The current issue and full text archive of this journal is available on Emerald Insight at:

[Link]/[Link]

IJLM
28,1 Benchmarking freight rates and
procuring cost-attractive
transportation services
194 Seong-Jong Joo
College of Business, Central Washington University-Des Moines,
Received 30 January 2015
Revised 26 June 2015 Des Moines, Washington, USA
Accepted 3 August 2015
Hokey Min
Department of Management, Bowling Green State University,
Bowling Green, Ohio, USA, and
Carlo Smith
College of Business, Central Washington University-Des Moines,
Des Moines, Washington, USA

Abstract
Purpose – The purpose of this paper is to help shippers determine a negotiation yardstick for transportation
price and formulate wise transportation outsourcing strategies by examining the presence of freight rate
differentials for shippers and identifying their main causes. This paper also develops a framework for
benchmarking freight rates based on the actual data.
Design/methodology/approach – This paper proposes an additive dummy regression model to determine a
statistical significance in shipping charges between different shippers. Unlike the traditional least square
analysis, the proposed model is designed to avoid a biased assessment of the impact of an explanatory variable.
Findings – Through a series of empirical data analysis and hypothesis tests, the authors discovered that the
fixed portion (minimum base charge) of shipping charges differed depending on the shipper’ individual
contract, while the variable portion (fuel or accessorial charge) of shipping charges remained the same
regardless of the shipper’s individual contract. As such, shippers who are unaware of flexible but
unpredictable transportation pricing practices and unprepared for freight rate negotiation can suffer from
higher shipping costs as compared to their peers. Thus, the authors conclude that the success of
transportation outsourcing, carrier selection, and freight rate negotiation strategies tends to rest on the
shipper’s ability to understand transportation cost structures and then determine the benchmark freight rate
considered “fair” and “reasonable” for a given service.
Originality/value – This paper is one of the first to examine shipping cost differentials between different
shippers and determine what causes such differentials. In doing so, this paper attempted to assess the
potential impact of freight rate negotiation and carrier selection strategies on shippers’ transportation costs in
current deregulatory environments where shippers were given a greater freedom to negotiate freight rates
with carriers and an increased opportunity to save their transportation costs.
Keywords Transportation, Pricing
Paper type Research paper

1. Introduction
The passages of a series of landmark transportation deregulatory acts such as the Motor Carrier
Act of 1980, the Surface Freight Forwarder Deregulation Act of 1986, the Negotiated Rates Act
of 1993, the Trucking Industry Regulation Reform Act of 1994, and the Interstate Commerce
Commission (ICC) Termination Act of 1995 brought sweeping changes in transportation pricing,
carrier market entry, and operating territories in the transportation industry. In particular, these
acts allowed the shipper to enjoy the unprecedented range of bargaining room for freight rates
The International Journal of
Logistics Management and exploit transportation cost saving opportunities. For example, between 1977, a year before
Vol. 28 No. 1, 2017
pp. 194-205
the ICC started to relax its control over the industry, and 1982 two years after the Motor Carrier
© Emerald Publishing Limited
0957-4093
Act of 1980 was introduced, rates for truckload-size shipments fell by about 25 percent in
DOI 10.1108/IJLM-01-2015-0030 inflation-adjusted terms. The General Accounting Office found that freight rates charged by
less-than-truckload (LTL) carriers had fallen by as much as 10-20 percent, with some shippers Cost-attractive
reporting declines of as much as 40 percent (Moore, 2008). Thanks to declining freight rates, the transportation
US transportation costs were reduced to 3.9 percent of GDP in 2013 from 7.4 percent in 1980 services
(Case Logistics, 1980; Penske Logistics, 2014). Since trucking-related costs comprised 77.2 and
77.4 percent of total transportation costs and 47.5 and 48.4 percent of total logistics spend in
2013 and 2014, respectively, trucking-related cost savings could be translated into total logistics
cost savings (Gilmore, 2014; Penske Logistics, 2015). Indeed, transportation cost (e.g. trucking- 195
related cost) patterns often paralleled with logistics cost patterns and transportation
cost accounted for the largest portion of overall logistics cost (Penske Logistics, 2014, 2015).
A similar pattern can be found in other countries such as South Africa (Havenga and Simpson,
2014). Although transportation deregulation dramatically increased transportation cost saving
potentials, it created ambiguity and confusion regarding the fairness and reasonableness of
freight rates for a given transportation service. Such ambiguity and confusion have often led to
transportation price (shipping charge) differentials for the same service due to wide ranging
negotiation outcomes. A similar problem has existed in Europe in that the freight rate for a
certain route was once determined based on the lump-sum costs. Such a transportation pricing
scheme created a difficulty for the shipper in ascertaining the primary cause of a freight
rate increase and in understanding cost justification for the fixed freight rate. To ease
such a difficulty, very few freight rates have been regulated in the European Union countries
for the last decades, although Eastern and Central European countries have the remnants of
transportation tariff regulations (The European Conference of Ministers of Transport, 1998).
With the less restriction on transportation pricing across the world, the shipper bears the
burden of figuring out what price range will be targeted as the most attractive benchmark
prior to rate negotiation.
Thus, it is imperative for shippers to understand how their rates would be compared to those
of other shippers and to see whether their rates charged by carriers are fair and reasonable for a
given service. Skinner et al. (2009) once examined the fairness of a truckload rate at the lane level,
but no prior study in the published literature identified the benchmark freight rate at a company
level and checked to see whether rate differentials were affected by transportation outsourcing
and/or carrier choice strategies. Given this research gap, the primary objective of this study is to
develop ways to estimate benchmark freight rates and then maximize transportation cost saving
opportunities for shippers. To achieve the study objective, we first look at the transportation cost
structures (i.e. variable, semi-variable, and fixed costs) as summarized in Table I and then
evaluate three major strategies for carrier selection and outsourcing transportation services;
bidding or purchasing in the spot market; maintaining private fleets; and long-term contracting.
If a shipper selects a bidding option, it eliminates a need for investment in fixed assets and pays
only the base rates plus extra fees, such as fuel surcharges and accessorial charges (e.g. direct
and inside delivery). In this case, however, the shipper loses control over carriers and the
subsequent flexibility. If a shipper keeps its own private fleet, it can reduce variable costs but will
incur the investment costs for buying/leasing transportation equipment. Long-term contracts
with the contract carriers are compromises between first two options.
Shippers may choose one of the three aforementioned options or combine more than one
of those options. Since such a choice will be influenced by estimated freight charges,

Cost category Cost components

Variable cost Direct labor, fringe benefits, fuel, insurance premium, equipment
part/component, scheduled maintenance, and taxes Table I.
Semi-variable cost Licenses, sales, and advertisement Transportation
Fixed cost Depreciation, overhead, administrative cost, utilities, and property/building cost structures
IJLM shippers should first determine the benchmark rate by calculating shipping costs for the
28,1 same or similar shipping profiles (e.g. similar shipment sizes/weights, load types, shipping
lanes, shipping distance, number of stops, and needed trailer types). In addition, we compare
shipping costs between shippers with different outsourcing strategy so as to see if that
strategy matters for finding most attractive freight rates.
Although transportation cost structures can be classified into three categories, the typical
196 basis of transportation pricing in practice do not necessarily include all three categories of
costs. For example, semi-variable cost components such as sales and advertisement cost can be
negligible since many carriers (especially small, regional carriers) cannot afford to spend a lot
for sales and advertisement. Also, a license fee will be the same across the carrier and will not
vary significantly from one year to another. For that reason, a license fee does not make a
difference in competitive transportation pricing. In practice, transportation pricing is often
determined based mainly on minimum base charges, fuel surcharges, and accessorial charges.
In other words, transportation pricing does not necessarily reflect all three categories of costs.
As such, it is extremely difficult for any shippers to project what the right freight rate is and
how it will be determined. This uncertainty and ambiguity poses additional challenges for
freight rate negotiation. One way to deal with these challenges is to figure out the role of
aforementioned charges in transportation pricing. With that in mind, we designed a new form
of a regression model that used those charges as independent variables and estimated the
intercepts and coefficients of those independent variables to assess their impact on
transportation pricing (i.e. freight rate determination).

2. Relevant literature
The Internal Revenue Service and accounting auditors allow shipping costs to be part of
business expenses: thus, shipping costs affect the company’s profitability. Especially,
shipping costs are known to be one of the biggest expenses for many small companies
(Porter, 2012). Despite declining oil prices, shipping costs have been on the rise for the last
few years. For example, both United Parcel Services and FedEx raised their shipping rates
by 4.9 percent in late 2014, following the earlier rate hike by the United State Postal Services
([Link], 2014; Huddleston, 2014). Considering the significant impact of shipping costs
on the company’s bottom-line, the way shipping costs were determined deserves more
attention from shippers and other transportation service users. Nevertheless, prior studies
investigating shipping cost dynamics or freight rate savings via negotiation are still almost
non-existent with an exception of a handful of pioneering studies that attempted to identify
factors affecting shipping costs and investigate why shipping rates varied from one service
to another and one shipper to another even for the same service.
For instance, Skinner et al. (2009) employed econometric models to analyze the shipping
costs on truckloads carried by dry or reefer vans based on aggregated data from multiple
companies and then assessed the impact of fuel surcharges on freight rates. They found that
distance or mileage explained 75-80 percent variation in shipping costs, with an adjusted
coefficient of determination, or R2 of 94 percent. Also, they observed that geographic
characteristics of the shipping market accounted for the 8-10 percent variation in shipping
costs. Kay and Warsing (2009) developed a non-linear regression model to predict LTL rates
based on publicly available data including tariffs and industry average discounts. They also
evaluated truckload (TL) and LTL shipping options based on their impact on total shipping
expenditures and logistics costs by making a tradeoff between transportation and inventory
carrying costs. Özkaya et al. (2010) used data from Schneider Logistics to predict and
benchmark LTL rates that were heavily discounted by 50-75 percent from the published
rates or tariffs. In their regression analysis, they considered weight, travel miles, freight
class, origin and destination region, and carrier type as potential predictors for the LTL
rates. They found that miles and weight were the most important variables for predicting
the actual LTL rates. In addition, they concluded that intangible factors such as freight Cost-attractive
desirability, negotiation power, and economic value (of freight) were statistically significant transportation
for explaining the actual LTL rates in their model; and actual freight class was more services
important than the contracted freight class for estimating LTL rates.
Rather than examining the rate structures of for-hire carriers, Galletti et al. (2010)
analyzed the private carrier’s rate structures and attempted to benchmark shipping costs
incurred by the use of the company’s own fleet. Focusing on the case of ocean carriers 197
(especially oil tankers), Randers and Göluke (2007) attempted to forecast shipping freight
rates based on the patterns of fleet capacity adjustments and fleet utilization over multiple
years and discovered that cargo, vessel, and contract types influenced shipping freight
rates. More recently, Özkaya et al. (2010) used a regression-based methodology to estimate
the LTL market rates with high reliability based on an extensive database of historical
shipments from the USA. Their model successfully combined the quantitative data with
qualitative market knowledge to produce better LTL market rate estimates which could
allow shippers to identify key transportation cost drivers and the subsequent cost saving
opportunities. Although these prior studies help build a theoretical foundation for
understanding the freight rate structures and predicting the freight rates, most of them
failed to establish the benchmark freight rate that shippers can target and bargain for before
coming to the negotiation table and failed to offer practical guidelines for developing wise
freight rate negotiation and contract management strategies. Likewise, in the traditional
benchmarking literature, a vast majority of the supply chain benchmarking efforts focused
primarily on the development of performance indicators (e.g. order fulfillment rate and order
cycle time) or the measurement of comparative operating efficiency of organizational units
instead of finding a yardstick for logistics service pricing such as transportation pricing
(Beamon, 1999; Gunasekaran et al., 2001; Wong and Wong, 2008; Shepherd and Günter,
2011; Sillanpää, 2015). This study is intended to fill this research gap.

3. Research methodology
To identify what may have caused shipping cost differentials between shippers and how
shippers’ different transportation outsourcing strategies may have affected their shipping
costs, we collected and analyzed two shipping companies’ data that were obtained from the
Supply Chain Consortium in Raleigh, North Carolina. With respect to strategies for carrier
selection and outsourcing transportation services, these two companies employed a strategy
of mixing spot market purchasing with long-term contracting and thus did not consider
private fleet usage options. The actual data provided by these two companies (shippers)
contained transportation transactions that occurred between them and their carriers during
the 2011 contract period. To maintain consistency and compatibility between two different
shippers’ data, we limited those transactions to direct truckload shipments hauled by dry
van in the same market. This sample data consisted of a total of 230 observations for two
shippers (actually retailers), dubbed L Shipper and W Shipper throughout this paper to keep
their confidentiality. These two shippers are national retailers of consumer goods in the
same industry that distributed and sold the same or comparable products. The transactions
we chose for analysis were taken from these shippers at the same point of time in April,
2011. Products that they shipped were primarily non-seasonal. Accordingly, their freight
rates were insensitive to seasonal variations. The sample included 101 cases of transactions
for L Shipper and 129 cases of transactions for W Shipper with other specific information
such as consignees’ names, transaction identification numbers, shipping origins,
destinations, base charges, fuel surcharges, other accessorial charges, total charges, miles
traveled, the number of stops, transportation modes, and trailer types. Based on this
information, shipping costs of two companies are broken down into two categories: fixed
and variable costs. Notice that we did not include semi-variable costs such as license fees
IJLM since license fees could be incorporated into the fixed cost and were negligible for
28,1 comparison purposes. Fixed cost components include one-time charges and/or charges
invariant by distance (e.g. minimum charges assessed by carriers regardless of distance,
fuel surcharges, other accessorial surcharges associated with residential delivery and
loading/unloading). Variable costs include distance-based shipping rates such as per-mile
freight rates. The rationale being that if the premise of equal shipping costs regardless of
198 different outsourcing strategies turned out to be true, both fixed and variable costs
(shipping charges) of two shippers are expected to be the same. Thus, we developed and
tested the following hypotheses using regression models:
H1. Fixed costs of L Shippers are the same as those of W Shipper, regardless of their
differences in transportation outsourcing (or carrier choice) strategies.
H2. Variable costs of L Shippers are the same as those of W Shipper, regardless of their
differences in transportation outsourcing (or carrier choice) strategies.
In these regression models, the intercepts of regression lines represent fixed costs, while the
slopes of the regression lines represent variable costs. By comparing the values of intercepts
and slopes of the regression models, we can discern any differences in those shipping costs.
To elaborate, when a shipper totally relies on a spot market for acquiring transportation
services from common carriers on an as-needed basis, variable costs or costs per mile would
be relatively high, while fixed costs are minimal. On the other hand, when a shipper uses its
own fleet, the portion of fixed costs would be much higher than that of variable costs. In the
case of using a contract carrier with a long-term contract, fixed and variable costs can be
more equally allocated than those of the two aforementioned shipping options (i.e. private vs
common carriage use). More importantly, notice that even if two shippers made the same
carrier choice, their shipping costs still may differ because the variable cost component of
shipping costs may differ depending on shipping distance. Figure 1 illustrates the impact
of aforementioned carrier choice options on shipping costs with three possible scenarios of
outcomes resulting from the proposed regression models. To be specific, Figure 1(a) shows
an example that includes two regression lines with different intercepts but the same slopes,
indicating that two shippers incur different fixed costs but were charged the same rate per
distance or mile. Figure 1(b) exhibits a case where two shippers have the same fixed costs,
but were charged with different shipping rates per mile. Figure 1(c) displays a situation
where two shippers incur different fixed and variable costs.
To verify the underlying premise that distance (miles traveled) affects shipping rates for
truckloads of products in the same category, we developed another regression model with
total charges (Charge) as the dependent variable and miles (Mileage) as the explanatory
variable, along with a binary (intercept) dummy variable representing the presence of
shipping rate differentials. This allows us to confirm any statistical significance in the

(a) (b) (c)

L Shipper L Shipper L Shipper

Charge ( )

Charge ( )

Charge ( )

W Shipper W Shipper W Shipper

Figure 1.
Regression lines for
different carrier
choices
Mileage Mileage Mileage
difference between two intercepts (shipping charge differentials). This regression model can Cost-attractive
be mathematically expressed as: transportation
yi ¼ b0 þb1 xi þgdi þei (1) services
where yi is a dependent variable for shipper i; xi is an independent variable for shipper i; di
is a binary dummy variable for an intercept (fixed shipping charge for shipper i); β0 is the
constant; β1 is the coefficient of the independent variable; γ is the coefficient of the dummy 199
variable; εi represents an error term of the regression model. When we pool the data sets of
shippers and estimate regression Equation (1) with d1 ¼ 0 for L Shipper and d2 ¼ 1 for W
Shipper, Equation (1) for L Shipper becomes yi ¼ β0 + β1xi + εi, while Equation (1) for
W Shipper becomes yi ¼ ( β0 + γ) + β1xi + εi. Depending on the value of γ, the intercepts of
two models may vary and thus yield different intercepts. Since regression Equation (1)
still cannot explain how different outsourcing strategies influence shipping rate
differentials, we decided to introduce another dummy variable, producing the regression
model with two dummy variables for both intercepts and slopes specified below:
yi ¼ b0 þb1 xi þb2 xi d i þgdi þ ei (2)
where β2 is the coefficient of the slope dummy variable, and xidi is the slope dummy variable,
also known as an interaction term. By confirming the significance of β2, we could confirm the
slope difference between two groups. To be specific, in the case of L Shipper with d1 ¼ 0,
regression Equation (2) becomes yi ¼ β0+β1xi+εi. On the other hand, for W Shipper d2 ¼ 1,
Equation (2) becomes yi ¼ ( β0 + γ) + ( β1 + β2) xi + εi. Herein, β0 and γ represent the amount of
fixed shipping charges assessed for W Shipper, while β1 and β2 represent variable charges
paid by W Shipper. One may argue that it is possible for us to estimate two regression models
separately without using dummy variables. If so, however, it would be meaningless for us to
compare regression coefficients, since a goodness-of-fits of two models is incompatible. Thus,
it should be noted that we pooled the data and estimated the coefficients in one single
equation. Another motivation for using this dummy regression model is to account more fully
for the dependent variable, by making the errors smaller, and to avoid a biased assessment of
the impact of an independent variable (e.g. mileage), as a consequence of omitting another
independent variable (e.g. shipper preference) that is related to it (Gujarati, 1970).
Herein, we included two dummy variables in the model so as to test any differences
between intercepts and slopes in the regression model. The inclusion of dummy variables
allows us to verify the difference in shipping costs that may be due to the varying carrier
choices or transportation outsourcing strategies employed by different shippers.

4. Results and discussions

We tested a series of regression models by choosing Charge as a dependent variable and
Mileage and two dummy variables as independent variables. For an intercept dummy
variable, we assigned 0 to L Shipper and 1 to W Shipper. We constructed the slope dummy
variable by multiplying the intercept dummy variable by Mileage. Table II shows
descriptive statistics for this regression model results.
It is somewhat surprising to find that the median Mileage was 754 miles, defying the
common rule of thumb that rail is the more cost-efficient mode of transportation for any
shipping distance over 600 miles than trucks. Perhaps, the shippers chose trucks instead of rails
due to their door-to-door service capability and schedule flexibility. Another reason may be a
reduced risk of shipment damages due to less commotion by trucks as compared to rails. Also,
notice that two variables showed a great degree of variability (e.g. ranging from $255 to $4,969
of shipping charges, 55 to 2,923 miles of trips) as summarized in Table I. As expected, Figure 2
displays a strong and positive correlation between the two variables, Charge and Mileage.
IJLM Variable Minimum Maximum Median Mean SD Sample size
28,1
L Shipper
Charge (in $) 255.00 2,900.00 652.00 1,141.14 688.63 101
Mileage 55.00 2,285.00 960.00 749.22 578.29 101
W Shipper
200 Charge (in $) 350.00 4,969.10 906.00 1,711.49 1,181.07 129
Mileage 83.00 2,923.00 1,462.00 1,214.74 887.65 129

Table II. Total

Descriptive statistics Charge (in $) 255.00 4,969.10 1,120.48 1,461.02 1,033.07 230
for variables Mileage 55.00 2,923.00 754.00 1,010.32 800.04 230

5,500.00
L Shipper

5,000.00 W Shipper

4,500.00

4,000.00

3,500.00
Charge ( )

3,000.00

2,500.00

2,000.00

1,500.00

1,000.00

500.00

Figure 2.
A scatter diagram for 0.00
0 500 1,000 1,500 2,000 2,500 3,000
charge and mileage
Mileage
This result is congruent with the finding of Skinner et al. (2009) who confirmed that distance Cost-attractive
(miles) explained a 75-80 percent variation in shipping costs, with an adjusted coefficient of transportation
determination of 94 percent. Similarly, Özkaya et al. (2010) observed that mileage (along services
with shipment weight) was the most important variable for estimating actual freight rates.
Also, to examine how significantly the spot-market purchase of common carrier services
and/or the extensive use of contract carrier services influences shipping costs, we conducted
three-step hypotheses tests; separate estimation of regression coefficients for each shipper; 201
examination of differences in intercepts (or fixed costs) of the regression equations by
pooling data for both shippers and introducing a dummy variable for intercepts;
examination of differences in slopes (or variable costs) of the regression equations by
including an additional dummy variable in the pooled data.
Table III summarizes the results of regression analysis for L Shipper with the adjusted
R2 of 0.907 which indicates that 90.7 percent of the variance in the dependent variable
(Charge) can be predicted from the independent variables. The result also shows that
estimated constant and coefficient were significant at α ¼ 0.01.
Table IV exhibits the result of regression analysis for W Shipper with the adjusted R2 of
0.902. The result also shows that estimated constant and coefficient were significant at α ¼ 0.01.
To compare these two results pictorially, we drew Figure 3 which displayed that the
intercepts of the two regression lines were different, but the slopes of those lines ran in parallel.
However, since these results still did not suggest any statistical differences in intercepts, we
introduced a dummy variable (Shipper) for the intercepts: 0 for L Shipper, and 1 for W Shipper
and tested the statistical similarity or dissimilarity of the intercepts by pooling data for the two
shippers. As recapitulated in Table IV, the result reveals that Shipper, the dummy variable for
intercepts, turned out to be statistically significant. In other words, a difference in the fixed costs
between two shippers was statistically significant. Therefore, we rejected H1. Indeed, we found
that the fixed shipping costs of L Shipper were higher than those of W Shipper. This finding
indicates that L Shipper was charged more than W Shipper for equivalent transportation
services due to a greater minimum base charge and accessorial/fuel surcharges (Table V).
To test H2, we added another dummy variable for slopes, Shipper × Mileage, by entering
0 for L Shipper (0 × Mileage) and Mileage (1 × Mileage) for W Shipper, respectively. We ran
this modified regression model with two dummy variables: one for intercepts and the other
for slopes. The result of this model is presented in Table VI.
The result reveals that Shipper turned out to be marginally significant (at α ¼ 0.10) and a
dummy variable for the slopes of the two shippers was insignificant ( p-value ¼ 0.937). The
adjusted R2 of this model was 0.907. Thus, this result supports H2. Also, we found no
heteroscedasticity indicating that the variability of Shipper variable is constant over time, since
a plot of error terms against predicted values showed no discernible patterns. To summarize, we

Unstandardized coefficients
B SE t-value Level of significance
Table III.
Constant 424.368 51.374 8.260 0.000 Regression result
Mileage 1.700 0.054 31.259 0.000 for L Shipper

Unstandardized coefficients
B SE t-value Level of significance
Table IV.
Constant 242.859 74.182 3.274 0.001 Regression result
Mileage 1.693 0.049 34.294 0.000 for W Shipper
IJLM
28,1 6,000.00
L Shipper

W Shipper
5,000.00
202

4,000.00
Charge ( )

3,000.00

2,000.00

1,000.00

0.00
0

200

400

600

800

1,000

1,200

1,400

1,600

1,800

2,000

2,200

2,400

2,600

2,800

3,000
Figure 3.
Regression lines for
L and W shippers
Mileage

Unstandardized coefficients
B SE t-value Level of significance
Table V.
Regression result Constant 428.129 50.440 8.488 0.000
with a dummy Mileage 1.695 0.037 46.207 0.000
for intercepts Shipper −187.293 59.002 −3.174 0.002

Unstandardized coefficients
B SE t-value Level of significance

Constant 424.368 69.592 6.098 0.000

Table VI. Mileage 1.700 0.074 23.076 0.000
Regression result with Shipper −181.509 94.367 −1.923 0.056
two dummy variables Shipper × Mileage −0.007 0.085 −0.079 0.937

found that a difference in fixed costs between two shippers was statistically significant as
illustrated above. That is to say, unsuspected shippers who are unware of benchmark freight
rates can be charged more than other shippers, not necessarily because of discriminatory pricing
but because of unwise transportation outsourcing and carrier choice strategies.
5. Managerial implications Cost-attractive
In a typical business setting, order fulfilment costs account for the large portion of business transportation
expenditures. Their key components include shipping costs which directly affects the services
company’s profitability. In the era of globalization and stretched supply chains with
dispersed customer bases, the company’s survival can rest on its ability to control its rising
shipping costs. However, the complexity and uncertainty involved in the way shipping
(freight) rates are determined have created confusion and unintended discriminatory pricing 203
structures for some unsuspecting shippers. This paper is one of the first to examine the
presence of instability or significant swings in shipping charges as a result of greater
bargaining room created by a series of transportation deregulatory acts. As expected, we
discovered that different shippers were charged with different shipping rates, but only for
the fixed components of shipping charges such as minimum or base-line charges when the
shippers enter shipping contracts. From a practical standpoint, though price differentials
may not be unique to the trucking industry, this finding implies that carriers are likely to
give some shippers priority over others with a certain transportation outsourcing and/or
contract negotiation situation (e.g. limited negotiation leverage resulting from market
conditions, constrained carrier capacity, and uneconomical shipping volume). Also, some
carriers estimate their cost to serve each of their shippers (customers) and then identify
favorable shippers who would enable them to improve their shipping efficiency. For
example, shippers who load their shipments more quickly without increasing driver wait
time or incurring higher lumper costs, clearly label their packages, provide repeat business
opportunities, and offer better payment terms could be favoured by the carrier over others.
Although this study cannot fully uncover the dynamic transportation price mechanism
and key determinants for favorable pricing, the study results confirmed the presence of
price differentials that were affected by particular carrier choice and transportation
outsourcing strategy. Thus, to determine benchmark shipping rates for particular
transportation services, shippers themselves should conduct their own cost analysis and
estimate the minimum fixed charges for those services. In addition, to mitigate the carrier’s
bias against a certain shipper, we suggest that the shipper’s cost analysis should be based
on the activity-based costing concept as opposed to conventional fixed and variable cost
breakdowns. Another way to develop a benchmark rate is to share actual shipping charges
assessed for particular services across multiple shippers through shippers’ association and
then identify the lowest minimum charge as the benchmark rate. This benchmark rate can
help shippers increase leverages for freight rate negotiation. This benchmark shipping rate
also can be a basis for future price hedging. Shipping charge hedging can be an important
contractual tool that can reduce the shipper’s exposure to volatile and potentially rising
transportation charges. More importantly, shippers should be aware that such a benchmark
rate cannot be developed without sharing a history of negotiated shipping price data with
their peers. Furthermore, shippers may enjoy more favorable pricing options by placing
their freight for a competitive bid to a selected group of prospective carriers and then
figuring out the lowest shipping rate offered by these carriers. In particular, the use of
constraint-based bids which constrain carriers in different ways (e.g. penalizing for the
carrier’s failure to perform certain tasks such as on-time delivery and real-time shipment
status notice) may help the shipper weed out the low-cost carrier whose performance is
lagging behind the industry norm.

6. Concluding remarks and future research directions

The main contribution of this paper is its attempt to uncover shipping cost dynamics from a
shipper’s point of view and then identify the sources of shipping cost differentials between
different shippers. The increased understanding of shipping cost dynamics will not only
help shippers improve their freight rate negotiation leverage, but also help them set a
IJLM benchmark freight rate that is considered “fair” and “reasonable” to both shippers and
28,1 carriers. Therefore, this study can be valuable guidance or a precursor for more rigorous
benchmarking studies of transportation pricing. From a theoretical viewpoint, we proposed
a novel regression model with an intercept and slope dummy variables to discern any
statistical differences in fixed and variable shipping charges. This proposed additive
dummy regression model can also be used for benchmarking shipping costs within the same
204 transportation outsourcing strategy or across different transportation outsourcing
strategies. Built upon this current study, this study can be extended to include a longer
period (multiple years) of historical shipping data and data sources obtained from a large
number of shippers. Also, the future study can look into the extent of shipping cost
differentials in different industry sectors.

References
Beamon, B.M. (1999), “Measuring supply chain performance”, International Journal of Operations and
Production Management, Vol. 19 No. 3, pp. 275-292.
Case Logistics (1980), “11th annual state of logistics report”, unpublished report, Council of Logistics
Management, Oak Brook, IL.
[Link] (2014), “FedEx to increase shipping rates for express, ground and freight services”,
available at: [Link]
14/FedEx-to-Increase-Shipping-Rates-for-Express-Ground-and-Freight-Services/[Link]
(accessed January 18, 2015).
Galletti, D.W., Lee, J. and Kozman, T. (2010), “Competitive benchmarking for fleet cost management”,
Total Quality Management, Vol. 21 No. 10, pp. 1047-1056.
Gilmore, D. (2014), “State of the logistics union 2014”, Supply Chain Digest, available at: [Link].
com/ASSETS/FIRSTTHOUGHTS/[Link]?cid=8190 (accessed June 22, 2015).
Gujarati, D. (1970), “Use of dummy variables in testing for equality between sets of coefficients in two
linear regressions: a note”, The American Statistician, Vol. 24 No. 1, pp. 50-52.
Gunasekaran, A., Patel, C. and Tirtiroglu, E. (2001), “Performance measures and metrics in a supply
chain environment”, International Journal of Operations and Production Management, Vol. 21
Nos 1/2, pp. 71-87.
Havenga, J. and Simpson, Z. (2014), “Reducing national freight logistics costs risk in a high-oil-price
environment: a South African case study”, International Journal of Logistics Management,
Vol. 25 No. 1, pp. 35-53.
Huddleston, T. Jr (2014), “Sending packages via UPS will soon cost more”, Fortune, October 20,
available at: [Link] (accessed January 18, 2015).
Kay, M.G. and Warsing, D.P. (2009), “Estimating LTL rates using publicly available empirical data”,
International Journal of Logistics: Research and Applications, Vol. 12 No. 3, pp. 165-193.
Moore, T.G. (2008), “Surface freight transportation deregulation”, Economics Library, available at:
[Link]/library/Enc/[Link] (accessed
December 22, 2014).
Özkaya, E., Keskinocak, P., Joseph, V.R. and Weight, R. (2010), “Estimating and benchmarking
less-than-truckload market rates”, Transportation Research Part E: Logistics and
Transportation Review, Vol. 46 No. 5, pp. 667-682.
Penske Logistics (2014), “25th annual state of logistics”, unpublished report, Council of Supply Chain
Management Professionals, Oak Brook, IL.
Penske Logistics (2015), “26th annual state of logistics”, unpublished report, Council of Supply Chain
Management Professionals, Oak Brook, IL.
Porter, J. (2012), “10 ways to trim shipping costs”, Entrepreneur, October 17, available at: www.
[Link]/article/224619 (accessed January 12, 2015).
Randers, J. and Göluke, U. (2007), “Forecasting turning points in shipping freight rates: lessons from Cost-attractive
30 years of practical effort”, System Dynamics Review, Vol. 23 Nos 2/3, pp. 253-284. transportation
Shepherd, C. and Günter, H. (2011), “Measuring supply chain performance: current research and future services
directions”, in Fransoo, J.C., Waefler, T. and Wilson, J.R. (Eds), Behavioral Operations in Planning
and Scheduling, Springer Berlin Heidelberg, pp. 105-121.
Sillanpää, I. (2015), “Empirical study of measuring supply chain performance”, Benchmarking:
An International Journal, Vol. 22 No. 2, pp. 290-308.
205
Skinner, W., Maserang, D. and Caplice, C. (2009), “The future of transportation rates and fuel
surcharges – the shipper’s view”, paper presented at the annual global conference of the Council
of Supply Chain Management Professionals, Chicago, IL, September 20-23.
The European Conference of Ministers of Transport (1998), “Rail restructuring in Europe”,
unpublished report, ECMT Publications, Paris.
Wong, P.W. and Wong, Y.K. (2008), “A review on benchmarking of supply chain performance
measures”, Benchmarking: An International Journal, Vol. 15 No. 1, pp. 25-51.

Further reading
McKinnon, A.C. (2009), “Benchmarking road freight transport: review of a government – sponsored
programme”, Benchmarking: An International Journal, Vol. 16 No. 5, pp. 640-656.
Menachof, D. and Wassenberg, O. (2000), “The application of benchmarking techniques by road
transport companies in the United Kingdom and the Netherlands”, Transportation Journal,
Vol. 40 No. 2, pp. 40-56.
Wiegmans, B.W. and Donders, A.R.T. (2007), “Benchmarking European rail freight transport
companies”, Transportation Journal, Vol. 46 No. 2, pp. 19-34.

Corresponding author
Hokey Min can be contacted at: hmin@[Link]

For instructions on how to order reprints of this article, please visit our website:
[Link]/licensing/[Link]
Or contact us for further details: permissions@[Link]
Reproduced with permission of copyright owner. Further
reproduction prohibited without permission.