and deserves some representation on any policy board. Several times over the last 12 years proposals have been advanced to vest policy control in an independent board. As early as September 1970, the Office of Management and Budget recommended the establishment of a strong policy control board that would report directly to the U.S. Attorney General. The board was to include officials from the FBI, LEAA, and the States and represent all elements of the criminal justice community. The Board was to be structured so that the States would have an equal voice with the Federal Government. This proposal grew out of a conflict between LEAA and the FBI over control of the original CCH program (then known as Project SEARCH (System for Electronic Analysis and Retrieval of Criminal Histories) and funded by LEAA). In December 1970, the Attorney General assigned management responsibility for the CCH program to the FBI. 5 See Donald A. Marchand, et al., A History and Backround Assessment of the National (lime In fonnati;n C%nter n> Computerized Criminal History System, Bureau of Governmental Research and Service, University of South Carolina, June 1979, sec. III, The CCH Program: Its Origin and History, pp. 78-86.

PAGE 177

Continuing conflict between LEAA (and various States) and the FBI in part led to comprehensive legislative proposals in 1974 that included an independent Federal Information Systems Board. The board was to be responsible for the operation of a national CCH system and for promulgation and enforcement of regulations on the use of such a system. In addition, the board was to have an advisory committee with one representative from each State who would serve at the pleasure of the Governor. 6 In 1974 congressional hearings, LEAA and Project SEARCH supported the independent board approach, while the FBI opposed placing policy control in either an independent board or LEAA. The FBI testified that the single-State/multi-State alternative advocated by NCIC was preferable to the national index (pointer index) and that NCIC was best equipped to manage and operate a national CCH system. 7 As discussed in chapter 6, in large part because of these conflicts, efforts to enact comprehensive legislation were not successful. The 1978 DOJ survey of selected States (conducted from November 1977 through February 1978) found that many State officials expressly or implicitly recognized that in the longer term a Federal agency other than the FBI could provide the services expected of a national CCH system, although there was a clear consensus that the FBI should continue to provide such services in the foreseeable future. State officials were critical of the fragmented responsibility for criminal history records within the FBI and the organizational separation between Ident and NCIC/CCH 8 Coincidentally, on June 1, 1978, a presidential decision memorandum (PRM) was prepared for President Jimmy Carter by the President Reorganization Project on Federal Law Enforcement. Among other things, the PRM recS. 2963, The Criminal Justice Information and Control and Protection of Privacy Act of 1974. Ibid., pp. 97-100. Ibid., pp. 101-104. DOJ, Representative Viewpoints, op. cit., p. 70. Ch. 14Congressional Policy Considerations, Part II l 171 ommended that Ident and NCIC (as well as statistical programs such as the Unifor m Crime Reports) be combined into a ne w Bureau of Information and Statistics tha t would be organizationally separate from the FBI, although still a part of DOJ. The reorganization plan was never enacted 9 In mid-1979, OTA conducted a survey of State repository personnel in 42 States. At that time, repository officials from 22 States felt that the FBI was the logical place to locate management responsibility for a national system. However, seven conditioned their support for the FBI on increased State and user participation in system policymaking. The present NCIC/APB was not regarded as being sufficiently representative or authoritative, a t least with respect to the CCH program. Thus, these officials proposed that the board be reconstituted with participants from all States and be given policymaking (as opposed to solely advisory) authority. Officials from seven States indicated a preference for vesting CCH policymaking authority in an independent board or council. Some noted that there is no compelling necessity to locate a national index within the FBI. Rather, the participating States could again follow the NLETS model and create their own management and policymaking machinery outside of the Federal Government Four States indicated no preference, one expressed an equal preference for the FBI or NLETS, another indicated an equal preference for SEARCH Group or NLETS, and seven did not respond to the question on policy control. Table 33 illustrates the division of opinion that existed in 1979 among State repository Marchand, et al., History and Background op. cit., pp. 151-153. Steven W. Hays, et al., An Assessment of the Uses of Information in the National Crime Information Center and G)mputerized Criminal History Program, Bureau of Governmental Research and Service, University of South Carolina, October 1979, sec. IV, Inter-State and Intra-State Uses of NCIC and Identification Information, pp. 177-180.

PAGE 178

172 l An Assessment of Alternatives for a National Computerized Criminal History System Table 33. Preferences of State Repository Personnel for CCH System Structure and Policy Control a Policy Control FBI with Independent No FBI or SEARCH No System structure FBI revised board board preference NLETS or NLETS response Totals National repository . 1 1 Single-State/multi-State . 8 1 1 1 11 National index . . 6 6 6 1 1 4 24 Regional systems. . . 1 1 Total decentralization . 1 1 No preference . . . 4 4 Totals . . . . 15 7 7 4 1 1 7 42 a Forty-two states responding by telephone and/or mail to an OTA survey conducted in mid-1979. See app. C for a list Of State repository officials responding. SOURCE: Office of Technology Assessment personnel with regard to policy control and system structure for a national CCH. As of mid-1979, officials from States that had contributed records to NCIC/CCH were split between those preferring a national index versus a single-State/multi-State system, and between those preferring control by the FBI versus an independent board. Officials from States preferring a national index were almost evenly split among those favoring policy control by the FBI, by the FBI only if the APB is broadened and strengthened, and by an independent board. There also appeared to be a growing belief on the part of some State repository personnel that the FBI should not maintain any nonFederal criminal records, either automated or manual. They expressed the view that the FBI only needs to keep one fingerprint card for each offender, not a rap sheet. Some repository personnel also expressed a sense of frustration over the absence of a clear direction at the Federal level, as evidenced by the inability of the Federal Government to formulate a clear and consistent position on a national CCH system, at least up to that time. During 1980 and 1981, consensus on a system structure was substantially reached, at least among those segments of the law enforce ment and criminal justice community that are involved most directly. NCIC/APB, NLETS Board of Directors, and SEARCH Group have all endorsed the national index concept which, if fully implemented, would mean that all State records would be maintained by the States themselves. Only Federal records and an identification index, known as III, would be maintained at the national level, along with a national fingerprint repository. The results of the III pilot project (as well as the Phase 1 test) suggest that a national index is feasible. The December 1981 NCIC/APB action to accelerate the testing of the III on the recommendation of NCIC staff and the III Subcommittee, confirms the strength of that consensus.ll In June 1982, NCIC/APB endorsed plans to proceed with Phase 2 of III. However, the question of policy control has yet to be resolved. Further, while the U.S. Attorney Generals Task Force on Violent Crime has endorsed III, the task force also supported renewed consideration of a national repository and/or full message switching if III does not prove to be feasible. 12 At the Dec. 9-1o, 1981, meeting of the NCICS APB, they voted Un anirnously to combine Phases 2 and 3 of the Interstate Identification Index development plan. This means that if decentralization of single-State records of current CCH participating States proves successful, the next step would include testing of full decentralization in at least some States. *Attorney Generals Task Force on Violent Crime, Final Repor6 U.S. Department of Justice, Aug. 17, 1981, pp. viii and 67-69.

PAGE 179

Ch. 14Congressional Policy Considerations, Part II l 173 File Size and Content As noted earlier, both the FBI and NCIC/ APB (as well as SEARCH Group, Inc.) are now proposing that a centralized national index be created containing names and other identifying information such as personal identifiers (height, weight, race, etc.), social security numbers, State identification numbers, and Federal FBI identification numbers There are essentially four options in terms of the number and types of persons that could be listed in the index file. The first option is a so-called narrowband index, which would include entries for only certain types of offenders; for example, multi-State violent offenders. The estimated size of such a narrowband index would range from about 1 million to at most 2 million or 3 million entries, as shown in table 34. For example, the OTA record quality research on the Ident criminal history file found that 30.4 percent of individuals arrested were multi-State offenders (counting Federal offenders as multi-State), and about 11 percent of arrests were for violent crimes 13 Thus multi-State violent offenders would tota l about 0.7 million persons (i.e., 11 percent violent offenders of the 30.4 percent multi-State offenders in the 21 million person Iden t criminal file; or O.11 X 0.304 X 21 million persons) A second option is a so-called mediumband index that would include very serious property crimes as well as violent crimes such as are included in the FBI crime index. For example, multi-State FBI index offender s (which include violent offenders) would total about 2.6 million persons (41 percent FBI crime index offenders 14 of the 30.4 percent multi-State offenders in the Ident criminal file; 1 5 or 0.41 X 0.304 X 21 million persons). One 13 Based on 1979 OTA record quality research. [ncludes murder, manslaughter, forcible rape, robbery, aggravated assault, burglary, larceny, and motor vehicle theft. Ii]ncludes 11 percent violent crimes and 30.1 percent ProPerty crimes. Of the 168 records sampled from the Ident file and subsequently verified, violent crimes accounted for 113 arrests out of a total of 1,029 and property crimes accounted for 310 arrests out of 1,029. See ch. 3, table 1. Table 34. File Size Options for a National CCH System Based on Composition of Ident File Number of records in index Single-State and Only multi-State multi-State Index size option offenders offenders .. Narrowband (violent offenders) . 0.7 million 2.3 million Mediumband (FBI Crime Index offenders) . 2.6 million 8.6 million Restricted broadband (serious and significant offenders) . 6.4 million 210 million Unrestricted broadband (all offenders) . 10.9 million 36.0 million a All Numbers approximate. Based on arrests, not convictions, and includes Federal as well as State offenders Assumes 304 percent multl-State offenders SOURCE Off!ce of Technology Assessment serious problem with a mediumband index is that it would exclude entries for drug, weapons, financial (e.g., bribery or fraud), escape or unlawful flight, drunk driving, and other offenses generally considered to be serious but not included in the FBI crime index. A third option has been termed a restricted broadband index that would include entries for all persons arrested and/or convicted of one or more serious and/or significant crimes. The current Ident criminal file, with criminal records on 21 million individuals, is a good example of a restricted broadband file. The Ident file is restricted in the sense that nonserious offenses (e.g., disturbing the peace, drunkenness) are excluded.16 Finally, an unrestricted broadband index in theory could include entries on as many as 36 million persons the estimated number of people with criminal offenses ranging from minor misdemeanors to serious felonies.17 The current Ident file probably contains 5 to 10 percent nonserious arrests entered prior to 1975 when 28 CFR $20.32 went into effect. See Lynne Eickholt Cooper, et al., An Assessment of the Social Impacts of the National Crime Information Cknter and Computerized Criminal History Program, Bureau of Governmental Research and Service, University of South Carolina, October 1979, sec. I, Background on the Criminal Justice Process and Criminal History Records, pp. 52-77 and especially p. 63 and 76.

PAGE 180

174 l An Assessment of Alternatives for a National Computerized Criminal History System File size also depends on record retention policies. For example, at present NCIC/CCH maintains records indefinitely, with the exception of court-ordered deletions and expungements and records for individuals over 80 years of age. By comparison, the California State CCH system has established a 5-year retention period for records of misdemeanor arrests not resulting in a conviction or for which no disposition was received, and a 7-year retention period for misdemeanor convictions and felony arrests not resulting in a conviction or for which no disposition was received. Also, felony convictions are purged for individuals over age 70 with no arrests since age 60. Through these policies, California is able to keep the file size under better control, remove criminal record information that has become outdated and has little value to the criminal justice community, and protect individual privacy rights. Retention policies for entries in a national index may need to be considered for similar reasons. One other file content question relates to juvenile offenders. States vary in their handling of criminal history records on juvenile offenders. In the OTA 50-State survey, 11 States indicated that criminal history information on juvenile offenders is maintained in the State repository, while 38 States reported that juvenile offender information is not maintained (except, in some States, when juvenile offenders are tried as adults). Policies for entry and retention of juvenile offenders in a national index need to be established to prevent index use that might conflict with State laws on juvenile offenders. Record Qualit y Record quality is one of the most important aspects of any CCH alternative, yet it is one of the most difficult to achieve since all alternatives depend largely on information (whether name and identifiers, summary criminal history record, or full record) originating at the State and local levels. As discussed in chapter 8, a major problem with State CCH files is a lack of court disposition information. The implementation of procedures to ensure reporting of dispositions was found, not surprisingly, to be highly variable at the State level, as summarized in chapter 9. Ident currently requires disposition reporting within 120 days (after the disposition has occurred), 18 but Ident has found it very difficult if not impossible to enforce this requirement. OTA found that, as of mid-1979, about 30 percent of the arrest events in records disseminated from the Ident manual file lacked information on dispositions that had taken place and were listed in local records. Indeed, there is no easy way even to determine how many rap sheets in the Ident manual file lack disposition data, other than by a special audit as was conducted by OTA. AIDS is developing the capability to determine what percentage of records lack disposition data. NCIC/CCH also requires disposition reporting within 120 days. 19 Based on the OTA sample, however, as of mid-1979 about 27 percent of the arrest events in records disseminated from the CCH file lacked a court disposition that was recorded at the local level. NCIC/ CCH does conduct 418 computer edits to check for routine errors, periodic quality checks (including comparison with manual and/or automated Ident records), and automatic computer listing of criminal history records containing an arrest without a disposition if the arrest is 1 year old or more. These listings are used by participating States in followup actions to obtain the final dispositions. At the request of a participating State, NCIC/ CCH will prepare a computer tape of all records entered by that State, and will conduct offline searches to identify missing disposition CFR \ 20.37. g Ibid.

PAGE 181

Ch. 14Congressional Policy Considerations, Part II l 175 data. 2o However, the only sanction available to the FBI for noncompliance with the 120-day reporting requirement is a cutoff of NCIC/ CCH service, 21 which the FBI finds to be counterproductive and therefore does not exercise. Current NCIC/CCH operating procedures also require State control terminal agencies to conduct systematic audits of record quality to insure that files have been regularly and accurately updated. Where errors or points of incompleteness are detected, the control terminal shall take immediate action to correct or complete the NCIC/CCH record as well as its own State record. 22 However, the OTA 50State survey found that as of 1979 about three-quarters of the States had never conducted a record quality audit of either computerized or manual criminal history record systems. The existing NCIC/CCH standard might be further refined to require that systematic audits be conducted at regular intervals as a condition of participation. States could be required to provide documented and independently verified certification that record quality audits were indeed being conducted, and that appropriate followup actions were being taken. With a national index, the FBI would no longer maintain non-Federal records, and the problems of record quality in Ident and NCIC/ CCH would be reduced. But the quality of records maintained by the States would still be a matter of concern, as would be the quality of any index based on those records. Current Federal regulations hold agencies contributing records responsible for keeping the information complete, accurate, and current. Source agencies are required to submit dispositions to their State criminal history record repository within 90 days after the disposition has occurred. 20 See Oct. 22, 1981, statement by William A. Bayse, FBI Assistant Director, before the Subcommittee on Civil and Constitutional Rights, House Judiciary Committee, pp. 6-13. 2 CFR $20.38. FBI, IVCIC Operating kfanua~ pt. 10, p. 15. *See ch. 6. 23 28 CFR $ 20.21(a)(l). Thus, there is a 90-day period during which individuals may be subject to criminal justice decisions based on incomplete records that are, nonetheless, in compliance with Federal regulations. The regulations do require criminal justice agencies to query their State repository for the most up-to-date disposition data prior to dissemination of a record, except in cases where the State repository is technically incapable of responding within the necessary time period. 24 The record quality of a national CCH system could be strengthened by tightening the disposition reporting requirements and/or requiring confirmation of CCH records lacking disposition data with the originating agency prior to dissemination of the record. For example, the basic disposition reporting requirement could be reduced from 90 days to perhaps 30 days or even less. Some States with fully computerized State CCH systems and effective centralized reporting (where local agencies report all data to the central State repository) could probably comply with a tightened standard. In the OTA 50-State survey, computerized States such as Florida, Kansas, Minnesota, Oregon, Texas, Utah, Virginia, and Washington indicated that an average update time (time from occurrence of a criminal history event, e.g., arrest or disposition, to the updating of the criminal history file) of 30 days or less had already been achieved. For example, Minnesota indicated an average update time of 2 to 5 days, Virginia 7 to 21 days, and Washington 2 to 10 days. States with manual or semiautomatic CCH systems that lack effective centralized reporting might have considerable difficulty. Substantial system development, and in some cases enactment or revision of State law, would be required to bring these States into compliance with a tighter disposition reporting standard. Nonetheless, in the OTA 50-State survey, noncomputerized States such as New Hampshire, Vermont, Wisconsin, and Wyoming indicated that they too had achieved an average update time of 30 days or less. Ibid.

PAGE 182

176 l An Assessment of Alternatives for a National Computerized Criminal History System The OTA 50-State survey update found that as of 1982, 13 States indicated that less than half of all dispositions were ever reported, regardless of the time period. Computerized States indicated a significantly higher disposition reporting rate overall (70.6 percent) than did the noncomputerized States (56.3 percent). 26 It seems evident that a tighter disposition reporting standard could be established by law or regulation, but its full implementation would require a significant further commitment measured in manpower, dollars, and system improvements at the State and local Office of Technology Assessment, 50-State survey conducted in 1979-80. See ch. 9, table 19. Ibid. levels. The progress made by many States in recent years indicates that improved disposition reporting is possible. On the other hand, requiring confirmation of a CCH record lacking disposition (or other important) data prior to dissemination would be feasible but could be very time-consuming, to the point, in the opinion of some, of strangling the system. Confirming an out-of-State record-whether manual or computaized-ap parently can be an onerous task. However, NCIC/CCH already performs a large number of automated quality checks. And it is possible that disposition checks could also be automated, particularly with originating agencies in States with an online CCH file. Confirmation of CCH records could be required by law or regulation. Noncriminal Justice Acces s OTA has found that there is a significant amount of noncriminal justice use of criminal history record systems at both the State and Federal levels. For example, as of 1982, roughly 15 percent of all requests to State CCH systems were for nonc riminal justice purposes. As of 1981, about 53 percent of requests to Ident were from noncriminal justice users (30 percent from Federal noncriminal justice agencies and 23 percent from State and local noncriminal justice users). Most of the noncriminal justice use is for employment or licensing purposes. At the Federal level (for Ident and NCIC/CCH) such use must be authorized by Federal statute or Executive order or by State statute if approved by the U.S. Attorney General. 27 A major problem is that State statutes vary widely in terms of defining authorized users of criminal history records. Some State and local officials are concerned about the possibility that CCH information might be provided to users in other States who would be denied access in their own State. The definition of .- 7 28 CFR j 20.33(a)(3). criminal justice users also varies from State to State, although not so widely as the definition of authorized employment and licensing users. The problem of noncriminal justice access is further aggravated by the wide variation in State policies on sealing and purging. Some States purge or seal records of all arrest events that do not result in a conviction, whereas other States maintain all police contact information. Assuming that State laws on sealing and purging and noncriminal justice access to State files will continue to vary widely, four major options have been proposed for dealing with noncriminal justice access to a national index. First, it could simply be prohibited, as was the case for the III pilot and Phase 1 tests. Second, it could be permitted, but only under considerably stronger Federal guidelines than presently exist. Third, a separate index for such use could be established; and fourth, the status quo could be maintained. Prohibiting noncriminal justice access to a national index would conflict with the many Federal and State laws that grant a variety

PAGE 183

Ch. 14Congressional Policy Considerations, Part II l 177 of noncriminal justice users access to existing criminal history record systems. Given that such access serves many lawful and legitimate purposes, such as screening convicted violent or serious offenders from positions of public employment, a total prohibition is probably not realistic. The development of stronger guidelines for noncriminal justice access seems more reasonable. In New York, for instance, existing State statutes mandate that public agencies seeking access to the State CCH file must establish a probable connection between an occupation and criminal activities of potential job applicants. This approach could be mandated for all States by Federal statute or regulation, and might involve the establishment of oversight committees at the Federal and State levels to develop occupation-crime matrices. A third possibility is the establishment of a dual indexone for criminal justice use and one for noncriminal justice use. The criminal justice index would be based on arrest information. The noncriminal justice index could be based on disposition or conviction information only. Thus, persons such as public and private employers would have access only to index entries based on dispositions or convictions. Noncriminal justice inquiries that matched index entries based on arrest-only information would receive a no record response. No new data collection would be required since, in effect, the noncriminal justice file would be a subset of the criminal justice file. A dual index approach is technically feasible and straightforward in a computerized as opposed to manual-system. The FBI already has, in effect, a partial dual file, since arrest data over 1 year old without a disposition are not disseminated to State or local licensing and employment agencies, unless the arrest is known to be still under active prosecution. 28 An intermediate approach would permit noncriminal justice access to a national index for entries based on disposition or conviction information or on arrest information subject to confirmation that the arrest is still active (and that a disposition has not occurred). Under the status quo, noncriminal justice access to a national index would be permitted, presumably subject to existing Federal and State laws and regulations. However, this would require the programing of the index so that access would be consistent with the wide variety of State laws and regulations, and an agreement would be needed on whether the laws of the donor or recipient States, or of the Federal Government or States (for Federal noncriminal justice inquiries), will take precedence where the laws conflict. 28 When requests come in to NCIC/CCH for State and local employment or licensing purposes, the computer automatically checks to determine if all arrest information has a corresponding final disposition. If not, and the arrest is more than 1 year old, the arrest is deleted from the record prior to dissemination. See Bayse, op. cit., p. 8. Oversight and Audi t The purposes of new oversight mechanisms would be to help assure Congress, political executives, managers, courts, and the public that the national index (or any other national CCH system) is operating within the boundaries defined by law and regulation, and to help identify any system problems that might emerge in the course of operation. Oversight is closely linked to system audit, since audit is one of the strongest mechanisms for monitoring system performance. Several possibilities have been suggested. First, Congress could require an annual management report on the operation of a national CCH system. Such a report could include tallies of routine system activity, e.g., participation levels by State and number of inquiries by purpose, as well as statistics on any system irregularities, errors, and problems. Second, Congress could require periodic audits to help ensure compliance with whatever 1+ -. <

PAGE 184

178 An Assessment of Alternatives for a National Computerized Criminal History Syste m record quality and system access standards might be established. For example, periodic record quality checks of both Federal and State files, along the lines of those conducted by OTA for the purposes of this study,* would be necessary to accurately measure the level of compliance with system standards. These audits could check not only the content of records in Federal files against those in State files, but also the Federal and State records against local police arrest and court disposition data in order to determine the extent to which records are inaccurate, incomplete, and/ or ambiguous. In order to keep costs at a reasonable level, the audits would presumably be conducted by sampling Federal and State files on a rotating and perhaps unannounced schedule. Presumably audits would also extend to the use as well as the content of Federal and State files. OTA estimates that a two-person audit team could conduct about 12 reasonably comprehensive audits per year. Thus, five two-person teams would be required to audit each State once a year (50 States plus the District of Columbia and Puerto Rico, for a total of 52 audits per year), the national index and the Federal offender CCH file twice a year, and up to three other Federal CCH files (e.g., Treasury Enforcement and Communication System), plus the national fingerprint repository, once a year. Thus, the 10-person audit staff could be expected to conduct a total of 60 audits per year. Audit priorities could, of course, be adjusted to focus on files with significant irregularities, errors, and problems as identified by operating statistics and/or by the system manager (or, for that matter, by congressional or policy board oversight). The FBI has already agreed in principle to a GAO audit requested by Congress to ensure that the upgraded NCIC communications controllers (front-end processors) and host computers are not used for message switching.** *See ch. 8. **See ch. 5. However, Federal legislation may be necessary to provide GAO with adequate authority to carry out an audit involving direct access to criminal history records. In a 1975 letter to the Senate Judiciary Committee commenting on proposals for comprehensive criminal justice information systems legislation, the Comptroller General advised that: we believe explicit access to the necessary criminal history data should be provided to our office in this legislation because of the sensitive nature of the data involved. We also need access to the records of all non-Federal criminal justice information systems subject to the legislation for the purpose of evaluating the Attorney Generals or the Federal Information System Boards operations under the legislation. An explicit statement of congressional intent regarding this matter should preclude future executive agency reluctance to allow us access to documents we believe we must review to properly discharge our responsibilities. 29 Third, Congress could establish a hybrid audit structure. There are several possibilities. For example, GAO could audit just the national index and Federal offender file, independent State auditing agencies could audit State and local CCH files, and GAO (or some other outside organization) could compare the results of these audits for consistency. The existing Federal law and regulations place the responsibility for annual audit of State CCH systems with each State. Thus, any Federal audit authority, whether granted to GAO or some other body, would appear to require new legislation and/or regulations. Considerable support for such audit authority was expressed in numerous congressional hearings held in the mid-1970s and was reflected in several bills. None of these, however, was enacted. e Letter from Comptroller General of the United States to Chairman, Subcommittee on Constitutional Rights, Senate Committee on the Judiciary, Apr. 14, 1975.

PAGE 185

Ch. 14Congressional Policy Considerations, Part II l 179 Public Participatio n At present, NCICS APB is the mechanism designed to facilitate public participation in the existing NCIC/CCH system. APB is intended to serve as the formal liaison between system users and the system manager (currently the FBI). However, users are defined as those involved in the creation of the data base and who ultimately use it. This confines the definition of users to the criminal justice community. Within that community, it is largely law enforcement agencies that are represented on APB. As discussed earlier, APB could be substantially broadened and strengthened with increased representation from the States and from the rest of the criminal justice community, perhaps along the lines of SEARCH Group, Inc. When compared with NCIC/APB, SEARCH has roughly double the representation from nonlaw enforcement criminal justice agencies (as a percentage of total membership) and has representatives from all 50 States. As of December 1981, NCIC/APB members came from 19 different States. However, NCIC/APB does have a regional structure that includes representatives from all 50 States. Participation from outside the criminal justice community would also help ensure accountability. Alameda County, California, has a model of public participation in a CCH system known as CORPUS (Criminal Oriented Records Production Unified System). Designed and implemented by a local district attorney in cooperation with a broadly based representation of criminal justice and public decisionmakers, CORPUS integrates the information needs of the police, sheriffs, district attorneys, public defenders, courts and probation departments. The CORPUS system is unusual, partly because of the degree of information-sharing among criminal justice agencies that traditionally have not cooperated and that have been excluded from the design of criminal history systems. CORPUS is also unusual in terms of its governance. It is governed by a countywide Alameda County Committee on Criminal Justice and Data Processing (the Parent Committee). This committee includes a broad representation from the criminal justice agencies, as well as the public defender, probation officers, court administrators, five public members, and a member of the county Board of Supervisors. The five public members are appointed by the Board of Supervisors and represent a range of political points of view in Alameda County. The committee has 19 members and conducts monthly meetings that are open to the public. The governing or parent committee is supported by a staff composed of CORPUS administrators. It has regular standing subcommittees on security and privacy and a CORPUS operations committee that reports on day-to-day operations. The presence of public members as well as nonlaw enforcement personnel on the parent committee of CORPUS has resulted in broadly based concern and effort at understanding the privacy and due process implications of criminal justice information. It has also yielded highly developed auditing procedures and significant efforts to ensure the quality and security of information in the system. The parent committee also has the power to hire external auditors to conduct routine audits on the CORPUS system. One local Alameda County assistant district attorney commented: 30 The system of public governance which we have established here in Alameda County takes law enforcement and criminal justice information recordkeeping out of the closet and into the public light. 1m sure most criminal justice agencies around the country will resist this, but so do most other organizations. It all comes down to developing public confidence in the operation of these important information systems. Without participation, without independent insights and external auditors just as in financial institutions, there really can be no public confidence or trust in the operation of systems like this. 0 1979 interview with Alameda County, Calif., assistant district attorney; reaffirmed in 1982.

PAGE 186

180 l An Assessment of Alternatives for a National Computerized Criminal History System The CORPUS experience has been that inclusion of outside (noncriminal justice) individuals and groups is initially uncomfortable, in that issues or perspectives that might otherwise be avoided are forced onto the Parent Committees agenda. However, generally acceptable solutions have apparently been found that have stood the test of the inevitable public scrutiny. In many jurisdictions, public defenders do not have direct access to criminal history record systems, and are not considered part of the criminal justice community. CORPUS does permit public defender access. As noted by an Alameda County assistant public defender:31 One of the reasons that (we are) so supportive of the system is that from the very beginning, we were considered a necessary member of the Alameda County criminal justice community For a multi-State or national system to work effectively, the public and private defender organizations must be accepted as part of the criminal justice communities in which they reside. We were given membership on the (CORPUS) planning and implementation boards and committees and input into the design of the system. As a result, the public defender is a user agency. 3Sept. 22, 1981, letter from Alameda County Assistant Public Defender Duane A. Sciford. Comprehens i Perhaps one of the strongest measures to provide Federal direction and ensure accountability and control would be the enactment of comprehensive national criminal justice information system legislation. As noted in chapter 6, criminal justice information systems currently operate at the Federal level under the very general statutory authority provided by title 28, United States Code, section 534. Although more detailed regulations have been promulgated (and appear as title 28, Code of Federal Regulations, pt. 20), initiatives in the v The National Legal Aid and Defender Association believes that legislation is needed to mandate public defender access to a national CCH system so that defenders can check on the crimin al history records of defendants and witnesses for both the defense and the prosecution. No legitimate policy reason exists as to why the defense should not receive all available CCH information, if truth-seeking and honesty are part of the criminal justice process. 2 As for public participation, the III Evaluation Committee established by NCIC/APB is another example. Here, in addition to representatives of Federal, State, and local law enforcement, the courts, corrections, and prosecutors, the committee includes public members from such groups as the American Civil Liberties Union and the National Association for the Advancement of Colored People. The intent is to help ensure that a wider range of perspectives and affected interests are reflected in the NCIC/APB evaluation of and decisions concerning III. 32Mmm 16, 1982, letter from Jack J. Schmerling, Deputy Director, Defender Division, National Legal Aid and Defender Association. e Legislatio n early and mid-1970s to enact comprehensive legislation were not successful.* Comprehensive legislation could provide explicit authority for a national index or other national CCH system and include statutory guidelines for the operation and use of the system that are much more detailed than those currently available. Eleven areas that could be covered by comprehensive legislation are listed in table 35. Several have been discussed *% di9cu9gion in ch. 6.

PAGE 187

Ch. 14Congressional Policy Considerations, Part 11 l 181 Table 35.Subject Areas Relevant to Comprehensive Legislation on Criminal Justice Information Systems 1. 2. 3. 4, 5 6 Applicability: Federal, State, local Police, courts, corrections, other criminal justice Secondary users (private, public) Information covered: Arrest records (rap sheets) Conviction Correctional Investigative Intelligence Want/warrant Stolen property Collection: Content of records Restrictions on particular types of data elements Restrictions on method of collection Maintenance: Record quality (accuracy, completeness, timeliness) Security (data, people, physical) Separation of files Dedication (complete, partial) Transaction logs Listing of information systems (public notice) Retention: Purging by type of Information Sealing by type of information (e. g., conviction v. nonconviction, juvenile offender) Removal of disqualifications Right to state nonexistence of record Access: Individual (method of review/inspection, challenge judicial or administrative review of challenged information) SOURCE Off Ice of Technology Assessment earlier, and several others are discussed below. Many of these are quite controversial, as evidenced by the last 12 years of debate. However, even in the early 1970s some members of Congress believed that such legislation was a prerequisite for any effective national CCH system. For example, in introducing his own bill in 1974, Senator Sam Ervin noted that: the bill is quite detailed and attempts a resolution of all the major privacy and security issues which have arisen in the development of law enforcement data banks. It endeavors to balance the legitimate needs of law enforcement with the requirements of individual liberty and privacy. It would for the first time give firm statutory authority for criminal justice data banks, a major obstacle in the development of such systems. It would impose upon the data banks strict but manageable privacy limitations. Not the least important, the bill also attempts to solve fundamentally important questions of Federal-State reResearcher (method of use, challengejudicial or administrative) Media 7. Training: Data processing and recordkeeping personnel Primary users (use, interpretation) Secondary users (use, interpretation) 8. Dissemination by type of information: Primary users Secondary users 9. Penalties: Civil Criminal Administrative sanctions 10. Auditing/evaluation of: Use (primary, secondary) Record quality Operations Management Social impacts (privacy, confidentiality, and security) 11. Regulatory authority: Type (operating agency, special council/board, advisory group Responsibilities (consultation, study and advise, establish policy and procedures, oversight, audit) Membership (stakeholders included) Duration (permanent, temporary) Resources (executive director, staff, general appropriation/specific allocation) Powers (subpoena, hold hearings, mandate binding policies and procedures, audit, mandate reporting requirements) lationships in these comprehensive national information systems. 33 Given the vagaries of the criminal justice process, it would be a difficult challenge to ensure the accountability and control of a national index or other national CCH system in the absence of national legislation. Criminal law and the customs and traditions of the criminal justice process vary widely among the States and localities. Criminal justice agencies are under increasing pressure to apprehend, process, and dispose of persons suspected of committing a crime as expeditiously as possible. This is done through the use of strategies such as pretrial diversion, plea bargaining, presentencing negotiations, and career criminal programs designed presumably to improve the efficiency and effectiveness of an already overburdened and underfunded Congressional Record Senate, Feb. 5, 1974. *See ch. 9.

PAGE 188

182 l An Assessment of Alternatives for a National Computerized Criminal History System criminal justice process. As a consequence of all these factors, the process by which criminal history records are generated also varies widely. In addition to policy control, file size and content, record quality, and noncriminal justice access, national legislation could enhance the accountability and control of a national CCH system by addressing the following other areas. Access, Review, an d Challenge Procedure s At present, any individual has the right to access and review his/her criminal history record maintained by NCIC/CCH or Ident. However, if the record is believed to be incorrect or incomplete, the individual must seek correction by the source agency. Individuals may also direct a record challenge to the FBI, which will then forward the challenge to the source agency. The FBI will make any changes necessary only upon proper notification by the source agency .34 An alternative approach would be to: 1) hold the national index or other national CCH system manager fully accountable for the accuracy and completeness of records referenced in the index; and 2) give the national index manager the responsibility and legal authority to require the State repository to verify with the source agencyand correct if necessary any index entries and underlying records challenged by properly identified individuals. While this might appear to be potentially quite costly and burdensome, the experience of the many States with access, review, and challenge statutes has been that challengers are few in number.* It appears that individuals with records in criminal justice information systems are not very likely to exercise their rights of access, review, and challenge. While many of these individuals may have strong incentives to conceal their records, some may not know how to exercise their rights or may not even be aware of them. 34 28 CFR j 20.34, and 28 CFR $ 16.34. *See ch. 9. Criminal Penaltie s Under current Federal law and regulation there are no civil or criminal penalties for violation of NCIC/CCH system standards, except as provided by the Privacy Act of 1974; for example, for willful unauthorized disclosure of records that contain individually identifiable information. Any agency or individual violating Federal regulations on State and local criminal history information systems is subject to a fine not to exceed $10,000 and possible cutoff of LEAA funds. 35 As a practical matter, LEAA no longer exists and LEAA funding for CCH-related systems ended in fiscal year 1981. Any agency or entity failing to comply with regulations on the Federal systems and interstate exchange of criminal history information is subject to cancellation of NCIC/CCH and Ident services. 36 Accountability of a national CCH system could be strengthened by establishing criminal penalties, or at the minimum strong civil penalties, for violation of national system standards. Many States already have enacted both civil and criminal penalties, particularly for violation of various privacy and security statutes and regulations applicable to criminal history record information systems. Privacy Standard s Legal accountability of a national CCH system could be strengthened by bringing such a system more fully under the requirements of the Privacy Act of 1974. For example, this act requires each Federal agency with a records system to keep an accurate accounting of the disclosure of a record (including the date, nature, and purpose of each disclosure and the name and address of the recipient), 37 and to retain this accounting for at least 5 years or the life of the record, whichever is longer. 38 These two provisions are mandatory and have been implemented by the FBI. The Privacy Act also requires each agency to make S28 CFR $ 20.25. 3 CFR j 20.38. 37 5 USC $552a(c)(l), Privacy Act of 1974, public Law 93-579. 3 USC $552a(c)(2).

PAGE 189

Ch. 14Congressional Policy Considerations, Part II 183 the accounting available to the person named in the record at that persons request. 39 However, the act provides that law enforcement and criminal justice record systems can be exempted from this requirement, 40 as is the case with the FBI. As another example, the Privacy Act requires agencies maintaining a record system to publish a notice in the Federal Register, at least annually, that includes the agency procedures whereby an individual can be notified at his request if the system of records contains a record pertaining to him. 41 Here, too, the act provides a legal exemption for law enforcement systems, although both Ident and NCIC/ CCH do publish such notices. In sum, Federal law enforcement and criminal justice record systems, including Federal criminal history record systems, may exempt themselves from a number of the legal accountability measures in the Privacy Act. For example, should Congress wish to make it easier for individuals to request corrections in criminal history information contained in a national index or other national CCH system, the Privacy Act could be amended to remove certain exemptions now exercised by NCIC/ CCH. 42 Alternatively, provisions of the Privacy Act could be incorporated in new legislation to ensure that privacy standards are applicable to a national index, and that such standards also extend to the records maintained in Federal and State repositories on which index entries would be based. Technical Accountabilit y Measures to provide technical accountability include, for example, comprehensive transaction logs, systematic audits of local user agencies, and training of employees in the im USC $552a(c)(3). 40 5 USC $552a(j)(2). 4 USC $552a(e)(4)(G). 425 USC $552a(d) places responsibility for correcting records in the Federal agency maintaining g the record system. Moreover, law enforcement agencies can exempt themselves from this requirement and Ident and NCIC/CCH have done so by placing primary responsibility for record accuracy and completeness with the agencies originating the records. plementation of applicable State and Federal regulations. As discussed in chapter 9, as of 1979, the transaction logging, local auditing, and training procedures varied widely and frequently fell short of fully accounting for the flow of criminal history information within the States.* In the OTA 50-State survey, to which 49 States responded, dissemination logs maintained by State repositories contained the following information: name or identification number of requesting agencies (46 States); type of information disseminated (43 States); purpose of request (34 States); requestors terminal identification number (32 States); name or identification number of person requesting information (32 States); and user agreement or authority (13 States). Major problems included lack of funding, wide variations in the interpretation of State and Federal regulations, and absence of the necessary statutory or policy mandate. In addition, there are inherent difficulties in implementing logs and audits at the local level. Some States, such as Minnesota and New York, have implemented detailed logging procedures. Minnesota has added to its State repository transaction log the name of the police or other officer making the request for criminal history information. This automated personal identifier will assure the central State repository the capability to at least identify who it was in a local agency who requested and used the information. This method should be much more accurate than the maintenance of manual logs alone. New York State has implemented a central State repository transaction log that includes, in addition to the typical agency identifiers and purpose code requests, the personal identifier for the officer making the request, and in addition, a case number for which the request was made. Systematic review of transaction log books is accomplished by randomly selecting requests for criminal history information from the log, going to local agencies that received the information and from *J$~e nearly ~ States maintain transaction logs, asof1979 nearly two-thirds review the logs ordy when a specific abuse is indicated. Only a few States conduct systematic audits of user agencies. See ch. 9.

PAGE 190

184 An Assessment of Alternatives for a National Computerized Criminal History System there directly to the person who made the request, and reviewing the files in which the information was used. Additional problems are raised when considering audit trails for local criminal history systems. There may be hundreds of police agencies in a large State, each with different local criminal history rap sheets that often may be simply recorded on 3 X 5 cards. An examination in California showed how difficult it was to develop audit trails for such card systems. Thus, technical accountability would appear to be possible but difficult to achieve in a national CCH system. Because the records or entries in such a system would originate largely from and be used by State and local law enforcement and criminal justice agencies, audit trails and transaction logs would be needed at the State and local agency levels. Congress could mandate and fund the development of computer-based user audits, transaction logging procedures, and the required training programs through enactment of comprehensive legislation establishing a national index or other national CCH system. Uniform Crime Codes an d Record Format s Another problem with the interstate exchange of criminal history records is the wide variability in the classification and coding of crimes and in criminal history record formats. The more complex and variable the crime codes and record formats, the more difficult the records are to understand, particularly for out-of-State users. Comprehensive legislation could establish a national crime coding standard along the lines presently used by NCIC, and could promulgate guidelines for criminal history record formats. Surveys conducted by the State of Florida and NCIC have found that summary records are frequently too abbreviated to be fully useful, while the full records may have more details than are really needed. NCIC is developing a new record format that might strike a better balance. III Development Pla n The III pilot test with the State of Florida was completed during July through September 1981. Based on the generally favorable results, Phase 1 of the III development plan was completed during February and March 1982 with the addition of five other fully participating States.* In December 1981, NCIC/ APB concurred in the NCIC staff recommendation to combine Phases 2 and 3 of the original III plan. In June 1982, NCIC/APB recommended that the FBI proceed to implement the consolidated Phase 2. As revised, in Phase 2, already participating States will be tested on a fully decentralized basis (multiState as well as single-State records decentralized) and currently nonparticipating States will be added to III on a decentralized basis. This revised plan will provide a more clear-cut test of the national index concept, more vivid*These states included Michigan, North Carolina, South Carolina, Texas, and Virginia, in addition to Florida. ly demonstrate the FBIs commitment to a national index CCH alternative, and eliminate the need for the nonparticipating States to join III first on a single-State/multi-State basis before shifting to a national index basis at a later date. In Phase 3 of the revised plan (Phase 4 of the original plan), the III concept would be fully implemented. One unresolved question concerns whether Phase 2 should include a test of III with no message switching as well as with partial message switching. As discussed in chapter 10, III was originally proposed to involve no message switching. For a variety of reasons, the III pilot test with the State of Florida and Phase 1 with Florida plus five other States were conducted using a form of partial message switching known as automatic inquiry referral (AIR). While there may be significant advantages to AIR, several of the advantages that are

PAGE 191

Ch. 14Congressional Policy Considerations, Part II 185 claimed cannot be effectively substantiated without a test that would compare operational results with and without AIR. For example, the State of Florida (and other members of the III Subcommittee of NCIC/APB) believes that AIR would improve the overall III response time, reduce the burden on inquiring States by eliminating the need to send a second round of messages to States holding the desired records, simplify the verification of authorized users, standardize the format of inquiry messages, and provide greater security for messages than would be available over NLETS. NLETS has disputed the concern over security, noting that even with AIR, the actual summary records are transmitted over NLETS. In addition, NLETS is clearly capable of handling both the inquiry and record traffic. Use of NLETS might avoid any possible overload of NCIC, and might encourage greater participation of States without an automated interface. An operational test could perhaps resolve some of these concerns. The NCIC staff, NCIC/APB, and NLETS have initiated discussions of whether and how such a test might be carried out. In view of the high percentage of name hits during the III Phase 1 test that did not match the subject individual, 43 NCIC is proposing a two-step inquiry when unique identifying numbers are not available. Thus, in case of a name hit, NCIC would return only the identification segment to the inquiring agency to confirm the hit prior to referring the inquiry on to the State or States of record. The inquiring State could then exercise the option of using NLETS rather than NCIC to contact the State or States holding the record, and thus obviate part of the need for AIR. The second unresolved question concerns the quality of III, especially during Phases 2 and 3. One plan is to establish the index initially by extracting names and identifiers from the 5.8 million AIDS records and the 0.8 million CCH records not in AIDS. This would result in an initial III containing 6.6 million 43 NCIC, Draft Report-Preliminary Findings of the FebruaryMarch 1982 III test, April 1982, p. 18. entries. Additional entries would be provided by participating States. Another plan would be for States to either enter their own index records or have AIDS make the entries on their behalf. In any event, III quality would ultimately depend on the quality of the records in the Ident/AIDS and NCIC/CCH files for Federal offenders and on the quality of the State criminal history files, on which the index is based. OTA research presented in chapter 8 found that, as of mid-1979, more than a quarter (27 percent) of the records disseminated from the NCIC/CCH file were missing a court disposition, and about one-fifth contained inaccurate information. OTA also found that, as of mid-1979, over 30 percent of the records disseminated from the Ident file lacked a court disposition and about one-fifth percent contained inaccurate information. At the State level, OTA found that on the average about 35 percent of dispositions were not reported. Assuming that these figures are still reasonably valid, roughly one-third of the III entries on the average would be expected to be based on arrest data where a disposition had occurred and was recorded at the local level, but had not been reported to the State repository or to the FBI. This potential record quality problem highlights the role of III as an index to arrests, not to dispositions or convictions. A hit on the index would simply mean that the subject was arrested at some time for something in the State indicated. A hit would not say anything about whether the individual had charges dropped or dismissed or was acquitted, convicted, incarcerated, or served time and was released. In view of these concerns, Congress may wish to direct that record quality research be conducted as part of Phase 2 of the III development plan. This would help to determine where III currently stands with respect to 44 See NCIC staff paper prepared for the Nov. 3-4, 1981, meeting of the Interstate Identification Index Subcommittee of the NCIC Advisory Policy Board, Topic *7, p. 6. *SW chs. 8 and 9.

PAGE 192

186 An Assessment of Alternatives for a National Computerized Criminal History System record quality and the extent to which probensure a higher index quality. For example, lems exist in the records on which the index one possibility would be to create the index is based. No such research was included in the from scratch based on current and updated in111 pilot test or Phase 1. Congress may also formation submitted directly from the States wish to request the development of alternative and meeting some set of minimum record qualplans for establishing III in ways that would ity standards. AIDS/CCH Consolidation Pla n As noted in chapter 4, the FBI is currently operating two filesAIDS and NCIC/CCH that maintain computerized criminal history records. 45 Between 1979 and 1981, the percentage of NCIC/CCH records also held in AIDS increased from about 44 to 58 percent. Thus, the two files duplicate each other to a significant and growing extent. This duplication reflects the unique need of Ident to develop a capability for automated fingerprint identification, the low level of State participation in the NCIC/CCH file, and the absence of a clear and agreed on long-range plan. AIDS and NCIC/ CCH have now developed to the point where the U.S. Attorney Generals Task Force on Violent Crime recommended that the FBI prepare plans to reduce duplication between AIDS and NCIC/CCH and that such plans take into account the results of phase 1 of the III development plan. 46 Any AIDS/CCH consolidation is likely to have a significant impact on the cost of FBI criminal history and identification services and could be an integral part of a national CCH system. Therefore, Congress may wish to request the preparation of plans for AIDS/ CCH consolidation that would be consistent with the national index and other national CCH system alternatives under consideration, and that would provide a clear basis for comSee U.S. Comptroller General, The FBI Operates Two Clnnputerized Criminal History Information System% U.S. General Accounting Office, Washington, D. C., 1979. U.S. Attorney Generals Task Force on Violent Crime, Final liepor~ U.S. Department of Justice, Washington, D. C., Aug. 17, 1981, pp. 67, 69. paring costs of the various alternatives at least with respect to the Federal share. Should Congress make a decision on a specific national CCH system, DOJ could be directed to prepare a plan to consolidate AIDS and CCH in a way that is consistent with costeffective implementation of the alternative selected. For example, from a technical viewpoint, a national index plan could: l l l l l l establish a national index on serious offenders as a new NCIC file; provide initial inputs (screened to meet quality standards) to the index from the NCIC/CCH file and Ident/AIDS; subsequently return all NCIC/CCH records to the States, except for records on Federal offenders; phase out the AIDS automated rap sheet function; include the AIDS automated fingerprint identification service as another new NCIC file; and seek further inputs to the index from the States (this could be in the form of computer tapes). The inputs would have to be screened to remove any that did not meet applicable quality standards. NCIC could be restructured as shown in table 36. These functions could be located in a new National Criminal Information and Identification Division of the FBI that would combine the existing Ident Division and NCIC Section and perhaps other related activities. After reorganization, the FBI would maintain full records on Federal offenders only. The only other criminal information available would be hot file information, index informa-

PAGE 193

Ch. 14Congressional Policy Considerations, Part II l 187 Table 36.Possible File Structure of a Reorganized National Crime Information Center File Nos. File content 1-8 Existing NCIC hot files 9 CCH full record file limited to Federal offenders 10 Existing CLIS b file 11 AIDS fingerprint file with interface to residual manual file c 12 National index file d lncludlrlg stolen vehicles, stolen guns, stolen Ilcense plates. wanted Persons, stolen securities, stolen boats, and mlss!ng persons bcomputerlzed Laboratory Information SYstem CAf S o known as the Nat!onal F[ngerprlnt File dAlso known as the Interstate Identlflcatton Index SOURCE Office of Technology Assessment tion, and fingerprint identification information. Congress could authorize this reorganization through new legislation that could also, for example: l l l l l l l l l establish statutory guidelines for operation and use of the system; require periodic outside audit by GAO and/or an independent board; broaden and strengthen the NCICS APB or establish an independent board to exercise policy control; specify limits on file and index size and content (e.g., CCH full record file limited to Federal offenders); mandate record and index quality and system access standards (e.g., with respect to disposition reporting and noncriminal justice access); authorize Federal funding and technical assistance to strengthen State and local CCH systems, field audits, and court disposition reporting; reiterate and revise the title 28 privacy and security regulations where necessary; codify NCIC/CCH operating procedures where appropriate; and specify message switching authority and responsibilities. Finally, when considering AIDS/CCH consolidation plans, Congress may also wish to examine the pros and cons of shifting management of a national CCH system out of the FBI to a new bureau within DOJ, to an independent Federal agency, or to a consortium of States. It has been argued that the FBI is burdened with too many contradictoryor at least conflicting-responsibilities. The FBI is primarily an investigatory and law enforcement agency, but also bears a heavy responsibility for the maintenance of criminal records and the production of criminal statistics. In addition, it offers training programs for State police officials, maintains an extensive forensic laboratory, and provides a considerable amount of technical assistance to State and local (as well as Federal) law enforcement agencies. Since any national index or other national CCH system would still depend on AIDS to a large extent for fingerprint identification and presumably on NCIC for computer and communication support, adequate management and technical coordination might prove to be difficult unless AIDS and NCIC were also moved to another agency. But this could possibly weaken the commitment and cooperation of State and local law enforcement agencies, which have been based in part on their longstanding involvement and rapport with the FBI. On the other hand, judicial and correctional agencies, for example, might be more inclined to actively participate in a national CCH system if it were managed by someone other than the Nations preeminent law enforcement agency. However, judicial and other nonlaw enforcement criminal justice agencies could be given a substantial role in a broadened and strengthened NCICS APB or an independent board established to exercise policy control over a national CCH system. Alternatively, the current NCIC APB could be left as is with its jurisdiction limited to the NCIC hot files. A new and separate APB, with substantial representation from prosecuting attorneys, judges, defense attorneys, correctional and probation/parole officials, and the general public, could be given jurisdiction over a national CCH system. Under either of these conditions, and with strong legislative guidance and congressional oversight, the FBI might be in the best position to enforce system standards and operating procedures and man-

PAGE 194

188 An Assessment of Alternatives for a National Computerized Criminal History System age the Federal component of a national sysnal justice services agency, as opposed to a tern. Nonetheless, the experience of the several police or law enforcement agency, should be States (e.g., New York) that have successfulexamined for relevance to control and managely located their State CCH system in a crimiment of a national CCH system. Private Carrier Role in National CCH Syste m Congress may wish to review the role of private communication carriers in a national CCH system. At present, the NCIC communication lines are leased from private carriers; and NLETS is a private nonprofit corporation funded and controlled by the States that also uses dialup or leased lines from private carriers. Two questions have arisen. First, for the national index alternative, would new technical equipment and systems available horn private vendors eliminate any need for partial message switching (automatic inquiry referral)? It may be that new computer technology alone, or in combination with new communication technology, could substantially reduce or eliminate problems with inquiry formatting, verification, and followup, and simultaneously provide greater security. The growing availability of small, inexpensive, easily programmable, yet powerful computers means that States with manual criminal history files will be able to automate more easily and at less cost than previously thought, and will be better able to fully participate in a national index or ask-thenetwork CCH system. Second, would the national index or ask-the-network alternative be more cost-effective using a privately offered communication network? For example, several private carriers now offer nationwide data communication networks, both packet switched and message switched, broadcast as well as narrowcast, using satellite as well as landline transmission links and providing security measures such as data encryption. The possible benefits from the greater use of privately offered new technology and services are significant enough to warrant further consideration. Congress may wish to commission more detailed research on the private sector role in a national CCH system.

PAGE 195

Appendixes

PAGE 196

Appendix A Status of NCIC Hot Files The National Crime Information Center (NCIC) hot files (e.g., on wanted persons, missing persons, and stolen vehicles, articles, guns, license plates, securities, and boats) are heavily used by Federal, State, and local law enforcement agencies. As of September 1981, hot file transactions were exceeding 300,000 daily and approaching 10 million monthly. No one has conducted a systematic measurement of the benefits of the hot files. However, the consensus of opinion among law enforcement officers interviewed by or expressing an opinion to OTA was that these files represent an invaluable tool in the apprehension of wanted persons and the recovery of stolen property. Anecdotal evidence on NCIC hot file hits supports this view, as do the results of other surveys cited in chapter 11. The status of the NCIC hot files warrants consideration in three important areas. First, various parties (including Federal agencies and individual Members of Congress) periodically ask the Federal Bureau of investigation (FBI) to establish new hot files, for example, on violent offenders, missing children, parolees, and dangerous persons. Most recently, in November 1981, the U.S. Secret Service asked the FBI to establish a file on persons judged to represent a potential threat to protectees, including the President. This file would help the Secret Service monitor the movements of or keep aware of the location of dangerous persons. Whatever the merits of these proposals, such proposals could involve the use of NCIC to gather intelligence data on or track individuals not formally charged with a current criminal offense. Such use of NCIC might lead to unwarranted invasions of privacy and improper detentions or arrests. 3 A review of the entire process by which such proposals are made and evaluated could be useful. If new legislation were developed for a national CCH, consideration might be given to including a specific statutory framework for the NCIC hot files. The quality of these files (discussed below) and criteria for accessing them might also be addressed. At present, there apparently are no legal or policy prohibitions on the indirect dissemi*See ch. 4. See Nov. 24, 1981, letter from the Secret Service Director to the FBI Director. See May 19, 1981, letter from William A. Bayse, FBI Assistant Director, to Jerry Morgan, Chief U.S. Probation Officer, Southern District of Georgia. Ibid. nation of hot file information to noncriminal justice agencies or private individuals. 4 Second, even with the upgrade of the NCIC host computer and front-end processor,* the FBI is concerned about the ability of NCIC to handle the growing volume of transactions and programing requirements. The NCIC Section and the NCIC APB have initiated a long-range planning effort on the future technology and operational needs of NCIC. In addition to a full range of technical options, it would be important to ensure that the needs of the entire criminal justice community, as well as any congressional decisions on a CCH system, are factored into the planning effort. Third, NCIC continues to experience some problems with the quality of the hot files. For example, during a short period in 1981, FBIs Identification Division (Ident) collected (on a nonrandom, unsystematic basis) FBI identification numbers for about 75 fugitives who had been apprehended and cleared from the Ident file but had not yet been canceled from NCIC. 5 NCIC notes that there could be several reasons for this; for example, a different name might have been used by the fugitive, extradition limitations might warrant retention of the record, or the apprehending agency might have failed to check NCIC. 5 NCIC repeatedly urges users to cancel their records from NCIC when a fugitive is no longer wanted. Promptly canceling these records will protect the rights of the citizen and eliminate the possibility of false arrest. 6 NCIC operating procedures require confirmation of all hits on the hot files. Despite a policy mandating immediate (within 10 minutes) response to a request for confirmation, complaints of untimely responses have been received by the FBI, NCIC has advised users that (d)elay in responding to such a request could subject the agency to a lawsuit (and damage awards for false arrest) or result in the release of wanted and missing persons or stolen property not being returned. Delays in entering records into the hot files, sometimes up to several months after a warrant is issued or stolen property reported, have See FBI, minutes of the June 17-18, 1981, meeting of the NCIC Advisory Policy Board, Topic +12, pp. 35-36. *See ch. 5. Mar. 26, 1982, telephone discussion with Conrad Banner of the FBI. aNCIC Newsletter. September 1981. ACIC Newsletter, February 1981. 191

PAGE 197

192 An Assessment of Alternatives for a National Computerized Criminal History System also been experienced. Again, NCIC has advised users that delayed entry reduces or eliminates the possibility of apprehending wanted persons, locating missing persons, and recovering stolen property, as well as unnecessarily endangering law enforcement officers, 8 The only NCIC hot file examined in detail by OTA was the wanted persons file. OTA found that of the 394 warrants that could be verified (out of a random sample of 405 selected on Aug. 4, 1979), 5.8 percent (23 warrants) had been cleared or vacated at the local level prior to August 4, 1979. As shown in table A-1, 12 of the 23 warrants had been cleared or vacated more than a month prior to August 4, and 7 of these 12 more than 6 months prior to August 4. In recent years, the FBI has taken actions to improve the currency and accuracy of the wanted persons file. It routinely conducts a large number of record quality checks, and repeatedly urges users to enter, update, and cancel records on a timely basis. It systematically requests originating agencies at the State and local levels to verify the warrants they have placed in the file. Still, OTA found that as of August 4, 1979, a possibly significant percentage of warrants in the file (5.6 percent, or approximately 7,400, 4 percent) were cleared or vacated. To the extent these individuals were at risk of being improperly detained and perhaps arrested, or detained and perhaps arrested but subsequently neither extradited nor prosecuted, both effective law enforcement and constitutional rights could have been compromised. Further improvements in the quality of the wanted persons file (and perhaps the other NCIC hot files) may be warranted. Possible actions include: l l l l establishing standards for retention of outstanding warrants; establishing tighter standards for retention of cleared or vacated warrants; requiring more frequent certification of warrants as still valid by originating agencies; and conducting periodic random audits of the wanted persons file to monitor record quality (currency and accuracy) and to verify that originating agencies are fully complying with certification requirements. At present, to the best knowledge of OTA, NCIC does not conduct any record quality audits of either the wanted persons file or the other hot files. Perhaps the most important potential action would be to initiate periodic audits to establish benchmarks, to identify problem areas, and to ,VCIC ,\ewsletLer, August 1981, Table A-l .NCIC Wanted Persons File: Record Quality as of Aug. 4, 1979 Number of warrants in general warrant file Aug. 4, 1979 . . . . . ..........127,500 Sample size (random sample, Aug. 4, 1979 . 405 Originating agency responses . . . . 394 Status of warrants in file Warrant cleared or vacated prior to Aug. 4, 1979 a . . . . . . 5.8% ( 23) Agency had no record of warrant b . 4.1 0/0 ( 16 ) Wanted but no warrant locatable . . 0.80/0 ( 3) Warrant cleared or vacated but date of clearance unknown . . . 5.1 % ( 20) Warrant cleared or vacated after Aug. 4, 1979 C . . . . . . 2.80/o ( 11) Warrant outstanding . . . . 81 .50/0 (321) Total . . . . . . . 100.1 /0 (394) aClearance dates for the 23 warrants were as follows 12/9/71, 12/30/76, 10/1 7/77. 6/13/78, 8/31/78, 12/7/78, 1/27/79, 4/5/79, 5/1/79, 6/1/79. 6/6/79, 7/2/79, 7/14/79, 7/19/79. 7/23/79, 7/26/79 (2), 7/30/79, 8/1/79 (5) bThe FBI has noted that use of the originating case agency number might have helped in further verifying these warrants C Warrants cleared or vacated after the date of the Sample were not considered in determining record quallty, s!nce the status of the warrants in the NCIC file might have changed between the date of sample and the date of verification Note on Methodology The sampling procedure for the NCIC Wanted Persons file was a random sample of the population of general warrants, a Iist of which is produced weekly by the FBI The population of warrants used in this research was produced Aug 4, 1979 A series of random numbers was used to select 405 warrants for verification Each of the warrants was verified by telephoning and/or writing the originating local agency as listed on the records selected from the NCIC Wanted Persons file The statistical estimates of record quality produced in this research were tech. nically generalizable to the population of warrants contained in the NCIC fiIe as of Aug 4, 1979 The ability to estimate population parameters using randomly drawn samples IS a function of sample size as well as the underlying distribution of the variable being estimated. For the Wanted Persons file sample, the 95 per cent confidence interval for the true population parameters IS plus or minus 4 percent That Is, there is 95 percent confidence that the true population param eters of record quality Iie within plus or minus 4 percent of the estimates given in the table SOURCE: Office of Technology Assessment determine whether or not record quality is improving. Record quality audit is defined here as a systematic comparison of a random sample of records in the NCIC file with the corresponding records held by the originating agency. At present, NCIC sends a list of warrants every 6 months to each State terminal control agency, which in turn is responsible for obtaining certification by originating agencies within 75 days, Warrants that are not certified or certified as vacated or cleared are removed from the file. One possible problem is that vacated or cleared or otherwise invalid warrants may be maintained in the file for as long as 81/2 months (6 months plus 75 days) prior to the next certification. Another is that some originating agencies may not be fully complying with the certification requirements. Periodic random audits, such as are currently conducted by the State of California, 9 should help to encourage full and complete certification. Per Feb. 16, 1982, telephone discussion with Fred Wynbrandt, California Department of Justice.

PAGE 198

Appendix B State Officials Responding to OTA 50-%ate Survey Conducted i n 1979-80 Arizona Arkansa s California Delaware The following individuals were listed as responsible for completing the written survey questionnaire for the States responding: State Name and/or agency Alabam a Name not provided. Alabama Criminal Justice 1nformation Center Alaska Susan E. Knighton, SAC Director Criminal Justice Planning Agency Robert J. Edgren, Coordinator Arizona Criminal Justice Information System David Eberdt, Director Arkansas Crime information Center Barbara G. Myers/victor J. Paradis California Department of Justic e Lt. Jay R. Brackin, Colorado Connecticut Florida Georgia Supervisor State Bureau of Identification W. G. Buckley Colorado Bureau Investigation Lt. Robert Hull, of Commanding Officer Connecticut State Police Peggy Horvath, Deputy Director Division of Criminal Justice Information Systems Florida Department of Law Enforcement E. W. Manseau, Deputy Director Georgia Crime information Center Kentucky Louisiana Maine Maryland State Name and/or agency Hawaii Frederick R. Witte, Systems Analyst Hawaii Criminal Justice Statistical Analysis Cente r Illinois John T, Loverude, Assistant Bureau Chief Illinois Bureau of identification Indiana Capt. James S. Kinder Criminal Justice Data Section Iowa G. W. Shanahan, Chief Division of Criminal Investigation Kansas Michael E. Boyer, Director Statistical Analysis Center J. Carey Brown, Administrative Officer Kansas Bureau of investigation J. Bruce Lee, Administrative Supervisor Criminal Justice information Bureau of State Police Lt. Leon Millet, Supervisor Criminal Records Louisiana Bureau of Criminal Identification Robert E. Wagner, Jr., Director Maine Bureau of Identification Earl L. Gillespie, Coordinator Criminal Justice Information System Maryland Department of Public Safety and Correctional Services 193 I

PAGE 199

194 l An Assessment of Alternatives for a National Computerized Criminal History System State Massachusetts Michigan Minnesota Mississippi Missouri Montan a Nebrask a Nevad a New Hampshire New Jersey New Mexico New York North Carolina North Dakota Ohio Name and/or agency Louis Sakin, Executive Director criminal History Systems Board Dallas G. Piper, Administrator Identification Section Michigan State Police Don Love, Director Criminal Justice Information System Gordon W. Skelton, Systems Analyst Statistical Analysis Center Criminal Justice Planning Commission Bob Winslow, Deputy Director Missouri Department of Public Safety Herb Bruning, Chief Identification Bureau Lt. Wayne Rowe Identification Division Nebraska State Patrol Richard S. Staub Criminal Justice Specialist No name provided. Wallace P. Miller Bureau of Identification New Jersey State Police Sheila A. Cooper, Manager Statistical Analysis and Evaluation Center Capt. David Kingsbury New Mexico State Police Adam F. DAlessandro Deputy Commissioner New York State Division of Criminal Justice Services William C. Corley, Director North Carolina Police Information Network Richard S. Milde, Chief Agent State Bureau of Criminal Investigation J. R. Wogaman, Project Director Criminal Justice Information System Office of Criminal Justice Services Marlin Dowler, Chief Bureau of Identification State Oklahoma Oregon Pennsylvania Puerto Rico South Carolina South Dakota Tennessee Texas Utah Vermont Virginia West Virginia Wisconsin Washington Wyoming Name and/or agency Jim Wilson, Acting Director Research and Evaluation Oklahoma Crime Commission W. H. Freele, Director Bureau of Criminal Identification Oregon State Police Joseph Riggione, Director Governors Task Force on Criminal Justice Information Systems Edgar E. Perez Bracelty, Director Criminal Justice Information System Lee M. Thomas, Director Division of Public Safety Programs Don Licht, Director Division of Criminal Investigation Capt. Tony Hansberry Archie Harem Willis Whatley, General Counsel Criminal Justice Division Governors Office (with assistance from Texas Department of Public Safety) Del Mortensen, Director Bureau of Criminal Identification Billy J. Chilton, Director Vermont Crime Information Center Capt. W. R. Wagner, Jr. Records and Statistics Officer Virginia State Police M. Sgt. R. E. Yost, Records Officer Capt. Armstrong, Communications Division Department of Public Safety Larry Quamme, Director Wisconsin Crime Information Bureau Officer Wayne Harsh Identification Section Washington State Patrol David G. Hall, Director Division of Criminal Identification

PAGE 200

Appendix C State Repository Officials Responding to OTA User Survey Conducted in 1979 The following individuals responded by letter and/or telephone interview to the OTA user survey: State Name and agency Alabam a Eugene J. Akers, Manager Systems Development Alaska Arizona Arkansa s California Colorado Delaware District of Columbia Division Alabama Criminal Justice Information Center Sgt. Michael S. Radisch CJIS Security Officer Alaska State Troopers Department of Public Safety Capt. W. Woodard Arizona Department of Public Safety David Eberdt, Administrator Arkansas Criminal Justice and Highway Safety Information System Maj. Buren Jackson Arkansas State Police F. W. Johnston and Roy T. Iwata Bureau of Identification California Department of Justic e W. Gray Buckley, Agent in Charge Crime Information Section Colorado Bureau of Investigation Benjamin M. Miller, Chief Communications Center Delaware State Police Charles J. Corcoran, Deputy Chief Communications and Data Processing Division Metropolitan Police Departmen t State Florida Georgia Hawaii Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Massachusetts Michigan Name and agency Robert L. Edwards, Director Peggy Horvath, Deputy Director Division of Criminal Justice Information Systems Florida Department of Law Enforcement E. W. Manseau, Deputy Director Georgia Crime Information Center Jack Piper Honolulu City and County Police Department Dwight E. Bee Illinois Department of Law Enforcement Robert J. Stanton Indiana State Police Gary L. Stevens Iowa Department of Public Safety Maj. Stuart A. Elliot Kansas Highway Patrol Mr. Kraing Kentucky State Police Lt. L. G. Finn Louisiana State Police Sgt. John H. Parkin, Jr. Maine State Police Louis Sakin Maryland Department of Public Safety and Correctional Services Emile Thibault Massachusetts State Police Henry Sedmak, Exective Secretary Law Enforcement Information Network Policy Council Michigan State Police 195

PAGE 201

196 l An Assessment of Alternatives for a National Computerized Criminal History System State Minnesota Mississippi Missouri Nebraska New Hampshire New Jersey New Mexico North Carolina Ohio Oklahoma Oregon Name and agency Don M. Love, Director Criminal Justice Information Section Bureau of Criminal Apprehension Lt. Art Richardson, Director Mississippi Justice Information Center Capt. J. M. Luker, Assistant Director Criminal Division Missouri State Highway Patrol Capt. D. W. Schamp Nebraska State Patrol Capt. David Dempsey New Hampshire State Police Sgt. Donald Menzel, et al. Division of State Police Capt. Monroe K. Alexander, et al. New Mexico State Police William C. Corley, Director Police Information Network North Carolina Department of Justice Lt. Col. C. C. Hayth Ohio State Highway Patrol Steve D. Tires Oklahoma Law Enforcement Telecommunication Systems Department of Public Safety Lloyd A. Smith, Manager Law Enforcement Data System State of Oregon Executive Division State Pennsylvania South Carolina South Dakota Tennessee Texas Uta h Vermont Virginia Washington Wisconsin Wyoming Name and agency Benjamin R. Jones Pennsylvania State Police Lt. Carl Stokes, et al. South Carolina Law Enforcement Division Donald G. Licht Division of Criminal Investigation Capt. Tony Hansberry, et al. Criminal Justice Information System Tennessee Department of Safety H. A. Albert, et al. Texas Department of Public Safety Del Mortensen, et al. Bureau of Criminal Identification Utah Department of Public Safety Sgt. Billy J. Chilton, Director Vermont Crime Information Center Capt. W. R. Wagner, Jr. Virginia State Police Capt. George B. Tellevik, et al. Washington State Patrol Larry J. Quamme, Director Crime Information Bureau Division of Law Enforcement Services Wisconsin Department of Justice Robert E. Olsen Criminal Identification Division Office of the Attorney General

PAGE 202

Appendix D State Officials Contacted in OTA Followup Surve y Conducted in August 1982 Arizona Arkansas California Colorado Connecticut Delaware Florida Georgia The following individuals were contacted by telephone: State Name and agency Alabama Ron Cunningham, CCH Project Leader Alabama Bureau of Investigation Alaska Dick Carpenter Alaska Justice Information System Marilyn Crenshaw Records and Identification Section Alaska Department of Public Safety D. C. Britt, Manager Arizona Criminal Justice Information System David Eberdt, Director Arkansas Crime Justice Information System Vic Paradis and Nell Hutchinson California Department of Justice W. Gray Buckley Colorado Bureau of Investigation Connecticut State Police Joseph Halloran, Supervisor Bureau of Identification Lt. Neal Tilghman, Systems Analyst Bureau of Identification Delaware State Police Richie Grant Criminal Justice Information Systems Florida Department of Law Enforcement E. W. Manseau, Deputy Director Georgia Crime Information Center State Hawaii Idaho Illinois Indiana Iowa Kansas Kentucky Louisiana Maine Maryland Name and agency Liane Moriyama, Systems Analyst Criminal Justice Data Center Hawaii Attorney Generals Office Richard Burns, Chief Criminal Identification Bureau Idaho Department of Law Enforcement John T. Loverude, Assistant Chief Bureau of Identification Illinois Department of Law Enforcement Stella Hanley, Program Coordinator Central Records Indiana State Police G. W. Shanahan, Chief Criminal Investigation Division Iowa Department of Public Safety Michael E. Boyer Kansas Bureau of Investigation Gary Bush Criminal Justice Information Section Kentucky State Police Elizabeth Fahl Criminal Records Section Louisiana Bureau of Criminal Identification Bob Wagner, Director Maine Bureau of Identification Lamar Edwards, Director CriminaI Records Central Repository Maryland State Police 197

PAGE 203

198 An Assessment of Alternatives for a National Computerized Criminal History System State Massachusetts Michigan Minnesota Mississippi Missouri Montana Nebraska Nevada New Hampshire New Jersey New Mexico Name and agency Louis Sakin, Executive Director Criminal History Systems Board Dallas G. Piper, Administrator Identification Division Michigan State Police Kenneth Bentfield, Director Criminal Justice Information Systems Section Bureau of Criminal Apprehension Minnesota Department of Public Safety Dewey Weems, Director Records and Identification Criminal Investigative Bureau Mississippi Department of Public Safety Robert Bradley, Director Information Systems Division Missouri State Highway Patrol Herb Bruning, Chief Identification Bureau Montana Department of Justice Lt. Greg Schnasse Identification and Records Section Nebraska State Patrol Frank Adams Division of Investigation Nevada Department of Motor Vehicles Lt. Lynn Presby New Hampshire State Police Wallace P. Miller Records and Identification Section New Jersey State Police Capt. M. K. Alexander, Commander Records Division New Mexico State Police State New York North Carolina North Dakota Ohio Oklahoma Oregon Pennsylvania Puerto Rico Rhode Island South Carolina Name and agency Adam DAlessandro Deputy Commissioner New York State Division of Criminal Justice Services William C. Corley, Director Police Information Network North Carolina Department of Justice Richard S. Hilde, Chief Agent Bureau of Criminal Identification North Dakota Attorney Generals Office Marlin Dowler, Chief Identification Bureau Ohio Attorney Generals Office Paul Boyd, Chief Identification Division Oklahoma State Bureau of Investigation W. H. Freele, Director Bureau of Criminal Identification Oregon State Police Lt. Terry Clemens Records and Identification Division Pennsylvania State Police Luis M. Gonzalez, Acting Administrative Director Sgt. Rivera Criminal Justice Information System Puerto Rico Department of Justice Jack Craven, Chief Criminal Identification Bureau Rhode Island Attorney Generals Department Shelby Cote Systems and Program Manager South Carolina Division of Law Enforcement Ernest Euler Office of the Governor South Carolina Division of Public Safety

PAGE 204

. App. DState Officials Contacted in OTA Fe/low-Up Survey Conducted in August 1982 l 199 Texas Utah Vermont State Name and agency South Dakota Donald G. Licht, Director Division of Criminal Investigation South Dakota Attorney Generals Office Tennessee Arso Carson, Director Douglas Woodlee, Special Agent in Charge Criminal Records Unit Tennessee Bureau of Investigation Steve Elliot, Manager Fingerprint Records Bureau Texas Department of Public Safety Del Mortensen, Deputy Commissioner Dan Taylor, Acting Chief Bureau of Criminal Identification Utah Department of Public Safety Billy J. Chilton, Director Vermont Criminal Information Center State Name and agency Virginia Capt. W. R. Wagner, Virginia State Police Washington George B. Televik, Commander Jr Criminal Records Division Washington State Patrol West Virginia Lt. Flanagan, Communications Division Lt. Atkinson, Records Officer Bureau of Criminal Identification West Virginia Department of Public Safety Wisconsin Bob McGrath, Director Crime Information Bureau Division of Law Enforcement Services Wisconsin Department of Justice Wyoming David G. Hall, Director Division of Criminal Identification Wyoming Attorney Generals Office

PAGE 205

Index

PAGE 206

Inde x Alameda County (California) Criminal Oriented Records Production Unified System (CORPUS), 179-180 American Bar Association, 139 American Bar Foundation, 129 American Civil Liberties Union, 53, 180 Arkansas, 64 Arizona, 36, 102 Attorney General (U.S.), 31, 34, 35, 61, 74, 95 authority to operate Ident and NCIC, 61, 62 and message switching controversy, 53 Task Force on Violent Crime, 5, 96 AUTODIN network (see Department of Defense) Automated Identification Division System (AIDS), 8, 94 advantages of automation, 9 computer hardware used by, 54 development phases of, 44 as effort to automate the FBI Ident system, 39, 109 format of records in, 22, 26 Jet Propulsion Laboratory study of, 9-10, 161-162 number of records in, 3 proposed consolidation with National Crime Information Center, 18, 45, 168-169, 186-188 Bertillon system, 27 California, 36, 62, 82, 84, 154 case disposition reporting in, 94 fees for noncriminal justice access to CCH files, 141 improvements in on-line CCH files, 163 licensing boards authorized to use State criminal history information, 139 rate of arrest of minority group members, 137-138, 141, 142 response time on CCH inquiries, 162 retention period for CCH records, 174 Canada, 40, 82 CCH (see computerized criminal history systems) Central Intelligence Agency (CIA), 81 Coast Guard, 45 Colorado, 120, 162 Commercial Carrier Corp., 85 Commission on Criminal Justice Standards and Goals, 128 Comptroller General, 178 computerized criminal history (CCH) systems current status of, 3-8 evolution of, 31-36 Federal access to State systems, 78, 81-82 at the Federal level, 3-4 at the non-Federal level, 4, 46-48 Federal and State court rulings on, 67-69 Federal statutes and regulations, 62-65 legal and regulatory framework for, 61-74 national index for, 9 national system (see national CCH system) noncriminal justice access to, 17-18 origins of the national program, 35-36 oversight and audit of, 18 police departments with, 47 policy control over, 16-17 preferences of State repository officials, 10-11 public participation in governance of, 18 response time of, 9-10 record quality of, 10 recommendation of the Presidents Commission on Law Enforcement and Administration of Justice, 33 record quality of, 17, 89-96 repositories for, 46-47 State statutes and regulations on, 69-73 State and local management of, 99-105 users of, 77-85 Congress acts of (see legislation) hearings on privacy and security legislation, 73-74 House Judiciary Committee, 16, 53, 157 House Judiciary Subcommittee on Civil and Constitutional Rights, 127 policy options for (see policy considerations) prohibition against message switching, 51, 53 Senate Judiciary Committee, 5, 16, 53, 74, 157 Connecticut, 36 Constitution (U.S.), 14, 63, 148, 149 court rulings (see litigation) criminal history records access, review, and challenge procedures for State and local, 99, 102 arrest and court disposition reporting in State and local, 99, 100-102 correctional and parole use of, 126, 133-134 court disposition monitoring in State, 104-105, 154 criminal justice use of, 4-5 noncriminal justice use of, 5, 17-18 defense use of, 126, 132-133 dissemination of information in State and local, 99, 103 Federal statutes and regulations on, 62-65 file content of State and local, 99, 102 interstate exchange of, 4, 16 judicial use of, 131-132 locus of authority for State and local, 99, 100 nature of, 21-26 origins of, 3, 26-28 police use of, 125, 128-130 privacy and security protection, 8 private users of State systems, 85 probation use of, 126, 133 problems with manual files, 31-33 prosecutorial use of, 125-126, 130-131 quality of, 6-8, 104 203

PAGE 207

204 An Assessment of Alternatives for a National Computerized Criminal History System recommendations of the Presidents Commission on Law Enforcement and Administration of Justice, 33 repositories for, 3-4, 8-9, 23, 46 sealing and purging of State and local, 64, 99, 103-104 transaction logs and local audits, 105 criminal justice agency, 79 Delaware frequency of file access requests in, 102 response time on CCH inquiries, 162 Department of Defense AUTODIN network, 46 use of Ident services by, 80, 81 use of State criminal history information by, 82 Department of Justice, 8, 18, 27, 65, 74, 79 Bureau of Justice Statistics, 71 Justice Telecommunications System (JUST), 40, 46, 147 prohibited from engaging in message switching, 16, 53 regulations on Criminal Justice Information Systems, 79 Department of State Diplomatic Network, 46 use of Ident services by, 81 Department of Treasury, 8 Bureau of Alcohol, Tobacco, and Firearms, 45, 81 Bureau of Customs, 45, 79 Secret Service, 79, 81 Treasury Enforcement Communication System, 40, 45, 79, 147 use of Ident services by, 81 use of NCIC services by, 147 Diplomatic Network (see Department of State) District of Columbia, 40, 64, 90, 117, 130, 132 Edwards. Rep. Don, 74 Ervin, Sen. Sam, Jr., 74, 181 Executive Orders No. 10450, 14, 61, 64, 81 No. 10865, 14, 64 No. 12065, 14, 64 Federal Bureau of Investigation (FBI), 3, 6, 9, 51, 79, 81 assistance to States, 14 authority to upgrade NCIC communications controller, 51 authority to operate Ident and NCIC, 61, 62 challenges to records of, 65 COINTELPRO program, 146 COMINFIL program, 146 computerization of criminal records, 31 Computerized Criminal History Program (CCH), 36 contribution to NCIC operating costs, 55-56 Crime Index, 17 criminal history dissemination by, 64-65 exemption of criminal records from public disclosure under the Privacy Act, 65 expunging and sealing of records by, 63 first use of computer technology by, 34 Identification Division (see Ident) and message switching controversy, 53 National Crime Information Center (see NCIC) position on automatic inquiry referral (AIR), 115-116 role in fingerprint identification, 5 standards for use of criminal justice information systems, 61 surveillance activities of, 15-16 survey of State criminal justice record repositories, 46-48 Technical Services Division, 55 Uniform Criminal Records (UCR) Division, 34 use of Ident services by, 81 fingerprints, 5-6 disposition in pretrial diversion cases, 63 facsimile electronic transmission of, 9-10 Ident files, 5, 31, 83, 109, 111 National Fingerprint File (NFF) proposal, 10, 155, 164 processing time for, 5, 9-19, 32, 162 reliability of, 5 State contributions to record repositories for, 46 Florida, 36, 116, 117, 118, 184 Crime Information Center, 85 Department of Law Enforcement, 5, 127 fees for noncriminal justice access to CCH files, 141, 143 frequency of file access requests in, 102 Public Records Statute, 85, 143 response time on CCH inquiries, 162 test of single-State-multi-State concept in, 113 update time on CCH system, 175 General Accounting Office (GAO), 6, 80 audits to ensure compliance with prohibition against message switching, 51, 53 audits of Federal and State agencies, 148 study of State criminal history files, 93 General Telephone Co., 85 handguns, 5 Hoover, J. Edgar, 34 Hruska, Sen. Roman, 73, 74 Hughes Refrigerated Express, Inc., 85 IBM, 52, 54 Idaho, 47 Ident (FBI Identification Division), 3, 5, 6, 7, 8, 9, 21, 22, 55 authority to operate, 61, 62 background and security checks, 81 content of records, 62-63 creation by Congress, 27 criminal justice use of, 80

PAGE 208

Index l 20 5 noncriminal justice use of, 80-81 differences between NCIC and, 111 Federal Government users of, 77-78, 80-81 fingerprint file, 4-5, 44, 80, 109 format of records in, 22, 24 interconnection with NCIC files, 39-40, 44-45 international users of, 78, 84-85 Jet Propulsion Laboratory study of, 89 multi-State offenders in criminal file, 31 as a national repository, 109, 110, 111-112 percentage of multi-State offenders in files of, 33 private users of, 78, 85 processing time for fingerprints, 9-10 record quality, 89 record updating, 63-64 right of individuals to review and challenge records, 65 State and local users of, 78, 82-83, 99 III (see Interstate Identification Index) Illinois, 82, 161 Immigration and Naturalization Service, 45, 81 Internal Revenue Service, 45 Criminal Investigative Division, 147 use of Ident services by, 81 use of NCIC services by, 147 International Association for Identification, 6 International Association of Chiefs of Police (IACP), 27, 35 International Police Association (Interpol), 45, 84 Interstate Identification Index (III), 4, 5, 45 development plan, 18 file size and content, 17 officials preferences regarding, 10, 17, 110, 118 National Fingerprint File (NFF) proposal for, 10-11, 118, 164 national index system, 110, 118-119 pilot and Phase 1 test of, 110, 114-117 potential response times of, 10 regional and ask-the-network systems, 110, 119-121 Institute for Law and Social Research (INSLAW), 53, 57 Itel Corp., 52 Jacks Cookie Co., 85 Jet Propulsion Laboratory (JPL), 6 study of Automated Identification Division System (AIDS), 9-10, 161-162 study of Ident record quality, 89 Johnson, President Lyndon B., 33 Justice Telecommunications System (JUST), 40, 46 Kansas, 44, 175 Kansas City (Kansas) Alert 11 regional system, 69 Kennedy, Sen. Edward, 70, 73 Law Enforcement Assistance Administration (LEAA), 6 assistance for State computerized systems, 14-15, 31, 34, 51-52, 55, 56-57, 154, 158 demonstration project on interstate exchange of criminal offender files, 36 establishment of, 31, 33-34 and message switching controversy, 53 regulation of State and local CCH systems, 61, 69-71, 81-82 legislation Crime Control Act, 6, 8, 61 Criminal Justice Information Control and Protection of Privacy Act, 62, 74 Criminal Justice Information Systems Act, 74 Criminal Justice Information Systems Security and Privacy Act, 62, 73 Departments of State, Justice, and Commerce, the Judiciary, and Related Agencies Appropriation Act, 64 Department of Justice Appropriation Authorization Act, 53, 74, 115, 157 Florida State Public Records Statute, 85 Freedom of Information Act, 65 Justice Systems Improvement Act, 94 Omnibus Crime Control and Safe Streets Act, 31, 33, 35, 62, 70, 73 need for, 18 Privacy Act, 65, 81, 182, 183 Securities Exchange Act, 64 South Carolina Freedom of Information Act, 85 Title 28, United States Code, sec. 534, 61 litigation Green v. Missouri Pacific RR 1 3 Gregory v. Litton Systems, 7, 13, 138 illustrative Federal/State rulings on criminal records (table), 68 Long v. U.S. IRS 65 Menard v. Mitchell 64, 65, 68, 73 Menard v. Saxbe, 62-63, 68 primary focus of State and Federal rulings, 67 Tarleton v. Saxbe, 6, 63 Tatum v. Rogers, 7, 11, 94, 130, 131, 148 Maine, 47, 141 Maryland, 36, 47 Massachusetts, 47, 143 Mathias, Sen. Charles McC., Jr., 73 message switching alternatives for, 16 American Civil Liberties Union and controversy over, 53 Attorney General (U. S.) and controversy over, 53 automatic inquiry referral (AIR) and, 114-117 Department of Justice prohibited from engaging in, 16, 53 Congress and controversy over, 51, 53 definition of, 16 FBI and controversy over, 53, 109 GAO audits to ensure compliance with prohibition against, 51, 53 LEAA and controversy over, 53 opportunity for, provided by modern computers, 51, 52-54

PAGE 209

206 l An Assessment of Alternatives for a National Computerized Criminal History System prohibition against, 16, 51, 53, 109 Michigan, 115 Minnesota, 36, 154 frequency of file access requests in, 102 improvements on on-line CCH files, 163 progress in court reform, 159 response time on CCH inquiries, 162 update time for CCH files, 176 multi-State offenders, 31, 33 National Advisory Commission on Criminal Justice Standards and Goals, 69-70, 74 National Association for the Advancement of Colored People, 180 National Center for State Courts (NCSC), 57 National Chiefs of Police Union (see International Association of Chiefs of Police) National CCH system alternatives for, 108, 109-121 cost of, 14-15 impact on constitutional rights, 138, 147-149 impact on the criminal justice process, 11-12, 125-134 impact on employment and licensure decisions, 12-13, 137, 139-141 impact on Federal-State relations, 14, 16, 138, 142-144 impact on minority groups, 13-14, 137-138, 141-142 Interstate Identification Index pilot tests, 110, 114-117 national index for, 110, 118-119 national repository for, 109, 111 policy considerations (see policy considerations) possible impacts on the criminal justice process, 125-134 regional and ask-the-network systems, 110, 119-121 single-State/multi-State alternative, 109, 112-114 support for, 127-128 surveillance potential of, 15-16, 138, 145-147 National Crime Information Center (NCIC), 4, 6, 8, 22 Advisory Policy Board, 66, 67, 79, 95, 115, 168 communications controller, 51, 53 Comprehensive Data System program, 69, 70 computer hardware previously used by, 52 computer hardware upgrade, 52-54 computerized criminal history file, 31, 39, 42, 43, 66-67 content of records, 62-63 contextual description of, 39-48 costs to States, 56-58 costs of operation, 51-52, 55-58 Criminalistics Laboratory Information System (CLIS) file, 42, 39 establishment of, 34-35 Federal agency orders and procedures for using, 67-69 FBI authority to operate, 61, 62 FBI use for intelligence purposes, 146 Federal Government users of, 77, 79-80 as a Federal repository with national access, 109, 111-112 hot files of, 39, 42, 43, 56, 66, 127, 153, 191-192 information exchange via, 9 initial participating agencies for (table), 35 interconnections with other criminal justice information systems, 39-40, 42, 44-48, 146-147 international users of, 78 Interstate Identification Index (III), 4 and message switching controversy, 52-53 number of records in files of (table), 42 operating policies and procedures, 66-67 percentage of multi-State offenders in files of, 33 position on automatic inquiry referral (AIR), 117 private users of, 85 processing time for inquiries, 10 proposed consolidation of files with Automated Identification Division System, 18 quality of records, 89 record updating, 63-64 right of individuals to review and challenge records, 65 State and Federal agencies linked to (figure), 41 State and local users of, 78, 82, 99 system description, 40-42 system security, 67 technology used by, 51, 52-54 volume of transactions, 43-44 National Criminal Information and Identification Division, 18 National Law Enforcement Telecommunications System (NLETS), 4, 8, 84 Board of Directors, 115, 117 capacity of, 120 general description of, 40 information exchange via, 9 interconnection with NCIC files, 40, 45 and Interstate Identification Index (III) concept, 45 position on automatic inquiry referral (AIR), 117 National Legal Aid and Defender Association (NLADA), 132-133 NCIC (see National Crime Information Center) NCSC (see National Center for State Courts) Nebraska, 141 New Hampshire noncomputerized criminal history system, 175 response time on CCH inquiries, 162 New Jersey, 162 New Mexico, 47, 162 New York City, 26 New York State, 7, 36 fees for noncriminal justice access to CCH system, 141 purge and seal policies of, 104 rate of arrest of minority group members, 141 response time on CCH inquiries, 162 use of high-quality facsimile electronic transmission, 161

PAGE 210

Index l 20 7 NLETS (see National Law Enforcement Telecommunications System) North Carolina, 154 case disposition reporting in, 94 improvements in on-line CCH files, 163 Police Information Network, 163 test of Interstate Identification Index concept in, 115 Nuclear Regulatory Commission, 81 offender-based transaction systems, 57-58 operation costs by State (table), 58 Office of Management and Budget (OMB), 36, 53 Office of Personnel Management (OPM), 61, 81 Office of Technology Assessment (OTA), 5, 6, 10, 31 estimate of number of U.S. citizens with criminal records, 13 findings on multi-State offenders, 5 research on record quality, 8, 9, 89-96, 121, 153 survey on Federal agency policies on using NCIC, 67 survey on fingerprint card submission, 11 survey of officials preferences, 10-11, 119 survey on multi-State offenders, 33 survey on State and local CCH systems, 47-48, 84, 99-105, 128, 140-141, 193-194, 197-199 survey of State repository officials, 195-196 evaluation of NCIC request for new communications controller, 53 Ohio, 162 Oregon, 82 frequency of file access requests in, 102 update time for CCH files, 176 OTA (see Office of Technology Assessment) overcriminalization, 33 Pennsylvania, 47, 162 Philadelphia, 141 Phoenix, 120 police blotters (see Criminal history records) policy considerations, 17-18, 153-188 arguments for action or no action, 153, 155-156 comprehensive legislation, 18, 168, 180-184 consolidation of Automated Identification Division System (AIDS) and National Crime Information Center (NCIC), 18-19, 168-169, 186-188 file size and content, 17, 167, 173-174 further study on use, quality, and cost of criminal history information, 153-154, 157 importance of complete, timely, and verifiable information, 161-162 improving record quality, 162-164 Interstate Identification Index (III) development plan, 18, 168, 184-186 noncriminal justice access, 17-18, 168, 176-177 oversight and audit, 18, 168, 177-178 policy control, 17, 167, 169-172 possible impacts on the criminal justice process, 161-164 private carrier role in national CCH system, 188 public participation, 18, 168, 179-180 record quality, 17, 167, 174-176 select a national CCH system, 154-155, 160-161 shifting preferences on a national CCH system, 164 strengthen State/local CCH systems, 154, 158-159 Postal Service, 79 Presidents Commission on Law Enforcement and Administration of Justice, 33 Pretrial diversion, 63 Privacy and security Criminal Justice Information Systems Security and Privacy Act, 62, 73 development of standards for, 69-70 implementation of standards for, 70-71 Project SEARCH (System for Electronic Analysis and Retrieval of Criminal Histories), 31 early efforts to develop guideline for State and local CCH systems, 69-70 States originally participating in, 36 role in development of State statutes and regulations on State criminal record systems, 61, 74 Puerto Rico, 40, 90 rap sheets (see criminal history records) record quality research methodology, 89-91 findings, 91-96 Rinker Materials Corp., 85 Royal Canadian Mounted Police (RCMP), 84 Ryder Truck Lines, 85 Santa Clara County (California), 69 SEARCH Group, Inc., 10, 147, 148 findings on release of arrest and conviction data to private employers, 139 and message switching controversy, 53 technical assistance by, 158 security (see privacy and security) South Carolina response time on CCH inquiries, 162 State Freedom of Information Act, 85 test of Interstate Identification Index concept in, 115 South Dakota, 47 Supreme Court (U.S.), 69 Texas, 36 test of Interstate Identification Index concept in, 115 update time for CCH files, 176 Treasury Enforcement Communications System (TECS) agencies served by, 45-46 interconnection with NCIC fries, 40, 45 use by Interpol, 84 Tunney, Sen. John, 74

PAGE 211

208 An Assessment of Alternatives for a National Computerized Criminal History System United Parcel Service, 85 U.S. Court of Appeals, District of Columbia, 6 U.S. Parole Commission, 133 Utah, 176 Vermont frequency of file access requests in, 102 noncomputerized criminal history system in, 175 Virginia frequency of file access requests in, 102 test of Interstate Identification Index concept in, 115 update time for CCH files, 176 Virgin Islands, 40 Washington, D. C., 120 Washington (State), 36, 47, 176 West Virginia fees for noncriminal justice access to CCH files, 141 response time on CCH inquiries, 162 Winn Dixie Stores, Inc., 85 Wisconsin, 176 Wyoming, 162, 176


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EDR69SSCL_4N6SLF INGEST_TIME 2017-05-22T19:40:05Z PACKAGE AA00055244_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES