December 1985 Bulletin 858 (technical)
The Determinants of
Motor Carrier Freight Rates for
Florida Perishables
Richard Beilock
Agricultural Experiment Stations
Institute of Food and Agricultural Scien es
University of Florida, Gainesville
December 1985
Bulletin 858 (technical)
The Determinants of Motor Carrier Freight Rates
for Florida Perishables
by
Richard Beilock
The author is an Assistant Professor of Food and Resource
Economics. Address: G091 McCarty Hall, University of Florida,
Gainesville, Florida 32611.
TABLE OF CONTENTS
Page
LIST OF TABLES............................................ iii
LIST OF FIGURES......................................... iii
ABSTRACT.... ............................................ iv
REVIEW OF LITERATURE ..................................... 2
How rates change over distance....................... 5
Responsiveness of rates to number of pickups and
drops, carrier type and method of arrangement......... 9
Influence of cargo value and shipping urgency
on rates............................................. 10
EMPIRICAL MODEL .......................................... 17
DATA...................................................... 19
THE RESULTS............................... ........ ...... 20
SUMMARY AND CONCLUSIONS ................................. 27
FOOTNOTES...................................... ......... 29
APPENDIX 1...... ......................................... 31
APPENDIX 2............................................... 34
REFERENCES.............................................. 36
LIST OF TABLES
Table Page
1. Total claims paid by motor common carriers
of general freight................................ 14
2. Net loss and damage costs of motor common
carriers of general freight....................... 14
3. Carrier responses to March and June
1983 survey................... .................. .. 21
4. Rate estimation equations dependent variable:
truckload freight rate........................... 25
LIST OF FIGURES
Figure
1. Flat, linear, and tapering rate -distance
relationships..................................... 6
2. Representation of price discrimination
resulting from rate tapers across distance......... 8
ABSTRACT
Freight rates for perishables have long been observed to
be extremely variable both across time and within the same time
period. Several reasons for this variation have been hypothe-
sized, little research has been devoted to the subject area.
This study examines the roles of load and shipment characteris-
tics, carrier status, and methods of arranging for carriage on
rates. The results reveal that load and shipment characteris-
tics are the dominant factors in rate determination. Of par-
ticular interest is the discovery that the rate structure re-
sembles the value of service or commodity value-correlated rate
structures commonly found in regulated carriage.
Keywords: Exempt trucking, Florida, freight rates.
The Determinants of Motor Carrier Freight Rates
for Florida Perishables
by
Richard Beilock
Trucking is the primary mode for shipping Florida perisha-
bles. During the shipping season, over a thousand trucks a day
may leave Florida, each with a unique destination and carrying
one or more of a variety of perishable agricultural products.
Carriers, receivers, shippers and truck brokers are involved in
the markets for transportation services. All parties interact
in a setting free from government regulation of rates, entry,
or service quality.
Little is known regarding the factors which influence rate
differentials between loads, both within the same time period
and across time. Such information is of importance on several
levels. First, a better understanding of the ratemaking pro-
cess would help shippers and carriers to adjust their opera-
tions to enhance their revenues. Second, it would be of value
to economists and policymakers in assessing the probable im-
pacts of policies relating to the marketing and distribution of
perishables. Finally, information about this transportation
market may provide insights into other transportation markets.
In particular, the workings of an unregulated, competitively
structured transportation market, such as that for Florida
perishables, may indicate how other still regulated or recently
deregulated transportation markets will evolve in the absence
of government intervention.
This study investigates the effects of carrier status,
method of load arrangement, and load and shipment characteris-
tics on freight rates. The specific objectives are to deter-
mine:
1. how rates change with the distance of haul.
2. if rates are sensitive to the number of pickups and
drops, and if carrier type and/or the method of arranging
carriage influences rate levels; and
3. if rates are influenced by cargo value and shipping
urgency.
In the next section of the bulletin the literature regard-
ing the three objectives will be briefly reviewed. The purpose
is twofold. First, the section will establish the current
level of knowledge regarding the issues addressed in the objec-
tives. Second, the importance of these questions will be made
clearer. In the next section the data employed in the study
will be described. This is followed by a description of the
methodology adopted for the empirical estimation. Finally, the
results are presented, and conclusions are drawn.
REVIEW OF LITERATURE
The literature addressing rate determination in transpor-
tation and, in particular, in unregulated markets is considera-
ble but disappointing, at least with regard to empirical work
addressing the concerns of this study. Due to difficulties
obtaining data regarding unregulated markets and to the domi-
nance of regulated freight in the overall transportation pic-
ture, most rate studies have concentrated on regulated markets
(e.g. Benishay and Whitaker, Oum, Sugrue et al., Spady and
Friedlaender, and Cherry and Garston). Most of the studies
employed highly aggregated data which precluded detailed analy-
sis of rate differentials due to load and shipment characteris-
tics, carrier type and method of arranging loads.
In the agricultural literature the picture is only some-
what brighter. The majority of the transportation work has
focused on grain and soybean movements, rather than perisha-
bles. Again the majority of the studies have focused on regu-
lated rail movements of these goods and the data have been
highly aggregated. Moreover, when dealing with questions re-
lated to grain/soybean transport via the unregulated barge and
truck modes, researchers have typically assumed rates to equal
the estimated accounting costs of an "average" load (e.g. Sor-
enson, Martin, Michaels et al., and Fuller and Shanmaugham).
Only one study in grain transport (Johnson) explored the role
of service quality differentials.
Empirical research on the transportation of perishables
has emanated primarily from three sources: the USDA, Washing-
ton State University/University of Hawaii, and the University
of Florida. The USDA has conducted several studies. However,
they have primarily employed a costing approach which ignores
individual shipment and load characteristics (e.g. Klindworth
and Brooks and Boles).
Of more direct bearing is the work done by a group of
researchers at Washington State University and the University
of Hawaii. In a series of articles spanning almost 20 years,
they have explored the intensity of intermodal competition for
perishables (Miklius), carrier and trip characteristics which
determine whether exempt loads are acquired (Pederson, Mittle-
hammer, and Casavant), and service characteristics which influ-
ence the demand for transportation of perishables (Miklius;
Miklius, Casavant and Garrod; Miklius and Casavant; and Garrod
and Miklius). It is clear from their work that the demand for
perishables transport is highly sensitive to service character-
istics. Further, the degree and even the direction of this
sensitivity vary from commodity to commodity. For example,
Miklius, Casavant, and Garrod found that for shippers longer
transit times for apples are often desirable, whereas speed is
always sought when transporting cherries. The reasons cited by
the authors are as follows. Apples are a long-lived fruit with
relatively low per unit value. A lengthy transit time may be
advantageous as it provides more storage-in-transit. Cherries,
on the other hand, are valuable and highly perishable. Slow
transit is costly, therefore, both because of high carrying
(interest) costs associated with the cargo and losses due to
spoilage. Therefore, disaggregation of data to preserve the
identity of loads of different types with differing service
characteristics is important in perishables transportation
research.
Research by Beilock and Shonkwiler, and Beilock, Kohbur-
ger, and Morgan) at the University of Florida has reaffirmed
the importance of taking into account differences in load
types. Rates for Florida tomatoes, grapefruit, and sweetcorn
and for California lettuce and citrus were found to vary both
with respect to average levels and to fluctuations over time.
This work, however, used data which did not allow for differ-
ences in transportation service characteristics, load arrange-
ment, or carrier type.
At the Transportation Research Center of the University of
Florida's Department of Civil Engineering, a study of Florida's
perishable transport system was carried out in 1980 and 1981.
Their results indicate the following:
1. carriers are not compensated for making additional pick-
ups and drops, and
2. owner-operators tend to get high cost-low net revenue
loads, while fleet carriers typically are able to secure
low cost-high cost revenue loads (Maze and Pavlovic et
al.).
These results are surprising, considering the atomistic struc-
ture of the industry and the apparent ease and magnitude of
information flows. When the State of Florida last registered
carriers of perishables (1979) there were over 20,000 carriers
and over 200 brokers. Moreover, the Florida Department of
Agriculture and Consumer Services lists in excess of 2,000
active receivers of Florida perishables (perishables are typi-
cally sold f.o.b. origin, with the receiver paying for the
freight). Information regarding market conditions is constant-
ly transmitted via telephone, at truck stops and over citizens'
band radios. Under such near-classical competitive conditions
one would not expect chronic nonpayment for additional services
(pickups and drops) or that one type of carrier would be con-
sistently favored over another due to the ability to secure
more complete market information.
The conclusions reached by the Transportation Research
Center study team were arrived at largely from informal discus-
sions with carriers, brokers, and receivers rather than by the
use of statistical techniques. The theoretically surprising
nature of their conclusions led to a reexamination of these
issues in this current study.
In the remainder of this section, largely nonempirical
work addressing the three objectives of this study is dis-
cussed. This is done to provide a rationale for the approach
which was taken and to familiarize the reader with the reasons
why these issues are of concern.
1. How rates change over distance
In Figure 1 are presented three types of distance-rate
relationships: flat, linear, and less-than proportional or
tapering. Flat schedules are usually attributable to some form
of institutional influence. The U.S. Mail First Class Service
is a prime example. Letters are transported anywhere in the
nation for the same charge. Linear schedules are what one
would expect to find. Once the "fixed" or distance-unrelated
costs, such as pickup and delivery, are accounted for (in the
intercept) there may be little reason to suppose that driving
the 100th mile would be any less costly than the 1000th mile.
However, several researchers have noted that rates for regu-
lated motor carriers exhibit tapers over distance (Bresseler
and King, Pegrum, and Lockin). There are three principal rea-
sons for explaining rate tapers. First, different equipment
may be employed for hauls of different lengths (Bresseler and
Freight
Rate
Linear
Tapering
Flat
Distance
Figure l.--Flat, linear, and tapering rate-distance relationships.
King). If larger, more efficient equipment is employed for
longer hauls then tapering rates would not be surprising.
A second explanation for tapering rates is price discrimi-
nation (Phlips). In Figure 2, two demand schedules for pickup
and delivery services are shown, DL and DS. The vertical dis-
tance between DL and DS represents the additional costs of
transportation to a more remote point. In other words the
demand for pickup and delivery services is depicted as a demand
derived from the demand for the entire transportation service
(which is assumed to be identical for long and short hauls).
Assuming constant marginal costs (MC) for pickup and delivery
services for long and short hauls, the profit maximizing rates
for pickup and delivery services for long and short hauls are
PL and PS, respectively. As MC is the same for both long and
short hauls, short haul shippers are being charged a higher
premium over costs than are long haul shippers, that is, there
is price discrimination. It should be noted that this explana-
tion rests on some fairly restrictive assumptions concerning
elasticities of demand for transportation services at different
points. Moreover, price discrimination would be difficult or
impossible in an atomistically structured market such as that
for trucking services for produce.
The final explanation for tapering rates is that such
rates are similar to volume discounts common for several, if
not most goods and services. Search and repositioning costs
between loads may be considerable. Longer trips naturally mean
fewer searches and repositionings per unit time.
If tapering rates are discovered in the empirical portion
of this study, it is felt that the last explanation would be
the most reasonable. The first explanation is invalid as the
same type of equipment is normally employed over all distances.
The second explanation is not feasible considering the competi-
tive structure of the industry.
D = demand for pickup and
delivery services
P = price for pickup and
delivery
MR = marginal revenue
MC = marginal cost for pickup
and delivery
L,s = subscripts denoting
long and short hauls,
respectively
S\D
\ s
Figure 2.--Representation of price discrimination resulting from
rate tapers across distance. (Adapted from Phlips,
Chapter 2.)
2. Responsiveness of rates to number of pickups and drops,
carrier type, and method of arrangement
In a competitive environment there should be little reason
to hypothesize that additional services, such as multiple pick-
ups and drops, would not command higher rates; or that rates
should differ for comparable services depending upon carrier
type. However, the aforementioned conclusions reached by Maze
and Pavlovic et al. that these conditions do exist raise these
questions. Moreover, there are considerable concerns through-
out the industry that the method of arranging for carriage is
not neutral with respect to rate levels. Of particular inter-
est is the role of brokers. Brokers are extremely important in
Florida (estimated by Pavlovic et al. and by Beilock and
Fletcher to account for between 55 and 65 percent of all per-
ishable shipments). Carriers interviewed by the author have
expressed their feelings that brokers affect rate levels. Such
sentiments are also echoed in a study by Taff. Taff reports
that brokers actively set rates using various devices such as
rate sheets and minimum per mile charges. No actual evidence,
however, is presented to support the hypothesis that rates for
loads arranged by brokers differ from those which are otherwise
set.
The answers to the questions of the effects of pickups and
deliveries, carrier type, and method of load arrangement on
rate levels are important in assessing market efficiency. If
rates are not compensatory for additional services, such as
pickups and drops, then inefficiencies will likely occur as
shippers demand more service than is justified by the actual
additional costs (i.e., real marginal costs exceed the marginal
benefits). Such a situation could occur if truckers are re-
peatedly deceived into believing that shipments entail fewer
pickups and drops. Undoubtedly such circumstances do occur,
but the likelihood of chronic carrier underestimates of service
requirements seems small. Indeed, there would have to be ex-
tremely limited information flows and a fairly constant flow of
carriers ready to be taken advantage of by underestimated ser-
vice requirements. Poor information flows would also be indi-
cated if carrier type and the method of arranging loads affect-
ed rate levels. Such indications of information problems would
cast doubts on the advisability of deregulation in other trans-
portation markets. Indeed, the contention that information
costs are high in the absence of rate and entry regulation has
been a major argument of those opposing economic deregulation.
However, as information appears to be readily available to all
participants, it is strongly anticipated that these indications
of information problems will not materialize in the analysis.
3. Influence of cargo value and shipping urgency on rates
A rate structure which is positively correlated with the
value of the cargo is known as value-of-service pricing, Eco-
nomists have traditionally explained the existence of value-of-
service pricing in terms of price discrimination.
To the economist, value-of-service pricing means
price discrimination. A firm with monopoly power
and the ability to separate demand for its product
into separate markets can increase its profits by
charging higher rates to its customers with less
elastic demands. (Olson, p. 395)
Blame for the existence of such rate structures has almost
universally been placed at the feet of state and federal regu-
latory systems. Indeed these regulatory agencies traditionally
have included cargo value as an allowable criterion for pricing
and have promoted centers of legalized price collusion (rate
bureaus)1 which facilitate pricing discipline in the industry.
Typically, regulators have defended this practice on the
grounds that it is necessary to ensure carrier viability or to
further social goals (such as subsidizing raw materials
freight). However, value-of-service pricing and the regulators
who supposedly condone and support it have undergone increasing
criticism, as reported by Wilson (pp. 138-9):
As the ICC noted in its first annual report: "It
was seen not to be unjust to apportion the whole
cost of service among all the articles transported
upon a basis that should consider the relative
value of service more than the relative cost."
Note that this relative downplaying of cost, pre-
sumably MC, emphasizes "justice"--not efficiency--
as the overriding criterion. Indeed, the commis-
sion has seldom stressed the latter. It is this
failure plus the persistent support of a non-cost
oriented rate structure that has resulted in gener-
ations of economists disparaging ICC regulations.
Recently, however, an alternative explanation has been
offered to the value-of-service as price discrimination argu-
ment. DeVany and Saving suggest that observed price differen-
tials could be explained in terms of compensation for the costs
of an unobserved service characteristic such as prompt, expe-
dited service. If this explanation does account for all or a
large proportion of value-of-service rate structures, such rate
structures would be expected to persist in the event of deregu-
lation. This has obvious policy implications in transportation
as it weakens one of the rationales for deregulation. Examina-
tion of a competitively structured, never regulated transporta-
tion market, such as the one examined in the current study
offers an excellent opportunity to investigate the question.
The explanation of value-of-service pricing as a form of
price discrimination will now be outlined. Begin by making the
following assumptions:
1. The quantity demanded of a good (QD) equals the quantity
transported (QT).
2. The selling price (PQ) equals the sum of the F.O.B. plant
price (PFOB) plus the transportation rate (PT) (i.e.,
there is no freight absorbtion, positive or negative),
and
3. A perfectly elastic supply (ES) exists for the commodity.
Under these conditions:
(1%AQD Q T
D %= APQ %A(PFB + PT)
where: ED = own price elasticity of demand for Q
%A = percentage change
Holding PFOB constant, ED may be rewritten as:
%AQT
(2) ED p
PT %AP
P T
Rearranging the terms in (2) yields
PT %&QT
(3) E
(3) ED = = ET
Q T
where ET = elasticity of demand for transportation.
Relaxing the assumption that E = i, it can be shown that:
P
(4) ET= R ED
where: R = 1 if ES = and
0 < R < 1 if ES < (see Appendix 1).
The important point regarding the above is that, ceteris
paribus, ET is absolutely smaller, the smaller is PT/PQ. Given
PT and reasonably small variations in ED's across commodities
or no strong negative relationship between PQ and ED across
commodities, shipper/receivers of higher valued commodities
would have less elastic demands for transportation than their
counterparts with lower valued commodities. Therefore, PQ may
be employed as a crude proxy for ET in facilitating price dis-
crimination schemes.
There are two major nondiscriminatory explanations of
value-of-service pricing. The first is that with different
cargo values, carriers incur different insurance costs. Car-
riers are liable for the goods they transport.2 The more valu-
able the cargo, the greater the liability. Therefore, it would
not be surprising to observe rate differentials commensurate
with increased insurance costs.
The insurance or liability argument is similar, in many
respects, to DeVany and Saving's expedited service argument
which will be discussed below. Both arguments are based on the
premise that the qualities or types of transportation services
for high and low valued goods are different, and that it is
these differences which are reflected in the price. However,
unlike prompt, expedited service, it is a fairly straight-
forward matter to measure costs resulting from damage claims
and insurance. In a recent study of general freight motor
carriers, Fauth develops such measures (Tables 1 and 2). Both
gross claims paid, and net claims and insurance costs are con-
sistently under two percent of operating revenue. Fauth,
therefore, concludes:
Statistics on loss and damage claims paid by carri-
ers indicate that such costs represent a very small
part of total freight charges and do not justify
commodity value as a major dimension of the rate
structure (p. 69).
In 1977 DeVany and Saving offered an alternative explana-
tion of value-of-service pricing. They argued that shipper/
receivers of urgently needed cargoes are more willing to pay
for prompt, readily available service than are those shipping
less urgent cargoes. For nonperishable goods the value of the
Table 1. Total claims paid by motor common carriers of general
freight.
Operating revenues Claims paid Claims as of
Year ($000's) ($000's) percent of revenue
1974
1975
1976
1977
Source:
12,170,387
11,481,875
13,299,567
14,989,850
Fauth
173,904
164,130
172,927
195,327
1.43
1.43
1.30
1.30
Table 2. Net claims and insurance costs as a percentage of
operating revenue of motor common carriers of general
freight.
Net claims paid and insurance costs
Quarter ended as percentage of operating revenue
March 31, 1975 1.55
December 31, 1975 1.30
June 30, 1976 1.31
September 30, 1976 1.17
December 31, 1976 1.24
June 30, 1977 1.19
September 30, 1977 1.33
Source: Fauth
_1________1___1____
cargo is a reasonable indicator of urgency, as the interest
costs associated with shipping delays vary directly with value.
To supply prompt, readily available service necessitates the
existence of excess capacity ready to load on demand. Relative
queue lengths (i.e., amounts of excess capacity) are regulated
because carriers will join longer queues only if the costs
associated with the additional expected wait are less than or
equal to the premium over the rates being offered at shorter
queues. Therefore, shipper/receivers requiring prompt service
must pay more to compensate carriers for the additional costs
associated with longer waits. As there is nothing in this
explanation which is at variance with the operation of competi-
tive markets, value-of-service pricing should be found in un-
regulated, as well as regulated markets.
Unfortunately it is rarely, if ever, possible to observe
such queues. These queues normally would not take the form of
lines outside a shippers' gates. Rather, vehicles would be
parked at their terminals, truck stops, or the operators'
homes. The researcher, then, must rely on indirect evidence of
this phenomenon. Both the insurance cost and the expedited
service arguments are based on the premise that value-of-
service price structures arise due to underlying cost differen-
tials between transport services for low and high valued car-
goes. If value-of-service price structures can be shown to
exist in unregulated, competitively structured markets, then
cost differences between transport of low and high valued car-
goes would be indicated. This follows because in a competitive
market sellers cannot discriminate between buyers on the basis
of differences in their demand elasticities. Therefore, simi-
lar charges would be expected across buyers for similar ser-
vices. Price differentials from buyer to buyer would be indi-
cative, then, of differences in the services delivered, rather
than of an exercising of market power on the part of the sell-
er.
This approach has been employed in two studies of the
deregulated general freight motor carrier industries: Blair,
Kaserman, and McClave in Florida, and Beilock and Freeman in
Arizona. In both studies pre and post deregulation rates were
examined across several routes, carriers, and weight and com-
modity classes. In both studies rate differentials between
commodity classes, which are in part based on value, were found
to persist after deregulation. These findings were cited as
supportive of the expedited service hypothesis.
While Blair, Kaserman, and McClave's and Beilock and Free-
man's findings are consistent with the expedited service hypo-
thesis, they fall well short of being conclusive evidence for
two principal reasons. First, while there is no obvious reason
for hypothesizing that insurance costs would be abnormally high
in these markets, it cannot be stated with certainty what pro-
portion of the observed rate differentials are due to insurance
cost differentials or to expedited service cost differentials.
This failing is not particularly serious for two reasons.
First, as just stated, it is highly unlikely that insurance
costs in both the Arizona and Florida intrastate markets would
greatly exceed those found by Fauth for the nation as a whole.
Moreover, even if it is insurance cost differentials and not
expedited services which are responsible for these results, the
basic point still holds that price discrimination is not the
culprit.
A far more serious criticism with using the intrastate
Florida and Arizona markets to make inferences about the source
of value-of-service price structures is that it is not entirely
clear that these markets are not operating as though rate bu-
reaus still exist. There are two reasons for suspecting this.
First, the habits of decades may not die over just a few years.
The fact that most carriers still employ the standard commodity
classifications in itself suggests a reticence to deviate from
long-established norms. One reason for such reticence in the
initial months after deregulation in Florida was that some
carriers threatened legal action to reverse the sunsetting of
the regulatory system. Also, the usage of class rates may be
convenient for those shippers and carriers having both unregu-
lated intrastate and regulated interstate freight. Indeed,
many carriers may be employing rates published by rate bureaus
in neighboring states or for interstate runs as guideposts for
their intrastate rate policies. The second reason is that in
both states the general freight carriage industry is highly
concentrated. While no exact data exist, it is evident that 80
or perhaps 90 percent of intrastate freight in each state is
hauled by the top 10 or 12 carriers. Taken on a route by route
or area by area basis, it is not uncommon to have virtually all
traffic handled by one, two or three carriers. This industry
structure raises the possibility of tacit or overt collusion.
While no evidence exists of collusion in these markets, the
possibility must be recognized. Moreover with so few firms,
even without collusion, sufficient monopoly power may be pre-
sent to facilitate price discrimination.
To obviate the preceding problems, what is needed is a
competitively structured, never regulated market. Only here
would a value-of-service rate structure due to price discrimi-
nation be impossible. Therefore, evidence of the existence of
such a structure would indicate cost differentials in the
transport services given for low and high valued cargoes. In
the remainder of this bulletin a study will be described of
such a transport market: the Florida-origin produce and orna-
mentals interstate truck market.
EMPIRICAL MODEL
The competitive structure of the industry suggests that
rates (PT) are likely to correspond closely to costs. There-
fore, for the empirical analysis, PT is specified as being a
function of the variables related to costs. This modeling
approach is common in transportation literature (e.g. Beilock
and Shonkwiler, Binkley and Harrar, Ferguson and Glorfeld, and
Perkins). The price necessary to bring forth the services
depends upon the direct or variable input costs (INP) and the
opportunity costs of alternative uses (PA). INP will depend
upon such factors as the quantity to be hauled (Q), the dis-
tance (D), and the quality of service (QUAL):
(1) PT = PT (INP(Q, D, QUAL,), PA)
Assuming that fuel efficiencies, speeds, and labor costs
are similar across trucks, distance (D) is a reasonable proxy
for running or on-the-road costs. As virtually all of the
vehicles had fully loaded, 42 to 45-foot refrigerated trailers,
the quantities carried (Q) can be assumed to be constant.
Moreover, all of the produce and ornamentals require similar
care in handling. Therefore, the major observable variations
in service quality (QUAL) per trip were the number of pickups
(PKUP) and drops (DROPS). Unfortunately, the promptness of
service as represented by the queue length or the amount of
excess capacity offered was not observable. However, following
DeVany and Saving (1977), in a competitive market the time-
sensitivity of the commodities may be substituted for queue
length. In this analysis, time-sensitivity is proxied by the
average per day loss in cargo value calculated as the per hun-
dredweight farm level price divided by days of storage life
(DAYLOSS).3,4 DAYLOSS would also serve as a proxy for the
level of risk inherent with each cargo. This follows because
DAYLOSS is, essentially, the level of liability (value) times
an index of the probability of a claim. The inverse of shelf
life indicates sensitivity to delays in transit and rough han-
dling, as produce becomes more susceptible to bruising and rot
as it ripens. Therefore, while the expedited service hypothe-
sis is viewed by the author as likely to be the most important
unobserved service, it must be acknowledged that the effect of
insurance costs is not separable in the analysis.
In order to capture possible rate differences resulting
from carrier type and experience, and/or the manner in which
the load was arranged, the following are included as independ-
ent variables:
OWNOP = 1 if carrier is an owner-operator, 0 otherwise;
SHUTTLE = I if carrier is based in Florida or in one of the
destination states for the haul, 0 otherwise;
BROKER = 1 if load arranged by a broker, 0 otherwise; and
OEXPER = number of years of driving experience if an owner-
operator, 0 otherwise.
SHUTTLE is included on the premise that a carrier based at
either end of the haul would be more likely to possess more
complete market information. BROKER is included to determine
if rates received are dependent upon how the load is arranged.
The driver experience variable, OEXPER, is constructed as an
interaction term with OWNOP, because in general, fleet drivers,
unlike owner-operators, generally have little to do with pro-
curing loads. The general form of the relationship estimated
was:
(2) PT = PT(D, PKUP, DROPS, DAYLOSS, OWNOP, SHUTTLE,
BROKER OEXPER).
To test for nonlinear changes in rates over distance, both D
and its square, DSQ, were employed in the model. Finally, an
intercept shifter was employed to capture overall differences
in rate levels between March and June due to differences in
market activity. This, in effect, was intended to proxy dif-
ferences between the survey months in alternative activities
available to carriers (PA).
DATA
On March 28 and 29, 1983, and again on June 1 and 2, 1983
interviews were conducted with northbound truckers carrying
produce and ornamentals passing through the Florida Agricultur-
al Inspection Station on U.S. Interstate 95. In March and
June, 198 and 215 interviews were completed, respectively. The
interviewing was performed from 5:00 PM to midnight on each of
the days (when most of the outbound truck passing occur).
Very specific questions were asked regarding the rates re-
ceived, cargoes, and service characteristics of the currently
carried load, as well as certain characteristics of the carrier
(see Appendix 2). To avoid selection bias, during the inter-
view period all truckers carrying produce and ornamentals were
interviewed. The interview strategy employed had an added
advantage in that the quality of the information obtained
should be high for two reasons.5 First, the shipping documents
could be examined (they are temporarily surrendered by the
truckers during the inspection process) and, second, truckers
were questioned about current or just past events. With only 5
truckers refusing to be interviewed, the response rate was,
essentially, 100 percent. A summary of responses is presented
in Table 3.
THE RESULTS
In Table 4, the results of the estimation process are
presented. The Goldfeld-Quandt test for heteroskedasticity
across distance (D and DSQ) and DAYLOSS was carried out, but
none was indicated.6 The equation explains 58 percent of the
variation in PT as indicated by the coefficient of determina-
tion. Moreover, all parameter estimates have the theoretically
anticipated signs or have standard errors which are large rela-
tive to the absolute value of the estimated paramter.
The estimated parameter associated with the June intercept
shifter is positive and over four times the magnitude of its
standard error, indicating a tighter truck supply situation in
June than in March. This is not surprising when it is consid-
Table 3. Carrier responses to March and June 1983 survey.
Survey month
Item March June
Number of carriers
(percentages in parentheses)
Carrier status:
Owner-operator 111 (56) 105 (56)
For-hire fleet 52 (26) 57 (27)
Private fleet 32 (16) 52 (24)
Other 3 (2) 1 (0)
Percent of time hauling
exempt commodities:
Owner-operators 70 67
For-hire fleet 65 60
Private fleet 82 95
All carriers 70 73
Percent of time trip leasing:
Owner-operators 14 18
For-hire fleet 13 16
Private fleet 8 5
All carriers 13 14
Percent of time permanent leasing:
Owner-operators 11 11
For-hire fleet 2 5
Private fleet 10 0
All carriers 11 7
Percent of time using
own authority:
Owner-operators 5 4
For-hire fleet 20 19
Private fleet 10 0
All carriers 11 6
Firm location:
Florida 64 (32) 55 (26)
GA, NC, SC 57 (29) 72 (33)
VA, WV, MD, DE 20 (10) 21 (10)
PA, NJ, NY 27 (14) 32 (15)
Table 3. Carrier responses to March and June 1983 survey
(continued).
Survey month
Item March June
Number of carriers
(percentages in parentheses)
Firm location:
New England 10 (5) 10 (5)
Lake States1 8 (4) 4 (2)
Canada 7 (4) 8 (4)
Other2 5 (2) 10 (5)
Percent from Florida:
Owner-operator 41 32
Other carriers 20 19
All carriers 32 26
Destination:
GA, NC, SC 39 (20) 61 (28)
VA, WV, MD, DE 31 (16) 32 (15)
PA, NJ, NY 75 (38) 82 (38)
New England 26 (13) 14 (7)
Lake States1 3 (1) 3 (1)
Canada 18 (9) 14 (7)
Other2 6 (3) 9 (4)
Percent with an inbound load:
Owner-operator 73 54
Fleet carrier 96 70
Private carrier 59 29
All carriers 77 50
Inbound load arranged by:3
Use of authority 41 (27) 25 (23)
Trip lease 45 (29) 35 (32)
Broker 32 (21) 14 (13)
Shipper 11 (7) 13 (12)
Receiver 3 (2) 5 (5)
Other 21 (14) 16 (15)
Table 3. Carrier responses to March and June 1983 survey
(continued).
Survey month
March
June
Number of carriers
(percentages in parentheses)
Percent with ICC authority:
Owner-operator
Fleet carrier
Private carrier
Made any intrastate hauls
while in Florida?
Yes
No
Principal commodity:
Citrus
Tomatoes
Other
Load configuration:4
Mixed
Straight
Pickup:
Single
Multiple
Drops:
Single
Multiple
Arrangement
Broker
Shipper
Receiver
Other
of outbound load:
Item
4 (2)
194 (98
3 (1)
212 (99)
(36)
(9)
(56)
(3)
(25)
(72)
47 (24)
151 (76)
116 (59)
82 (41)
119 (60)
79 (40)
42 (20)
173 (80)
172 (80)
43 (20)
119 (51)
105 (49)
(56)
(22)
(16)
(6)
- ------ ------
Table 3. Carrier responses to March and June 1983 survey
(continued).
Survey month
Item
March
June
Number of carriers
(percentages in parentheses)
Load arranged prior to
returning to Florida:
Yes
No
Time necessary to arrange load:
less than 1 day
1-3 days
Over 3 days
Average time to arrange load
(days)
100 (51)
98 (49)
161 (81)
20 (10)
17 (9)
.99
131 (61)
84 (39)
(88)
(12)
(0)
O1hio, Indiana, Michigan, Wisconsin, Minnesota, Illinois,
Missouri, and Kentucky.
All states west of the Appalachians and excluding the Lake
States.
percent of those with inbound load in parenthesis.
4Citrus is considered to be one commodity. Therefore a "mixed"
grapefruit and oranges load is considered here as a "straight"
load.
Table 4. Rate estimation equations dependent variable: truck-
load freight rate.
Independent
Variable
Intercept
June intercept shifter
Average daily loss per cwt.
Distance (miles)
Distance squared
Number of pickups
Number of drops
Owner-operator
Year of experience, if owner-operators
Carrier based at origin or destination
Load arranged by broker
F
R2
Number of observations
aOne hundred and fifty-one of the origin
incomplete data.
Estimated Parameters
(standard errors)
-480.6
(191.8)
184.5
(41.60)
50.83
(8.175)
1.856
(.3017)
-.0003456
(.0001198)
38.92
(15.08)
52.35
(14.83)
-31.33
(56.38)
-1.923
(2.020)
40.70
(41.01)
-29.14
(43.56)
34.83
.5812
262a
ial 413 observations had
ered that produce shipments from Florida peak in the late May-
early June period. At the time of the March survey, Florida
produce and ornamentals movements were approximately 3,000
truckload-equivalents per week, a rate which had been main-
tained since early January and would continue virtually un-
changed for several more weeks (USDA, 1980-1983). The USDA
Agricultural Marketing Service (AMS) characterized truck avail-
ability as "adequate" to "slight surplus" (USDA, 1983). At the
time of the June survey, the shipment rate had risen to around
7,000 truckload-equivalents per week, and availability was
classified by the AMS as "slight shortage" or "shortage."
Therefore, in June the reservation price of a carrier would
have been greater than in March, even if the direct variable or
accounting costs associated with acquiring a load had not
changed.
The positive parameter estimate associated with DAYLOSS
(50.83) has an absolute magnitude about six times that of its
standard error. This result lends strong support to DeVany and
Saving's contention that value of service pricing schemes can
evolve in competitively structured, unregulated markets. Again
however, it cannot be determined to what extent this is due to
costs associated with expedited service or to insurance. Em-
ploying the parameter associated with DAYLOSS, it is estimated
that for every 1,000 dollar increase in the average daily loss
per truckload PT rises by 118 dollars.7
Not surprisingly, PT is strongly and positively associated
with distance. The estimated parameter associated with D
(1.856) is over six times the magnitude of its standard error.
The estimated parameter associated with DSQ (-.00003456) is
also large in absolute terms relative to its standard error.
The peak in the rate function across distance is at 2,685
miles, well beyond the 1,705 mile maximum length of haul in the
data. These results indicate a tapering rate/distance schedule
similar to that shown in Figure 1.
The measures of carrier sophistication and status, market
familiarity, and method of load arrangement (OEXPER, OWNOP,
SHUTTLE, and BROKER) all prove to be of low explanatory value.8
The weak results for these variables had been expected in view
of the competitive nature of the market. Indeed, their lack of
significance may be viewed as evidence of market efficiency in
that there is no indication that similar services are priced
differently based upon supplier (as opposed to service) char-
acteristics.
Remuneration for each pickup and each drop is estimated to
be $38.92 and $52.35, respectively. Both parameter estimates
are large relative to their standard errors. The somewhat
higher estimated return for drops than for pickups was expected
due to the frequent practice at the northern terminals of
charging gate or entry fees, and of lumping (i.e., forced pay-
ment for unloading a vehicle).
SUMMARY AND CONCLUSIONS
The three objectives of the study were to investigate the
effect on rates of:
1. distance,
2. number of pickups and drops, as well as carrier type and
method of load arrangement, and
3. shipping urgency.
The results with respect to distance indicate a tapering rate/
distance gradient. Three possible rationales for such gradi-
ents are differences in equipment depending on distance, price
discrimination, and volume discounts due to the reduced fre-
quency of search and repositioning with longer hauls. As full
size tractor/trailers were used by all of the carriers in the
survey, the first rationale is not possible. Likewise the
second rationale is not feasible because of the competitive
nature of the industry.
The results indicate that rates rise by almost $40.00 per
pickup and over $50.00 per drop. Carrier status and experience
as well as the means of arranging carriage, however, do not
appear to have a significant effect on rates. All of these
results indicate that the market is operating efficiently in
that additional services (pickups and drops) receive compensa-
tion, and similar services are not priced differently depending
upon the carrier or method of arranging the load.
Another major focus of the paper was value-of-service or
cargo value-based rate structures. This form of pricing has
traditionally been thought to be limited to regulated carriage,
as it was assumed to be a form of price discrimination. In
1977 DeVany and Saving asserted that the presence of a value-
of-service pricing structure also could be explained in terms
of varying levels of compensation for varying levels of ser-
vice. In particular, it could reflect payment for the costs
inherent in providing expedited, ready-on-demand carriage.
They argued that shipper/receivers of time-sensitive freight
(high value or perishability) would be the most likely to se-
lect such service. If true, value-of-service price structures
would be expected to occur in unregulated, competitively struc-
tured markets. The results of this study lend support to that
view. Despite a highly competitive industry structure, freight
rates are found to be responsive to the urgency of the ship-
ment, as measured by the average daily loss in value due to
perishability. This result suggests that shipper/receivers of
high valued, regulated commodities may continue to pay premiums
for freight even in the face of deregulation or reduced regula-
tory controls on rates.
FOOTNOTES
1. Rate bureaus are private firms or associations through
which rates may be negotiated collectively among carriers
or carriers and shipper/receivers. These negotiations are
granted total or partial immunity from antitrust laws.
Rate bureaus also provide the service of publishing these
rates (publication is normally required for regulated
rates). Rate bureaus may represent one or more carriers in
rate hearings.
2. An exception to this is released rates, in which all or
part of the liability is removed from the carrier. Such
rates are not the norm, and when offered are at a discount.
In effect, then, released rates are simply reduced rates
for reduced service.
3. Farm level prices for the 24 different commodities recorded
were obtained from discussions with personnel at the Flor-
ida Crop and Livestock Reporting Service. Perishability
estimates for Florida commodities were obtained from Pav-
lovic et al (1980).
4. A possible alternative specification would have been to
include both the F.O.B. price (PRICE) and the inverse of
the storage life (IPERISH) as well as or in lieu of DAY-
LOSS. This was not done as DAYLOSS alone was felt to be
the correct representation of shipping urgency. Later
experimentation suggested that this specification was cor-
rect. PRICE and IPERISH are strongly correlated (.74).
The use of both variables separately would likely result in
multicollinearity problems, making discernment of the sepa-
rate effects difficult.
5. The circumstances under which the interviews were conducted
were conducive to promoting open, non-evasive responses.
The interviewers made it clear that they were with the
Florida Department of Agriculture and Consumer Services or
the University of Florida, and not with a law enforcement
agency. The carriers appeared to be relaxed and responded
readily and without hesitation. The negligible refusal
rate supports these observations.
6. The Goldfield-Quandt test involves ordering the observa-
tions in ascending order by the variable, X, being tested
for heteroskedasticity. The first and last N observations
are employed in separate regressions of the model. The
ratio of the error sum of squares of the regression for the
second N observations to that for the first N observations
will, on the assumption of homoskedasticity, have and F
distribution with N-K degrees of freedom in the numerator
and the denominator (see Johnston, pp. 214-21 for further
details).
Following this procedure by varying the first and last
100 observations the following F89,89 statistics are gener-
ated:
.80 for DIST
.86 for DAYLOSS.
Therefore, the null hypothesis of homoskedasticity is not
rejected.
7. Assuming 43,000 pound net cargo weight per truckload, for a
1000 dollar increase in average daily loss per truckload,
the average daily loss per hundredweight (DAYLOSS) in-
creases by 2.336 dollars. The resulting effect on the per
truckload freight rate (PT) is:
$2.336 50.83 = 118.43 dollars.
8. The joint test for significance of OWNOP, OEXPER, SHUTTLE,
and OUTBRK genetrates an F4,251 statistic of 1.07. This is
well below critical values for this statistic at conven-
tional levels. For an explanation of how to calculate this
statistic, see Johnston, pp. 135-55.
APPENDIX 1
The Relationship of the Demand for Transportation and the
Elasticity of Supply and Demand for the Commodity
Begin by assuming linear demand and supply functions:
(1) QD = a b PQ
(2) QS = c + d PFOB
(3) QS QD Q
where: QD' QS = quantities demanded and supplies,
respectively
PQ PFOB = demand and supply prices (net of transpor-
tation costs, PT)
a,b,c,d = positive parameters
Taking the inverse of (1) and (2) and using (3):
1
(4) P = (a-Q), and
(5) PFO 1 (-c+Q)
(5) FOB d
Define PT, the transportation prices as:
(6) PT =Q FOB
(6a) PT = (a-Q) Z (-c+Q)
(6b) PT ( Q( ).
Solving for Q yields:
(-= ( + P
b d
The elasticity of demand for transportation (ET) is:
-1 PT
(8) ET 1 1 Q
(+-- + )
b d
From (1) and (2) and using (3) we may derive the following:
EDQ
(9) b = and
PQ
ESQ
(10) d =--
FOB
where
ED = elasticity of demand for the commodity.
ES = elasticity of supply for the commodity.
Substituting (9) and (10) into (8) yields
EDES PT
(11) E = (P E-S FBED)
Recalling that PFOB PQ PT (11) may be rewritten as:
EDE PT
(12) E = ( PT )
T P P D
ES (1 --) ED
Q
or
ES PT
(12a) ET= (-- ) ED
E ( --) E
Q
T ES
(12b) ET = R ED -, where R =
T DbPED PQ' T
ES -(1- --p) ED
Notice that as long as ED < 0, R will be positive and less than
1. Moreover, R approaches unity as ES + -. Therefore, the
parenthetical expression scales down ET (in absolute terms)
as ES deviates from infinity.
APPENDIX 2
Motor Carrier Questionnaire
1. How long have you been a driver?
2. Which best describes your operation? (a) owner-operator
(b) fleet operation (c) private operation (d) agricultural
cooperative (e) other
3. Do you have ICC authority? YES NO
If yes, to haul what commodities?
4. Out of which state to you operate?
(Base plate state)
5. About what proportion of time do you haul Exempt Commodi-
ties? % Trip Lease %_
Permanent Lease %
6. What is/are your load(s) today, how much of each, where
did you make your pick-ups and what are your destinations?
NUMBER OF NUMBER OF
MAJOR COMMODITY AMOUNT PICKUPS DROPS DESTINATION
(List 2) or % full state
7. How did you get this load? (a) broker (b) direct contact
with shipper (c) direct contact with receiver (d) other
8. How long did it take to get it?
(From the time you started looking.)
9. Did you have this load arranged by the time you entered
Florida? YES NO
10. What are you getting for the load you are now carrying?
$
11. What did you bring into Florida?
12. From what state?
How did you get that load?
REGULATED:
used own authority
trip lease
other
EXEMPT:
broker
in Florida
shipper
receiver
other
13. What did you get on the load you brought into Florida?
(See question #11.) $
14. Did you make any intrastate hauls in Florida prior to
getting this load? YES NO
If yes, what?
15. Is there any area of exempt trucking that concerns you
today? (a) safety (b) lumping (c) nonuniformity of truck-
ing regulations from state to state (d) rates (e) other
REFERENCES
Beilock, R., and G. Fletcher. The Structure and Characteris-
tics of the Florida Exempt Perishables Trucking Industry,
Food and Resource Economics Report No. 109, University of
Florida, 1983.
Beilock, R., and J. Freeman. Motor Carrier Deregulation and
Tax Issues in Arizona, report prepared for the Arizona
Department of Transportation, 1984, 200 pp.
Beilock, R., J. Kohburger, and J. Morgan. "Short Term Truck
Rate Variations for California and Florida Produce Shipped
to the Northeast," Journal of the Northeast Agricultural
Economics Council 13,1 (1984): 25-32.
Beilock, R. and J. S. Shonkwiler. "Modelling Weekly Truck
Rates for Perishables" Southern Journal of Agricultural
Economics 15 (July 1983): 83-87.
Benishay, H., and G. Whitaker. "Demand and Supply in Freight
Transportation," Journal of Industrial Economics, 14,3
(1966):243-62.
Binkley, J., and B. Harrar. "Major Determinants of Ocean
Freight Rates for Grains: An Economic Analysis," American
Journal of Agricultural Economics 63(1982): 47-57.
Blair, R., D. Kaserman, and J. McClave. The Economic Effects
of Deregulation of Intrastate Trucking in Florida, report
prepared for Florida Attorney General's Office, 1983, 37pp.
Boles, P. Owner-Operator Costs of Hauling Fresh Fruits and
Vegetables in Refrigerated Trucks, U.S.D.A. Report No. 82,
1980.
Bressler, R., and R. King. Markets, Prices, and Interregional
Trade, John Wiley & Sons, Inc., 1978.
Cherry, R., and N. Garston. "Operating Ratio Regulation:
Control Theory Approach," Transportation Science 16,1
(1982): 67-82.
DeVany, A., and T. Saving. "Product Quality, Uncertainty, and
Regulation: The Trucking Industry," American Economic Re-
view 67(1977): 583-594.
Fauth, G. "The Role of Commodity Value in Motor Carrier Class
Rate Structures," Invited Paper, U.S.D.O.T. Conference on
Regulatory Reform in Surface Transportation, 1983, pp. 46-
71.
Ferguson, W., and L. Glorfeld. "Modeling the Present Motor
Carrier Rate Structure as a Benchmark for Pricing in the
New Competitive Environment," Transportation Journal
21(1981): 59-66.
Fuller, S., and C. Shanmaugham. Effect of Rail Deregulation:
The Case of Wheat Exports from the South Plains, Texas
Agricultural Experiment Station Bulletin, No. B-1385, 1982.
Garrod, P., and W. Miklius. "Estimated Value of Improvements
in Transport Services," Transportation Research, 17A,1
(1983):33-8.
Johnson, M. "Estimating the Influence of Service Quality on
Transportation Demand," American Journal of Agricultural
Economics 58,3(1976):496-503.
Johnston, J. Econmetric Methods, McGraw-Hill Book Company,
1972.
Klindworth, K., and E. Brooks. Shipping Alternatives for Mov-
ing Florida Fruits and Vegetables to Eastern and Midwestern
Markets, U.S.D.A., Office of Transportation, 1981.
Lockin, D. Economics of Transportation, Illinois, Richard D.
Irwin, Inc., 1972.
Martin, M. "Misallocative Effects of Value-of-Service Rail
Grain Rates," Transportation Journal 19,3(1979):74-83.
Maze, T. "The Value of Information in Unregulated Truck Ser-
vice Markets," Transportation Journal 20,2(1980):57-62.
Michaels, G., R. Levins, and J. Fruin. "Rail/Truck Competition
for Grain Traffic in Minnesota: Implications for Rate
Making," American Journal of Agricultural Economics 64,2
(1982):276-9.
Miklius, W. "Estimating the Demand for Truck and Rail Trans-
portation: A Case Study of California Lettuce," Agricul-
tural Economics Research 29,2(1967):46-50.
Miklius, W. ard: K. Casavant. "Estimated and Perceived Varia-
bility of Transit Time," Transportation Journal 15,4(1975):
47-51.
Miklius, W., K. Casavant, and P. Garrod. "Estimation of Demand
for Transportation of Agricultural Commodities," American
Journal of Agricultural Economics 58,2(1976):217-23.
Olson, J. "Price Discrimination by Regulated Motor Carriers,"
American Economic Review 62(1972): 395-402.
Oum, J. "A Cross Sectional Study of Freight Transport Demand
and Rail-Truck Competition in Canada," Bell Journal of
Economics 10,2(1979):463-82.
Pavlovic, K., G. Long, D. Reaves, and T. Maze. Domestic Trans-
portation for Florida Perishable Produce, Transportation
Research Center, Univ. of Florida, 1980.
Pederson, L., R. Mittelhammer, and K. Casavant. "Factors
Affecting Interstate Backhauling of Exempt Agricultural
Commodities by Regulated Motor Carriers: A First Look,"
Transportation Journal 19,2(1979):46-52.
Pegrum, D. Transportation Economics and Public Policy, Illi-
nois, Richard D. Irwin, Inc., 1973.
Perkins, M. A Truck Freight Model, Massachusetts Institute of
Technology, Center for Transportation Studies Report 80-15,
1980.
Phlips, L. The Economics of Price Discrimination, New York,
Cambridge University Press, 1983.
Sorenson, 0. "Rail-barge Competition in Transporting Winter
Wheat," American Journal of Agricultural Economics 55,5
(1973):814-19.
Spady, R., and A. Friedlaender. "Hedonic Cost Functions for
the Regulated Trucking Industry," Bell Journal of Economics
10, 1(1979):159-79.
Sugrue, P., M. Leford, and N. Glaskowsky. "Operating Economies
of Scale in the U.S. Long-Haul Common Carrier Motor Freight
Industry," Transportation Journal 22,1(1982):27-41.
Taff, C. "A Study of Truck Brokers of Agricultural Commodities
Exempt from Economic Regulation," Transportation Journal
19,1(1979):5-15.
USDA. Fresh Fruit and Vegetables Shipments, Agricultural Mar-
keting Service, (annual and weekly reports) 1980-1983.
USDA. "Fruit and Vegetables Truck Rate Report," Agricultural
Marketing Service, 1983 (weekly).
Wilson, G. Economic Activity of Intercity Freight Transporta-
tion, Indiana University Press, 1982.
This publication was produced at an annual cost of $1,309 or
a cost of $1.31 per copy to provide information on the deter-
minants of motor carrier freight rates for Florida perishables.
All programs and related activities sponsored or assisted by the Florida
Agricultural Experiment Stations are open to all persons regardless of race,
color, national origin, age, sex, or handicap.
ISSN 0096-607X
r~n~r~n~
|