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.