Content uploaded by Michael H. Belzer
Author content
All content in this area was uploaded by Michael H. Belzer on Nov 26, 2021
Content may be subject to copyright.
The Economics of Safety:
How Compensation Affects
Commercial Motor Vehicle Driver
Safety
Michael H. Belzer
Associate Professor
Department of Economics
2074 Faculty / Administration Building
656 W. Kirby
Wayne State University
Detroit, MI 48202
Michael.H.Belzer@wayne.edu
313-577-1328
Presented to
United States House of Representatives
Committee on Small Business
July 11, 20121
WORKING PAPER
1 This paper relies in part on a study conducted by Prof. Belzer along with his colleagues,
Daniel Rodriguez (University of North Carolina – Chapel Hill) and Stanley A. Sedo (University of
Michigan).
2
Executive Summary
On June 1, 2011 a discount intercity bus carrying 59 people to New
York’s Chinatown crashed, killing four people and injured more than 50 others.
The carrier had a long history of violations and crashes and a safety rating far
worse than the rest of the intercity bus industry. A driver fatigue rating of 86
on a scale of 1 to 100 meant that before the crash, federal officials had rated it
among the most unsafe bus carriers. Its driver fitness rating of 99.7 meant that
it ranked in the bottom 1%.
Sky Express should not have been on the road and after the crash the
FMCSA gave it an unsatisfactory rating and banned it from interstate service.
Though the ban was too late for the victims, under US regulations this still did
not prevent the company from continuing to operate intrastate. Safety
advocates’ calls to require seatbelts, stronger roofs, more driver training, and
other regulatory changes do not address the problems that led to the crash and
would not prevent future crashes.
Intensified competition created by deregulation, without proper
safeguards, created this safety problem. We do not have to repeal deregulation
to solve it, but we have to address the problems this intense competition
creates. If insanity is “doing the same thing over and over and expecting a
different result”, we are insane. Preventable crashes like this will happen again
for the same reasons, regardless of how many times we rework the algorithms
of CSA or scrap it and replace the entire program altogether. In short, the
safety problems that CSA attempts to address will not be remedied until we
begin to address the systemic problems in the trucking industry.
I have examined the link between commercial motor vehicle (CMV) driver
compensation and work pressure, and driver safety. Research establishes a
pay-safety link that is important for policy because it shows that economic
forces inherent in transport competition tend to produce unintended safety and
health consequences drivers and passengers.
My full report on the economics of safety applies to both truck and bus.
Transport deregulation brought lower consumer prices, but this bus crash
showed the dark side. Deregulation has increased competition among carriers
in all modes hauling both passengers and freight, and reduced compensation.
CSA in its current form places pressure on drivers without addressing
underlying causes. In the trucking industry, inadequate compensation for
drivers causes a misperception of a driver shortage that isn’t there, and causes
many to look for cheaper labor, such as that found in Mexico. Everyone who
has passed introductory economics knows that more drivers will be attracted to
trucking by a better job package, including compensation. Opening the border
to Mexican truck drivers will bring more of the same, as Mexican drivers
compete with American small business drivers and employees at ¼ the cost,
intensifying competition among US motor carriers and lowering driver hiring
3
standards. No regulation can overcome the effect of markets that drive down
price.
This creates a sustainability problem. The CMV driver’s workplace is the
public highway and unsafe drivers become a public hazard—–what we call a
“negative externality”. While people buy transport services for an apparent
market price, it does not include safety and health cost. Economic efficiency
requires that price incorporate all costs and benefits associated with
commercial movement, and failure to incorporate the full safety and
environmental cost sends incorrect signals to the market, creating an implicit
public subsidy of unsafe operators.
If the insurance market worked perfectly, the risk associated with low-
paying carriers would show up in higher cost insurance. This market does not
work well because insurance companies cannot rate motor carriers and charge
accordingly. Big crashes are low-probability, high-impact events that insurance
companies don’t like. FMCSA regulations only require truckers to carry
$750,000 of insurance per incident.
Economic forces play a strong safety role because carriers that pay more
money can hire better drivers. Efficiency wages paid by these carriers leads to
better performance because drivers know that their jobs are better than their
alternatives, providing incentives to drive safely.
These findings are consistent with economic theory because we expect
that carriers pay drivers their market value, determined by their personal
employment history, driving record, training and education, experience, driving
skills, temperament, and other factors. These factors explain the differences in
safety outcomes.
For every 1% in pay, we have found 1% to 4% better safety. Higher pay
produces better carrier and driver safety. We don’t yet know whether safety
pays, but clearly driver pay strongly predicts safety.
Low price doesn’t necessarily mean low-cost. Since in an efficient market,
price should include all costs, the environmental and safety costs associated
with cheap labor and cutthroat competition create unsustainable supply
chains that make everyone less well off.
Three solutions would go a long way to resolve this problem.
1. Get government regulators out of their silos. FMCSA and the Department of
Labor should cooperate with the industry to engage in a dialogue to promote
economic conditions that improve highway safety. The DOL has the
authority to regulate compensation and perhaps it is time to reconsider
certain exemptions for the trucking industry under the Fair Labor
Standards act.
2. Implement Chain of Responsibility regulations like those enacted by the
Australian Parliament to create a level playing field in a deregulated
environment. The owner-operator model is a valuable one and we need to
4
preserve small businesses in the trucking industry. Other nations, like
Australia, maintain a deregulated industry while supporting small business
truckers and without compromising safety. One way to do this is to address
underlying systemic problems such the failure to pay truckers for loading
and unloading.
3. Tighten regulations on subcontracting that balances the power between
contractors and trucking companies, as Australians have done. Court
rulings 40 years ago usurped legislative authority, disallowing traditional
cooperation among owner-drivers to negotiate with carriers. This would give
owner-drivers a fair shake. In short, help level the playing field by giving
small businesses more negotiating power to keep costs low and safety
benefits high.
5
The Economics of Safety:
How Compensation Affects
Commercial Motor Vehicle Driver
Safety
Michael H. Belzer
Associate Professor
Department of Economics
Wayne State University
Commercial motor vehicle (CMV) drivers usually paid on a piecework
basis, which is a source of confusion and misunderstanding for public policy
makers. While this is almost universally true for intercity truck and bus
drivers, it has become routine in recent years to pay local drivers – especially
owner-drivers – a flat rate per move or a percentage of revenue earned by the
shipment, rather than an hourly wage. Generally road drivers are paid by the
mile (or a percentage of revenue) and not paid for loading, loading, and other
delays (Burks et al. 2010). This leads to strong incentives to lie on one’s logs,
logging only “paid” time (driving time) on duty and logging all other work time
as “off duty” in order to conserve hours available to work. Since surveys
suggest that 25% of the average driver’s day is unpaid non-driving time, this
can easily mean that truckers can drive as much as eleven hours and work an
additional three hours more than they log every day, and still appear to be
legal. But they’re not. This undocumented fatigue and documented work
pressure contributes substantially to crashes.
I. Introduction
Compensation can influence worker behavior in several ways. Yellen
suggests that an employer paying higher than average “efficiency” wages (wages
above the market-clearing level that serve to attract a superior workforce) will
discourage workers from “shirking”, or failing to put full effort into their work,
since losing their job imposes a cost on the worker (they reduce their chances
of getting another good job and risk sinking to a lower tier company). If the cost
of monitoring workers is higher than that of increasing wages, Yellen argues
that this can be a cost-efficient way for the employer to elicit effort from
workers (Yellen 1984). In addition to the level of compensation, the type of
payment also can influence worker behavior. The “piecework” payment system
has a long history of providing an incentive for workers – especially transport
workers in general and contract workers in specific – to increase their effort
(Belzer 2000, 2011). While the efficiency wage argument appeals to the long-
6
run interest of the worker to maintain employment, the piecework system is
designed to create an immediate incentive to increase production by paying
higher wages to those workers who are more productive.
Almost all both truckload (TL) and less-than-truckload (LTL) intercity
drivers are paid by the mile or in some manner by the load, rather than an
hourly wage. This method of pay is so pervasive that in the industry, mileage
often is the sole determinant of compensation, regardless of what other work
the driver does.
The treatment of loading and unloading time is a good example. Drivers
frequently wait long periods of time for their loads, and in many cases must
load or unload their own freight. However, these drivers are underpaid, relative
to the value of their driving time, or not paid at all, for this work. This paper
raises the hypothesis that while these compensation practices may be useful in
getting drivers to work harder, they also create incentives that threatens public
safety and security (Belzer and Swan 2011).
Both the method and level of compensation in the trucking industry
create short-run economic incentives that may lead to unsafe driving practices.
These behaviors may include neglecting safety inspections and repairs as well
as driving too fast for conditions (and faster than legally allowed). Because long
work hours, especially when driving, is associated with intensified health and
safety risks, truck drivers’ hours of driving and hours of work (“hours of
service”, or “HOS”) have been limited since the 1930s (Belzer 2008; Belzer et al.
1999; Belzer et al. 2002; for a brief history of this regulatory framework, see
Belzer 2000).
Piecework compensation practices, along with unpaid non-driving labor
time, can lead drivers to work more than the number of hours allowed by the
hours-of-service rules. Drivers may require a minimum or ‘target’ level of
income that is necessary in order to meet basic living expenses. If the mileage
rate is sufficiently low so that this target cannot be reached, drivers may feel
compelled to work more hours than legally allowed, and economic theory
supports this expectation. The risk created by these incentives may be greater
under conditions where non-driving time earns a lower rate per hour relative to
that earned when driving, or not paid at all for loading and unloading. In these
instances, there is an incentive to underreport the amount of time spent on the
lower- or un-paid loading time in order to conserve available hours for the
relatively higher paid driving time. This underreporting of loading and
unloading time, combined with additional driving time to make up for this
unpaid time, means that drivers might often work – and drive – more hours
than allowed by law.
While this may provide short-run economic benefit to the drivers, in the
end it would cause truck drivers to provide an excessive supply of labor to the
marketplace for a fixed number of workers, driving wage rates down and
encouraging additional hours of work. Given a fixed labor market, each
7
individual driver will tend to work more hours than allowable and this
“sweating” of labor will encourage each individual driver to work even harder
and longer, increasing the number of hours provided to the market and
effectively expanding the labor market artificially, increasing all drivers’ crash
risk accordingly. These longer hours create safety concerns that affect not only
the industry but the broader population as well. If the market for individual
driver services insufficiently captures the cost of this additional safety hazard,
it would create a market imperfection that might have significant policy
consequences. In short, low driver wages and poor working conditions impose a
real and tragic cost to the nation through decreased highway safety.
II. Theory and Evidence
Introduction
Employee earnings levels and the method of compensation likely have an
influence on employee behavior. This research shows that the level and method
of compensating truck drivers affects their driving and non-driving behavior,
which ultimately influences their involvement in crashes.
Truck driver attitudes and behaviors have been studied in various
contexts. In most cases, the motivation for these studies is to understand the
immediate mechanisms that influence certain driver behaviors. These studies,
however, often focus on particular behaviors (e.g., speeding, working – and
especially driving – excessively long hours, and not getting enough sleep) rather
than confronting the factors that motivate such behaviors at different
organizational levels. Such factors can include economic pressures, personal
characteristics, pay rate, and the compensation method itself, among others.
From the driver’s perspective some consideration has been given to the
compensation issue and its influence on safety. Pay level has been studied
more consistently than pay method. Low levels of pay have been considered by
many as a motivator of long driving hours, illegal substance use, the onset of
fatigue, and other practices and phenomena (General Accounting Office - U.S.
Congress 1991; Hensher et al. 1991; Saccomanno, F. F., Craig, and Shortreed
1997). Other studies, however, have suggested that truck driver compensation
level has a less important role than the one regularly attributed to it (McElroy
et al. 1993).
Groups of drivers participating in different focus groups have
characterized the prevailing piece rate (per mile) compensation method as
limiting income and encouraging cheating (Cadotte, Sink, and Chatterjee 1997;
Mason Jr. et al. 1991) . Drivers readily identified the compensation system in
place as a motivation for unsafe driver behavior. Piece rate systems coupled
with hours of service regulations limit the income opportunities of drivers
(Chatterjee et al. 1994). Forty-five percent of respondents to a New York State
8
driver survey thought it would be useful to pay by the hour in order to reduce
driver drowsiness (McCartt, Anne, Hammer, and Fuller 1997b).
Management also has recognized the importance of better understanding
driver compensation. A 1995 mail survey of 1,464 drivers at 57 for-hire
truckload dry van, flatbed, refrigerated, and tank carriers showed that an
overall driver compensation factor emerged as the important dimension for
human resources improvement (Stephenson Jr. and Fox 1996). Similarly, in a
survey of 148 trucking company personnel managers, other researchers found
that managers believed that pay level was the most important factor in drivers’
choice of motor carriers for employment (Southern, Rakowski, and Godwin
1989).
Work pressure and economic pressure have contributed to workplace
hazards and even “disasters” across many industries and countries. A recent
report on the Massey mine explosion, for example, points directly to the role of
economic pressure and the very real drive for profits as a primary cause of
catastrophic industrial safety failures. A study by the West Virginia Governor's
Independent Investigation Panel charges that the 2010 explosion that claimed
the lives of 29 miners was an entirely preventable disaster that resulted from
the fact that “Massey Energy put coal production ahead of safety” (Berkes
2011; Governor’s Independent Investigation Panel 2011). Studies from
Australia, which is in many ways similar to the United States, have
consistently found economic pressure to be the root cause of safety problems in
the trucking industry (Quinlan 2001; Quinlan, Mayhew, and Johnstone 2006).
The same economic pressures have been found to be at the root of major safety
failures in Australian mining (Quinlan 2007; see especially paragraphs 818 and
836), U.S. oil extraction (Crooks 2011; National Commission on the BP
Deepwater Horizon Oil Spill and Offshore Drilling 2011; Urbina 2010), and in
airlines (National Transportation Safety Board - U.S. Department of
Transportation 2010; Young 2010)
The Role of Employee Compensation
Compensation generally acts as a pricing mechanism, but
compensation’s impact on employees, especially drivers, is much more
complex. As a method of allocating resources, employee earnings are a pricing
mechanism used to direct labor to its most productive use. This function, very
much in line with traditional microeconomics, explains variations in the
distribution of earnings as emerging from the interactions of supply and
demand where certain observable characteristics are taken into account.
A second role of compensation is to serve as a tool for social stratification
and cohesion. In this role, employee earnings are seen as a prime determinant
of standard of living. Earnings play the role of providing social legitimacy within
organizations and society. Compensation policies play a role in determining
what is a “fair” wage level (Akerlof, Rose, and Yellen 1988; Akerlof and Yellen
1988, 1990).
9
No previous study has utilized efficiency wage theory to explain CMV
driver safety. Compensation can serve as a management tool to elicit higher
employee effort and align employees’ core skills with the organization's
interests. Multiple theories attempt to explain the role of pay in the
employment relationship. They include the transaction cost perspective
(Williamson, Oliver E. 1975), where opportunistic behavior is to be minimized,
as well as the efficiency wage approach (Holzer 1990; Lazear 1990; Weiss 1990;
Yellen 1984), in which above-market wages result in desired behavioral
outcomes for a group of employees. These outcomes can range from reduced
shirking and enhanced effort (Yellen 1984) to adherence to hours of service
regulations, behaviors oriented towards reducing risk of fatigue and dozing
while driving, and generally safe-driving behaviors. However, safety research
generally has steered toward behavioral explanations and avoided economic
explanations, and efficiency wage theory may provide a better explanation for
outcomes.
Recent changes in wage structures, such as the impact of economic
deregulation, have created increased interest in the roles that compensation
plays in society (Rubery, Jill 1997). Belzer traced the post-regulation transition
from regulation-related truck industry segmentation to market segmentation,
and the resulting impact on industrial relations, including compensation
practices. He modeled wage levels as a function of a variety of firm-level factors
including industry segment, average haul, unionization, market share,
profitability, and location variables such as urbanism and regionalization.
Unionization and industry sector (LTL) were most strongly associated with
higher wages. He also found that market share affected wages positively
(consistent with previous findings) as did location (Southern carriers had
significantly lower wages) (Belzer 1995a).
Compensation Level
Compensation level is often framed in the context of a hierarchical
conception of pay (Milkovich and Newman 1993), where the compensation
system is disaggregated into its fundamental components, such as method,
level, changes in earnings over increasing job tenure and similar factors.
Employee compensation is understood as the overall employee earnings during
a specific period, including direct compensation (e.g., wages) and deferred
compensation (e.g., pension plans).
Direct Compensation
Organizations can have varying pay levels, depending on the flow of work
and the organization, yet we often observe pay differences between similar jobs
in similar organizations (Chen 1992; Leonard 1987; Seiler 1984). Weiss
provides a useful summary of issues associated with direct compensation
(Weiss 1990). The literature consistently shows that increases in relative wages
(after controlling for occupation and human capital) are associated with
increases in productivity.
10
In a series of studies of driver compensation using individual driver-level
data, cross-sectional motor carrier data, and individual driver survey data,
researchers showed that the relationship between compensation and safety
ranges from .92:1 in the cross-sectional study of 102 TL carriers to as much as
4:1 in the firm-level case study of JB Hunt (Rodriguez et al. 2003; Rodriguez,
Targa, and Belzer 2006; Belzer, Rodriguez, and Sedo 2002). In the Hunt study,
researchers found that for every 10% higher driver pay rate, at the mean,
drivers had a 34% lower probability of crash, month-to-month. In addition, for
every 10% of pay raise, drivers had a 6% lower crash probability (Rodrigue
2006; Rodriguez et al. 2003; Rodriguez, Targa, and Belzer 2006). In the cross-
sectional study, for every 10% higher compensation level for truck drivers
working for non-union truckload carriers, the carrier had a 9.2% lower crash
rate. The driver’s mileage pay rate explained half the difference and other
compensation factors explained the remainder (Belzer, Rodriguez, and Sedo
2002). Finally, an individual survey conducted by the University of Michigan
Trucking Industry Program showed that at the mean, a 10% higher
compensation level predicted a 25% lower crash probability for the year.
There is less agreement about the magnitude of the effects and whether
the increase in productivity can pay for the wage increase (Levine 1992). It also
is difficult to disentangle cause and effect, or whether the effect is due to
selection or performance incentive.
Efficiency wages
A “market-clearing” wage clears the market of unemployed workers –
absorbs all available unengaged labor or achieves full employment in a specific
labor market – at a compensation level sufficient to attract enough workers to
the jobs that pay enough to attract labor to do them. Markets do not clear
when companies offer workers a lower package of compensation than they
could get doing something else. This is why economists argue that there is no
such thing as a “labor shortage” in any labor market but rather a shortage of
compensation sufficient to attract labor. As demand for labor increases,
companies should be willing to raise wages enough to attract the necessary
labor.
Theorists of “efficiency wages” argue that some employers do not pay
market-clearing wages. Instead, they offer above market-clearing wages that
induce employees to be more efficient. This efficiency increase can occur in
several ways.
Reduction in shirking. Since employees have a higher compensation level
with efficiency wages than they would have otherwise, the cost of discharge due
to shirking behavior is higher. This reduces worker shirking because the job
they have already rewards them above the average market-clearing wage for the
industry, and if they lose their job because of poor performance they likely will
11
have to take an inferior job.2 Some research suggests that greater wage premia
are in fact associated with lower levels of shirking as measured by disciplinary
dismissals (Cappelli and Chauvin 1991; Yellen 1984) . However, shirking and
discipline also are dependent on whether a worker sees the relationship
between shirking and the difficulty in finding alternative employment (Groshen
and Krueger 1990).
Quality of workers. It is reasonable to expect, and empirical research has
shown, that high compensation levels attract more qualified workers than do
lower compensation levels (Chen 1992; Groshen and Krueger 1990). This is the
“creaming effect.” Acting as a mechanism for selection, the compensation level
attracts more productive employees. Positive consequences often associated
with having a more qualified pool of workers include the reduced need to
supervise employees and a reduction of employee shirking. For example,
Groshen and Krueger found that hospitals that paid high wages to staff nurses
employed fewer supervisors (Groshen and Krueger 1990). It is unclear,
however, if this is due to greater work effort from the average existing nurse
workforce (the efficiency wage) or because higher wages attract better nurses
who needed less supervision (the creaming effect).
Turnover costs. Higher wages may tend to reduce turnover. Turnover
costs include advertising, search, and training costs (Arnold, Hugh J. and
Feldman 1982; Becker 1975; Chen 1992; Cotton and Tuttle 1986; Salop and
Salop 1976) One study of high school graduates correlated higher wages with
longer job tenure (Holzer 1990). The turnover effects frequently are hard to
determine because few companies evaluate their recruiting programs well
enough to show that higher wages did in fact allow them to choose superior
applicants.
Wage-deferral
Scholars who advance the wage-deferral model argue that, in order to
invest in human capital, firms need to obtain long-term commitments from
their workers. Firms under-invest in employee training because of the turnover
threat. Requiring workers to share in the firm-specific investment in human
capital is a way of receiving this commitment. Such a sharing arrangement is
achieved, for example, by having workers earn below-market wages during the
early years of employment in the firm; during later years they earn above
market wage, reflecting a return on this investment. This is similar in nature to
the use of deferred compensation to encourage lower turnover, as shown later.
Proponents argue that the wage deferral profile can be used to favor older
workers (Ippolito 1991), dissuade workers from shirking (Lazear 1979), or
attract a higher quality of workers (Salop and Salop 1976).
2 In economic language, “shirking” is failure to work to one’s maximum capacity or, conversely
explained, to reduce one’s effort to match one’s own image of his/her value. If someone thinks
he/she is underpaid, then he/she will “shirk” to reduce output accordingly, in reciprocal
fashion.
12
Incentive theory
Incentive theory is related closely to efficiency-wage theories for
motivating higher employee effort. There are several incentive-based theories
among which content and process theories are very relevant. Content theories
focus on what motivates employees. The two most popular content incentive
theories, Maslow’s hierarchy of needs (Maslow 1954) and Herzberg's hygiene
theory (Herzberg 1966), include pay as an important factor in employee
motivation (Milkovich and Newman 1993). In the former, pay supplies a series
of basic needs: e.g., the need to acquire food and shelter. Beyond attending
basic needs, pay also can be associated with other higher needs, such as
recognition and satisfaction at the workplace.
Equalizing differences theory
This theory is based on the thought that low employee monitoring goes
hand in hand with low wages. The theory assumes that employees dislike being
monitored, and therefore closely supervised workers will insist on higher wages
because they need to be compensated for the lack of privacy. The romantic
notion of truck drivers as “highway cowboys” who enjoy a high degree of
independence to a degree supports the assumption of the equalizing differences
theory.
In the context of the trucking industry, the equalizing differences theory
may be linked to the argument behind Pedal to the Metal: The Work Lives of
Truckers (Ouellet 1994), though this link may not be straightforward and may
be ambiguous. In his book, Ouellet argues that truck drivers are a unique
group with specific tastes that are significantly different from the tastes of the
average workforce. Drivers who work for extrinsic value work for the money,
and earn more money in trade for greater supervision and lower status
equipment. Drivers who work for intrinsic value, on the other hand, will trade
substantially lower earnings to get independence. Recent data collected by the
author in cooperation with the Owner Operator Independent Drivers
Association strongly supports this hypothesis, since they are among the lowest
paid truck drivers in the U.S. (Belzer 2006), although this same result may be
attributable to the myth of the “American Dream” (Chinoy 1965) or the need to
“buy” a job, since a substantial fraction of trucking has shifted to
subcontractors across many sectors.
Fair wage theory
This is yet another conception of efficiency wages based on the idea that
“fairness” provides explanations for (a) wage compression, (b) the positive
correlation between industry profits and industry wages, and (c) the inverse
correlation between unemployment and skill. The fundamental hypothesis is
that in industries where it is advantageous to pay some employees highly, it is
considered fair also to pay other employees well and hence the “fair wage/effort
hypothesis” (Akerlof, Rose, and Yellen 1988; Akerlof and Yellen 1990; Milkovich
13
and Newman 1993; Rice, Philips, and McFarlin 1990). In other words, in some
industries and firms, high wages paid to one group must also be paid to
another or tensions may arise due to the perceived inequity. Other theories
incorporating the notion of fairness and similar social norms include the rent-
sharing (Levine 1992) and reciprocal-gift models (Burks 1999; Milgrom and
Roberts 1992). .
Compensation Method
We now move from compensation level to the way workers are
compensated. Compensation methods that deviate from the traditional time
rates and salaries have become more popular. Most of these new compensation
methods attempt to align the employee’s interests with those of the firm. While
performance-based methods have a long history in some areas of
manufacturing, the have become increasingly common in other industries and
particularly in the service sector. Piecework, where pay is related directly to
specific units of output, is a common performance-based pay measure, as is
incentive pay, which provides bonuses for meeting or exceeding a target
output. In the next section we focus on piece rates and time rates and their
implications for individual and firm productivity. We focus on these two
methods of direct compensation because of their prevalence in the trucking
industry.
Direct Compensation
Applied at the individual level, piece rates give individual financial
recognition to more productive or harder-working employees who are thus
encouraged to work more intensively. Because they are tied so closely to
output, piece rates provide incentives for employees to exert themselves to
produce more output and generate firm revenues.
Research on compensation methods and piece rates vis-à-vis time rates
has developed over nearly 40 years (Keselman, Wood, and Hagen 1974). In
most of the work reviewed, individuals receiving pay contingent on performance
were more productive than individuals on a time-pay basis (Fernie and Metcalf
1996; LaMere et al. 1996). For example, in a recent study of tree planters in
British Columbia, workers compensated under piece rates produced more, on
average, than those on time rates. Interestingly, however, the productivity of
piece-rate planters fell with the number of consecutive days worked; a similar
result was obtained in a study of copper miners (Paarsch and Shearer 1997;
Shearer 1996). This result becomes especially important in understanding the
effects of long daily and weekly working hours on the trucking industry, in
terms of both driver productivity and safety.
If piece rates produce higher output, one would think this should be
reflected in higher worker earnings. In a study of over 100,000 employees in
500 firms within two industries, Seiler (1984) examined the effect of piece rates
on employee earnings and the impact of incentives on earning. He observes two
14
incentive effects. First, incentive workers’ earnings are more dispersed (i.e., the
distribution is wider) than identical hourly workers’ earnings. Second, on
average the incentive workers earn 14% more money, controlling for other
factors. This premium is partly a compensation for the greater variation in their
income and partly a result of an incentive-effort effect (Seiler 1984).
Two interesting questions emerge from these results. First, does
contingent pay, or more broadly, do productivity-based incentives, actually
increase productivity (the motivation effect) or do they simply attract the most
productive workers (the sorting or selection effect) because they seek the
opportunity for greater earnings given their current level of human capital
(Blinder 1990; Lazear 1995)? This is similar to the issue raised by
compensation-level affects on workers’ productivity and behavior. Second, the
contingent pay passes part of the earnings risk to workers. Therefore, risk-
averse workers may prefer time-rates, which further strengths the sorting
mechanism described above.
Advocates of the sorting effect argue that piece rates differentially attract
workers who are harder working, or who are more productive, than are those
attracted by hourly rates, ceteris paribus. By eliciting higher effort levels, the
effect of piece rates on earnings produces an “earnings effect.” Piece rates also
affect other non-earnings situations. For example, a break or a visit to the
restroom has a high opportunity cost for the employee working in a piecework
compensation system; for a truck driver, who earns his living only when the
wheels turn, a rest-stop or “pit stop” during the day has a substantial
productivity and hence earnings cost. Therefore, given the choice, people who
are more apt to increase effort intensity and effort duration may choose piece
rate methods, while individuals who value the negative non-earning
consequences more than the positive earnings consequences of piece-rates may
tend to select time-based pay schedules. In a study of agricultural workers,
Rubin and Perloff found that the non-earnings effect captures the change with
age in a worker’s relative taste for piece rate work. For the very young and very
old, the non-earnings effect of age dominates the earnings effect (Rubin and
Perloff 1993). For trucking, with almost all intercity drivers and an increasing
fraction of intracity drivers working on incentive-based pay systems, the
“choice” may be to accept the piece rate system or choose another line of work.
Piece rate compensation is attractive to business because it seemingly
solves the problems associated with adverse selection and moral hazard.3 In
addition, by paying piece rates, the firm allows workers to receive the full value
of their own marginal product, thereby eliminating some of the firm’s a priori
3 In economics, “moral hazard” is the tendency of people to spend more of the money that is
not theirs or the time for which they do not pay. Moral hazard cuts both ways, however. From
the employer’s perspective, shirking is a moral hazard. From the employee’s perspective,
unpaid time is a moral hazard. From the trucking firm’s perspective as well as from the
driver’s perspective, unpaid loading and unloading time and shipper or consignee delays are
moral hazards. Shippers and consignees will waste such time because they do not pay for it.
15
need for information on productivity, thus reducing monitoring costs (or
transferring that cost to the worker). Arguably, these incentives may also
reduce the need for employee monitoring and observation to determine
individual merit or performance pay necessary when using other compensation
systems.
Piece rate compensation, however, can bring some disadvantages. As
indicated above, it introduces a source of randomness into workers’ earnings.
In addition, piece rates alone encourage employees to ignore other valuable
activities. As a result, piece rate workers are tempted to reduce quality to
increase measured quantity and engage in other non-productive activities
(Burawoy 1979). Another commonly cited disadvantage of piece-rate
compensation is the difficulty of observing actual productivity (information and
observation problems), which may lead to shirking behavior in the short term
(Gibbons 1987).
Bloom and Milkovich suggest that adverse selection and moral hazard,
as described above, only tells part of the story of the effects of piece rates. The
problem is one of “principals” and “agents”, where the firm is the principal and
the employee or subcontractor is the agent. That is, firms might act to align the
workers’ interest with their own through the use of payment incentives, but its
effect on agent behavior may be more complex than typically assumed by
agency-based research. The incentives and earnings risk-sharing tradeoff, for
example, might lead to the imposition of “greater uncertainty in the
employment relationships” or other adverse outcomes (Bloom and Milkovich
1995). Other responses to incentive payments may also affect the individual
and organizational climate. We review these in subsequent sections.
A 1991 National Research Council Panel study commissioned by the U.S.
Office of Personnel Management to assess the contemporary research literature
on employee job performance and performance-based pay concluded that
individual incentives (including piece-rates) can have positive effects on
performance, though the context of implementation remains important
(Milkovich et al. 1991). The report cites some negative consequences of
incentive pay, including the neglect of aspects of the job not covered in the
incentives, encouraging gaming or reporting of invalid data, and a potential
clash with group norms (as suggested by Burawoy above). Scholars conclude
that individual incentive plans are inappropriate in the presence of high task
complexity (Brown 1990, 1992) and the focus on quality rather than quantity.
For trucking, of course, the safety risk associated with piecework has been a
long-standing issue.
There is limited literature associating compensation methods and safety
outcomes. Hopkins, as cited in Hofmann, argued that incentive pay was not
the root of unsafe behaviors in several coal mines studied (Hofmann, Jacobs,
and Landy 1995). Instead, the organizational climate fueled unsafe behaviors,
as did the workers’ perceptions of the nature of the job (e.g., being unmanly to
16
be careful and safe) (Hofmann, Jacobs, and Landy 1995); Ouellet alludes to
this paradox in his research on truckers’ culture (Ouellet 1994).
Research on safety in the trucking industry has shown that
compensation level, however, is associated with safety, as drivers will tend to
work exceedingly long hours when compensation is low – contributing to safety
risk – and the ability to earn substantially more than in a comparable hourly-
paid job simply by sweating one’s labor and working more hours will make the
industry attractive to workers who cannot get comparable earnings elsewhere
(Belzer, Rodriguez, and Sedo 2002; Rodriguez et al. 2003; Rodriguez, Targa,
and Belzer 2006).
Deferred Compensation
The lower labor turnover found in large firms relative to smaller firms
has been cited by some as evidence that large firms pay workers above their
opportunity cost (Even and Macpherson 1996a). Large firms, they argue, can
afford efficiency wages. Several studies have disputed this claim by
investigating an alternative possible explanation: size-related differences in the
availability, portability, or generosity of pension plans (Even and MacPherson
1996b). Pensions, as wage-tilts discussed in the previous section, can be a
mechanism for encouraging long-term employment relationships beneficial to
firms. Other mechanisms, such as up-front fees and bonds, are rarely actually
observed, but steep age-earnings profiles and deferred compensation plans are
equivalent to bonding in their effects on behavior. Several scholars argue that
deferred compensation (e.g., pension plans, profit sharing, contribution thrift,
ESOPs) directly substitutes for employee wages (Lazear 1979, 1995; Salop and
Salop 1976). Arvin argues persuasively, however, that in imperfect capital
markets where individuals cannot borrow freely, deferred compensation and
wages are not perfect substitutes (Arvin 1991).
Research in the worker mobility literature finds lower turnover in jobs
covered by defined benefit pensions than in other jobs. Turnover is only about
half as great for workers covered by pension plans as for workers without
pensions, supporting the hypothesis that pensions (which act as deferred
compensation) discourage turnover. This relationship remains consistently
strong even after controlling for other factors such as pay level, union
membership, and tenure (Gustman and Steinmeier 1994). Ippolito found that
pensions increased tenure in the firm, on average, by more than 20 percent
(Ippolito 1991). Lazear argues persuasively that the pension plan’s vesting
provisions affect turnover the most and constitute the real incentive effect
(Lazear 1990). Other research shows that capital loss is the main factor
responsible for lower turnover in jobs covered by pensions, but self-selection
and compensation levels also play an important role. Allen provides direct
evidence that bonding is important for understanding long-term employment
relationships (Allen, Clark, and McDermed 1993).
17
This research on truck driver pay and safety will support these findings,
with the added caveat that few non-union truckload drivers and virtually no
owner-drivers can look forward to pensions. Their current rate of turnover, in
excess of 130% per year, supports this hypothesis as well. In sum, the only
truck drivers with defined pension benefits today work for unionized – generally
Teamster – motor carriers, and those pensions are at risk due to declining
participation rates at a vanishing number of unionized carriers. While one may
argue that deunionization has pushed the argument to the margins, the high
truck driver turnover rate and the alleged “truck driver labor shortage” (Global
Insight Inc. 2005) have helped to exacerbate the safety problems that safety
advocates have articulated.
A self-selection concern similar to the effect of efficiency wages also
occurs with pensions. Employees prone to have lower career mobility (such as
truck drivers) would tend to prefer deferred compensation. The study cited
above found virtually no association between firm size and labor turnover for
workers not covered by a pension (Even and Macpherson 1996a).
Two alternative interpretations are plausible. First, larger firms may tend
to select a method of compensation (Soguel 1995) that actually increases
turnover and crash rates (Brown 1990, 1992). Second, pensions were not
included in the study, so the correlation may be a result of the mere existence
of a pension plan or its vesting characteristics (Lazear 1990, 1995).
Several unresolved questions about deferred compensation remain. First,
the pension loss involved in quitting could be offset by a salary increase. This
means that deferred compensation is relevant in the context of the entire level
of compensation. Some scholars argue, for example, that firms offering deferred
compensation tend to have higher compensation levels overall. For this reason,
perhaps it is not the existence of deferred compensation (which is merely a
compensation method), but its existence in the context of other compensation
and the overall level reached (Gustman and Steinmeier 1993). Second, low
turnover rates have been observed for employees under both defined
contribution and defined benefits plans, which suggests that pension
portability is not an issue but rather this may reflect an unobserved sorting
mechanism that is causing the turnover reduction (Arvin 1991). This may be
an issue in trucking, however, since turnover generally is high in the non-
union TL sector and therefore drivers may be unable to vest and to take
advantage of defined contribution pensions (Belzer 2000). In other words, it
may not be the presence of a pension plan but rather the individual’s
anticipation of a pension (or anticipation of the absence of a vested pension)
that may govern turnover.
Finally, this discussion has assumed that compensation levels and
methods are independent of one another. Chen tested inter-industry wage
differentials across different methods of pay. He argued that his evidence
showed that efficiency wage considerations are less important for piece-rate
wages than for time-rate wages under three efficiency-wage-related models:
18
adverse selection or worker-quality, turnover, and shirking models. In the
main, he concludes that industry wage differentials are less prominent in
piece-rate compensation (Chen 1992). The importance of this finding will be
apparent in subsequent sections.
Other studies assume that compensation method is an exogenous
variable. A limited number of studies viewed compensation method as a firm
policy variable (Brown 1990, 1992; Gustman and Steinmeier 1994). Along
these lines, Brown found lower inter-industry wage differentials among workers
under piece rates than under time rates. Gustman and Steinmeier argue that
wages and pensions (or other forms of deferred compensation) are determined
simultaneously by the firm and therefore single equation models tend to bias
this relationship.
Economic Competition and Work Pressure
Compensation method and level of compensation may both be related to
the general economic pressures associated with competition. The customers of
trucking and other freight transportation operations are the shippers and
receivers (“consignees”) of goods; for passenger transportation, the customers
are those buying the tickets to ride the conveyance. Since deregulation, these
customers increasingly have become the controlling parties in freight and
passenger transportation. Indeed, conventional theory of welfare economics
considers markets efficient when consumer welfare is maximized. Shippers
and consignees effectively act as agents of the consumer, so theoretically our
system is working efficiently.
Problems arise when costs embedded in this competition lie external to
the market. This occurs when regulations governing the assignment of these
costs fail to incorporate all the cost. Deregulation of surface freight
transportation sought to promote innovation and competition but did not deal
either with the externalized environmental or safety costs. Indeed, evidence
suggests that metropolitan sprawl may have been encouraged by deregulation,
as the cost of port drayage dropped so low that shippers and consignees moved
their warehouse operations far away from ports of entry. In Southern
California, for example, commonly drayage trucks haul containers 100 miles
away to the Inland Empire, creating congestion (and the demand for more
highways), pollution, and safety costs that unregulated markets failed to
capture (Belzer and Christopherson 2008; Christopherson and Belzer 2009).
Problems also arise when work pressure created by competition causes
CMV drivers to make mistakes that lead to crashes. In a recent study, Belzer
found that interstate bus companies in the most competitive sectors –
“curbside” bus companies – have more than twice the safety risk than the
national average, and compare even more unfavorably with traditional
established intercity bus companies, both unionized and non-union (Belzer
2010a). Similarly, in a study of the carhaul sector of the trucking industry,
Belzer found that driver safety ratings, measured as driver out-of-service rates
19
and carrier-level analyses of safety management, were significantly better at
unionized carriers than at non-union carriers. Non-union carriers are more
likely to subcontract their work to brokers or owner-operators and pay lower
rates for the same work (Belzer 2010b).
In a study using the Large Truck Crash Causation Study dataset, Belzer
also found that work pressure strongly contributed to the CMV driver’s
likelihood of being assigned responsibility for being the last driver whose action
might have prevented a crash from occurring. Data for the cross-sectional
analysis of the causes of large truck crashes come from the Federal Motor
Carrier Safety Administration’s Large Truck Crash Causation Study (LTCCS).4
The LTCCS collected approximately one thousand truck crashes intensively,
collecting a substantial amount of information. While data were inadequate to
determine crash causation based on compensation, substantial evidence
supports the conclusion that work pressure contributes significantly to truck
crashes. Data were collected from 2005 through 2007 and this study was
completed in 2009. It shows that work pressure helps to predict whether the
truck driver is assigned the “critical reason for the critical event” associated
with the crash. For this study, Belzer consolidated all of the work-pressure
factors identified by the LTCCS data-gathering team into an index, and that
index, along with Aggression Count, Fatigue, Class Years, Safety Bonus, Hours
Driving, and Mileage Pay This Trip (as reported by driver) together predicted
15% of the likelihood that the CMV driver would be identified as the driver
responsible for the critical event that precipitated the crash (Belzer 2009b).
Work pressure, aggression, and fatigue were the factors positively associated
with crash responsibility.
Economic Competition and Subcontracting
Some researchers have focused on the role of subcontracting in
determining safety outcomes. While widespread in many industries,
subcontracting has been used intensively in trucking because the work
traditionally has been difficult to monitor, making subcontracting (like
contingent compensation) a useful way for a principals to structure
relationships with agents that align self-interest and reduce shirking and moral
hazard. It also is rooted in the history of the “teaming” business, since trucking
developed out of horse-drawn wagons, and it made sense for “teamsters” to
own and care for their own teams of horses and their own wagons.
Many scholars have long considered subcontracting a vehicle for labor-
market segmentation that creates a two-tier system of internal and external
labor markets as well as core and periphery labor markets (Doeringer and Piore
1971; Edwards, Reich, and Gordon 1975; Gordon 1972; Gordon, Edwards, and
4 For more on this study, see http://www.fmcsa.dot.gov/facts-research/research-
technology/analysis/ltccs.htm and the report of the National Research Council’s evaluation of
that study (Council et al. 2003), located at
http://trb.org/publications/reports/tccs_sept_2003.pdf.
20
Reich 1982; Osterman 1978; Piore 1973; Reich, Gordon, and Edwards 1973;
Rubery, J. 1984; Sabel 1979; Vietorisz and Harrison 1973). These conceptions
of the labor market, and of subcontracting, commonly find that primary or core
labor market participants, employed by firms, have significantly better
employment and compensation packages, including health and pension
benefits, than the packages of similarly situated subcontractors. Indeed, an
intensive analysis of owner-operator cost-of-operations in trucking recently
showed that owner-drivers who own and drive their own truck and do not
employ other drivers or operate multiple trucks earn approximately $21,000
annually in a combination of net profit and wages, which is about 60% of the
compensation earned by non-union employee drivers (Belzer 2006; Belzer and
Swan 2011). They often do not have health benefits and rarely have pension
plans (Belman and Monaco 2001; Belman, Monaco, and Brooks 2004).
In research conducted in the mid-1990s, and building on research
conducted by other Australian researchers (Williamson, Ann M. et al. 1992),
Mayhew, Quinlan and Ferris showed the relationship between safety and truck
ownership. Identifying problems such as the fragmentation of the industry and
the intense competition facing owner-drivers in Australia, they laid out a
paradigm that explains the health and safety risk posed by economic
conditions in this market, exacerbated by inadequate regulatory controls in
Australian long-haul trucking (Mayhew, Quinlan, and Ferris 1997; Mayhew
and Quinlan 1997). In a survey conducted a decade later, Mayhew and
Quinlan found that the problems facing owner-operators had, if anything,
intensified (Mayhew and Quinlan 2006), with even worse consequences for
subcontractor owner-drivers as well as other highway users. These findings
on the dangers of subcontracting have recently been supported by an
examination of growing safety problems in the subcontractor (“regional”) sector
of the U.S. airline industry (Young 2010) following the Colgan/Continental
Airlines plane crash in Buffalo, New York, in 2009 (see the full NTSB report;
National Transportation Safety Board - U.S. Department of Transportation
2010).
III Driver Compensation and Driver Safety: Evidence from Trucking
Research
This section addresses the empirical evidence linking compensation level
and method to worker safety in the trucking industry. First, we review studies
which focus on the effect of various firm characteristics on trucking safety, but
which do not directly address the role of compensation level and method. Next
we review the studies and papers that have included either compensation level
or method in the study of trucking crashes. We also extend the review to
include those studies that have correlated compensation with behaviors
traditionally associated with high crash rates, such as speeding and violation of
hours-of-service regulations.
21
In perhaps the most comprehensive study of compensation and safety,
Belzer, Rodriguez and Sedo studied the effects of compensation using three
methods: case study, cross sectional, and survey (Belzer, Rodriguez, and Sedo
2002). The authors looked at driver pay rates, driver raises, and retention in
their analysis of J.B. Hunt, using a semiparametric hazard function in an event
history analysis (a variant on survival analysis), finding that at the mean, for
every 10% in truck driver pay rates there was a 40% lower probability of driver
crash on a month-to-month basis (Rodriguez et al. 2003; Rodriguez, Targa, and
Belzer 2006). They also found in a cross-sectional study of more than 100
truckload motor carriers, using a logit model that at the mean, every 10% in
driver compensation was associated with a 9.2% lower carrier crash rate. This
study found that not only was driver pay rate significant, but so were the
number of hours of unpaid labor time per mile, the value health and life
insurance, and safety incentive bonuses (Belzer, Rodriguez, and Sedo 2002).
Safety Studies of the Trucking Industry: Firm-Level Characteristics
A study by the Office of Technology Assessment of the U.S. Congress,
Gearing Up for Safety, charted the complex possible causal paths of large truck
crashes in a comprehensive manner as early as 1988 (Office of Technology
Assessment - U.S. Congress 1988). This study traced the factors in the overall
causal mechanism influencing truck crashes to macro-social factors such as
societal values and market forces, and their impact on macro-structural
features such as government policy and legislation, motor carrier industry
segment goals, and shipping and distribution interests. The authors of this
study saw large-scale social forces and structures influencing two major sets of
micro-structural sources of organizational action. On the one hand, federal and
state agency actions such as regulations, roadway design, inspection and
enforcement had an influence. On the other hand, firm actions related to road
operations, driver selection and training, and vehicle maintenance and
specifications also played a role. Finally, at the level closest to the actual set of
crashes, these researchers focused on factors such as roadway conditions,
traffic conditions, other highway users, driver performance, vehicle
performance, load characteristics, weather and unpredictable situations.
Another causal model also identified management operating practices as a key
element in the crash causation chain (U.S. Department of Transportation and
Clarke 1987).
In both models, driver error, haphazard road conditions or equipment
failure were the immediate determinant of a crash. But Loeb et al. pointed out
that the direct causes of crashes “may have been influenced by a prior
occurrence (for example, insufficient driver training) that may have been
affected by an earlier policy action (for example, regulation on driver
qualifications). Furthermore, societal values or economic considerations may
have prompted adoption of a particular policy” (Loeb, Talley, and Zlatoper
22
1994). There has been increased attention recently to the importance of the
economic conditions facing the trucking industry, and how they can be
manifest in after-inflation declines in freight rates, tightening of schedules to
meet shipper demands, and increased interfirm competition (Belzer 2000;
Hensher, Batellino, and Young 1989; Quinlan 2001; Quinlan and Bohle 2002;
Quinlan, Mayhew, and Johnstone 2006; Quinlan, Wright, and National
Transport Commission 2008). The National Research Council’s Committee for
the Review of the Large Truck Crash Causation Study (LTCCS) conducted by
the U.S. Department of Transportation’s Federal Motor Carrier Safety
Administration (FMCSA) likewise expressed concern that data on many such
factors potentially influencing truck crashes should have been a priority of the
FMCSA (Council et al. 2003) , but FMCSA did nor collect data with which to do
an analysis (Belzer 2009b).
Despite awareness of the complexity of the policy environment and the
stochastic nature of the crash environment, the predominant sets of variables
found in large truck safety research have been driver characteristics and
behavior, load characteristics, vehicle characteristics, and roadway conditions.
Relatively little research attention has addressed motor carrier operations
(such as compensation level and method) and driver selection and training. Yet
both were identified as important in the OTA report (Office of Technology
Assessment - U.S. Congress 1988).
A new literature thus is emerging which seeks to take firm
characteristics such as these into account in modeling trucking safety. This
new literature identifies a number of firm-level characteristics other than the
compensation-related variables reviewed in the next section. These include firm
profitability, specific firm safety practices, fleet ownership, demographics of the
firms’ driver force, firm age, union presence, firm size and industry segment.
Firm profitability
Research suggests firm profitability is one firm characteristic related to
safety of transportation operations. Corsi, Fanara and Roberts found that net
operating income was not a statistically significant predictor of crash rates,
although there was an inverse relationship (Corsi, Fanara Jr., and Roberts
1984). Chow et al. found a suggestive association between a carrier’s financial
condition and its safety performance. They suggested that carriers close to
bankruptcy skimp on maintenance, use older equipment, and use owner-
operators (Chow et al. 1987). Blevins and Chow further studied the
profitability-safety relationship during the post-deregulation era. Using
bivariate analyses, they compared results for bankrupt and non-bankrupt
firms, and found that bankrupt firms did in fact spend less on insurance and
safety, maintenance, and equipment replacement, and also were more likely to
have unsatisfactory compliance ratings, but the results were not statistically
significant (Blevins and Chow 1988). Corsi, Fanara, and Jarrell found operating
ratio (operating expenses divided by operating revenue) as having a statistically
23
significant and positive relationship with crash rates for Class I and II carriers
in 1977 and 1984 (Corsi, Fanara Jr., and Jarrell 1988).
Seeking to improve on these earlier, rather inconclusive studies, Bruning
(1989) found that higher return on investment was associated with lower crash
rates. He used a 1984 database based upon Bureau of Motor Carrier Safety
records of crashes causing at least $2000 in property damage and federal
Financial and Operating Statistics from the Form MCS-50T report of 468 Class
I and II general freight and specialized carriers. Bruning made two linked
assumptions: (1) that managers substitute among various production-related
expenses in order to maximize profits, and (2) that the level of substitution of
such expenses as maintenance and training would be reduced given higher
flows of revenue. Bruning found that for large firms, carrier profitability was
inversely related to the crash rates for all general freight and specialized
carriers. He also found that profitability in preceding periods (measured in
1980 and 1982) explained safety performance in 1984 (Bruning 1989).
Moses and Savage utilized a large dataset of 75,577 federal safety audits
and crash records from the 1986-1991 period, but did not report statistically
significant effects for carrier profitability (Moses and Savage 1994). However, in
an earlier analysis the authors found that carriers identified in safety audits as
unprofitable did indeed have significantly more crashes (Moses and Savage
1992). Their analyses differed in the type of statistical procedure used and the
industry segments examined. They point out the importance of stratifying for or
controlling for firm size and industry segment.
Hunter and Mangum measured carrier financial stability using three
variables: revenue per mile; net debt to equity ratio, and operating ratio (total
annual operating expenses divided by annual gross revenue). They viewed
operating ratio as an indicator of a firm’s long-term profitability (Hunter and
Mangum 1995).
Golbe showed the difficulty of establishing such a relationship in any
industry (Golbe 1986). Golbe’s own cross-sectional study of the airline industry
found no statistically significant relationship between profitability and the
square root of total crashes, although note that the number of firms and
number of crashes is much smaller in the airline industry than in trucking. In
addition, higher levels of federal oversight of maintenance in the airline
industry may result in less between-firm variance in crashes. Most
importantly, however, Golbe concluded that data on firm risk preferences and
the specific cost and demand conditions in the industry are necessary in order
to test the relationship between profitability and safety (Golbe 1986).
Furthermore, Chow has pointed out that short-term profitability is but one
dimension of the financial condition of a firm, and may not reflect the longer-
range strengths or weaknesses of a firm (Chow 1989).
More recently, using driver compensation data from Signpost, motor
carrier crash data from the Motor Carrier Management Information System
24
(MCMIS), and from the US Department of Transportation’s (DOT) Financial and
Operating Systems (F&OS),5 along with the National Motor Carrier Directory,
Rodriguez, Rocha, and Belzer found that small motor carriers (fewer than 100
power units) with low liquidity and a lower share of employee compensation per
dollar of freight revenue, are at significantly greater risk of crash (Rodriguez,
Rocha, and Belzer 2004).
Direct measures of firm profitability are difficult to obtain for those firms
that do not submit financial and operating statistics to the federal government.
However, one proxy measure of firm financial condition is the ratio of sales
volume to power units or sales volume to number of employees, data which are
readily available over a period of several years for firms filing federal financial
and operating statistics as well as for firms of all sizes from Dun and
Bradstreet’s TRINC file.
Specific Firm Safety Practices
While safety best practices have never been established scientifically
(weighting all possible factors across firms over time), certain specific firm
safety practices likely have safety consequences. Oversight of the driver and
oversight of equipment, for example, appears to predict safety performance
(National Transportation Safety Board - U.S. Department of Transportation
1988). Moses and Savage identified as particularly significant several other
safety practices: compliance with requirements to file accident reports; taking
action against drivers involved in preventable crashes; and carrier ability to
explain hours of service rules (Moses and Savage 1994).6 However, such
studies often produce counter-intuitive results. For instance, like Moses and
Savage, Corsi and Fanara and Corsi, Fanara and Roberts also used safety
audit data to study the influence of firm safety practices (Corsi and Fanara Jr.
1989; Corsi, Fanara Jr., and Roberts 1984). They found a significant and
positive relationship between crash rates and carrier spending on maintenance.
They attributed this to another known factor, age of fleet: the older the fleet,
the higher the unavoidable repair expenses. Furthermore, in some of their
models, the authors found that substantial hours of service compliance and
demanding driver qualifications were associated with statistically significant
and higher crash rates. The authors explained this result by arguing that the
evolution of an unsatisfactory crash rate may lead to subsequent and costly
improvements in safety management practices, but that cross-sectional data
may not take into account a time lag in the eventual improvement of the crash
rate. More recently, research by Rodriguez, Rocha and Belzer suggests that
small firms with low liquidity and low driver compensation may have a
significantly higher risk of crash (Rodriguez, Rocha, and Belzer 2004). On the
other hand, these weak and sometimes contradictory results may indicate
5 The F&OS is an invaluable resource for motor carrier analysis that the DOT terminated in
2004.
6 Carrier-reported profitability again was not significant.
25
researchers are looking in the wrong place for safety effects; carrier profitability
may not drive safety.
Fleet Ownership
One important data element for firm-level studies is the proportion of a
firm’s fleet which is represented by company-owned vehicles driven by
company employees, leased vehicles driven by company employees, and
vehicles operated by owner-operators.
For Class I and II firms, Corsi, Fanara and Roberts (Corsi, Fanara Jr.,
and Roberts 1984) and Corsi, Fanara and Jarrell presented findings that
suggested that higher use of owner-operators was significantly related to higher
crash levels (Corsi, Fanara Jr., and Jarrell 1988). Chow also concludes that
higher proportion of owner-operators may negatively affect crash rates (Chow
1989). However, Bruning did not find a significant effect for the natural log of
the number of rented power units with drivers as a ratio of total power units
(Bruning 1989). With recent research showing that owner-drivers earn far less
money than do employee-drivers (Belzer 2006), the problem may not lie with
the use of owner-drivers themselves but rather with their low compensation
and the effects low compensation has on drivers’ pressure to take more work
and work too fast and too long.
Demographics of firm driver force
Individual factors such as driver age, experience, and job tenure can
contribute to both individual-level analysis as well as firm characteristics.
Since length-of-service with the firm is a data element in the MCMIS crash file,
a number of studies have sought to examine its impact. Although one study
sought to portray this as an indicator of firm turnover rates, the raw measure
used showed a significant and inverse relationship between length of firm
tenure and crash rates, with over half of nearly 200,000 DOT crashes involving
drivers with less than a year of tenure with the firm (Feeny 1995). Bruning also
found that drivers with less than one year with a reporting carrier accounted
for more than 50% of crashes in a similarly sized database (Bruning 1989).
Such measures cannot be treated as proxies for firm turnover, even in the
presence of controls for firm growth from year to year, nor may they be utilized
as measures of the minimum experience requirements for firm hiring. Belzer et
al. found that driver tenure is an important individual-level safety predictor
and that driver tenure reduces crash probability, ceterus paribus (Belzer,
Rodriguez, and Sedo 2002; Rodriguez et al. 2003; Rodriguez, Targa, and Belzer
2006).
Firm age
The ready availability of data on firm age suggests the value of the
inclusion of the year the carrier was established (and a calculated variable for
firm age) as a firm-level control variable in fire-level safety research. Such data
26
permit us to distinguish between a firms established before or after
deregulation. Corsi and Fanara found that the year of firm establishment, post-
deregulation, predicted crash rate in a multivariate model (Corsi and Fanara
Jr. 1989). This would suggest that firm experience plays a role in safety as
well, probably because it takes time to develop a safety culture and safety
management practices.
Firm size
Corsi and Fanara’s study of 2,000 safety audits found that, using
multiple regression, firm size correlated negatively to crash rates, with larger
firms having lower rates (Corsi and Fanara Jr. 1988). However, Even and
Mcpherson noted that the relationship between firm size and employee
turnover weakens when accounting for such factors as the nature of pension
coverage (Even and Macpherson 1996a). This finding suggests that research
must carefully assess the possibility of interactions between firm size and other
firm characteristics such as industry segment, union presence, and others.
Mixon and Upadyyaya used agency theory and its moral hazard
mechanism to suggest that managers of large firms with greater separation of
ownership and control are more likely to pursue better labor relations and
improved safety levels. However, the authors recognized that firm size is not
always the best measure of remote ownership (Mixon and Upadhyaya 1995).
An improved design might have compared publicly traded firms and firms
owned by holding companies with privately-held firms. While firm size was a
significant predictor of a proxy for safety (damage expenses), firm size may not
have a linear effect, the authors found.
Industry segment
There has been considerable attention paid to the similarities and
differences which can exist between different sectors of the trucking industry
and to the need to better understand the nature of industry segmentation
(Belzer 1994b, 1994a, 1995b, 1995a, 2000; Blevins and Chow 1988; Burks
1999). Yet despite the work of Moses and Savage, research still has not
distinguished conclusively among differential rates and causes of crashes in
different sectors of the trucking industry. The firm-level factors that can enable
the stratification of findings or a focus on a particular segment include for-hire
or private fleet; load mix (primary commodities hauled); trailer mix (primary
and secondary trailer types); truckload, LTL, or both; and average length of
haul. Such firm characteristics are readily available in industry directories as
well as from other sources.
Research on the effects of competition, discussed above, actually may tell
the story of industry segment differences. Horrace and Keane show that the
most competitive trucking industry sectors – produce, intermodal, and
refrigerated sectors – have the worst safety performance (Horrace and Keane
27
2004; Horrace, Keane, and Braaten 2002). This is consistent with Belzer’s
research, cited above related to the carhaul and intercity bus industries.
Summary
Moses and Savage note that “even among ostensibly similar firms there
may be ‘safe’ firms and ‘not-so-safe’ firms” (Moses and Savage 1994). The
design of the federal SAFESTAT system rested upon a similar assumption in
order to develop a national “safety fitness” program for the nation’s commercial
trucking fleet. The Progressive Compliance Program, a component paired with
SAFESTAT, was designed to identify “’sick’ (i.e. unsafe) carriers and provide
different treatments based on that diagnosis to nurse these ‘sick’ carriers back
to health” (John A. Volpe National Transportation Systems Center 1998).
Despite the advances in research on firm characteristics outlined above, the
definition of a “sick firm” remains unresolved. Furthermore, given the paucity
of longitudinal firm-level research, the question remains: are firms with high
levels of crashes at the present time unsafe or merely “unlucky?” Could a
significant year-to-year random variation in firm crash levels explain purported
trends? Finally, do some firm characteristics have a differential effect across
several years, such as whether a firm purchases a new fleet all at once (and
experiences the effects of fleet aging later) or replaces a portion of the fleet each
year (thus masking the effect of vehicle age and safety features)?
Sound research requires a full examination of firm-level characteristics,
along with the specific compensation level and method effects. We must
combine examination of existing records with prospective research, beginning
with some baseline year, to fully understand this problem.
Empirical Evidence for the Effect of Methods and Level of Compensation
in the Trucking Industry: Driver-Level Research
The unavailability of driver-level demographic data has contributed to
limitations to the empirical research in this area. Researchers, as a result, have
used either survey data gathered separately or have approached private firms
in order to have access to their human resources data. The limitations of both
approaches are readily apparent. Most survey data are not representative of the
population. Truck stop surveys, for example, may cause oversampling of
truckload for-hire carriers, over-the-road drivers, and drivers who use truck
stops for some other reason. In carrier-level findings, the results exclusively
apply to the population of drivers belonging to the firm and it becomes difficult
to make inferences to the truck driver population. Finally, data limitations on
the causes of the crashes observed rarely provide a data element that easily
distinguishes truck-at-fault from truck-not-at-fault crashes.
Despite these limitations, some researchers have studied the effects of
compensation on driver crashes and productivity. In one of the early and
definitive studies, Krass (1993) studied the economic environment of trucking
firms in order to explain truck-at-fault crashes in California from 1976-1987.
28
He used an ordinary least squares econometric model, relying on real wage
rates as an indicator, and found that safety declined after deregulation, and
that this decline was specifically attributable to the lower wage rates in the
industry. The results were highly significant, with an R2 greater than 95%
(Kraas 1993). Deregulation reduced safety outcomes because of structural
changes in the trucking industry attributable to a market failure for trucking
services; lower rates for trucking services did not incorporate higher costs of
increased safety risk and roadside inspections became less effective. Lower
rates earned by carriers probably led carriers to skimp on safety and drivers to
violate hours-of-service regulations at more than double the previous rate.
The reduced effectiveness of roadside inspections is consistent with
results found in subsequent research. This finding is especially consistent
with and helps to explain recent findings by Belzer and by National
Transportation Safety Board (NTSB) investigations that safety in the interstate
and international motorcoach bus industry has become a critical problem for
“curbside” and charter bus operators.7 Part of this problem is due to the
“needle in a haystack” or “whack-a-mole” problem faced by enforcement officers
attempting to use roadside inspections and carrier compliance reviews in an
industry characterized by very small firms with shifting ownership and
management structures—carriers never granted interstate and international
operating authority or “reincarnated” after having been placed out of service by
FMCSA (Belzer 2009a).
Beilock, Capelle and Page studied the effect of various driver-reported
firm characteristics on safety-related behavior of drivers and on firm crashes.
The data set comes from a survey of 1,762 truck drivers in the Florida
peninsula. They viewed speeding as providing an intrinsic pleasure-seeking
ability for some drivers, as well as being a way of maximizing leisure time (given
the predominant per-mile form of payment). The authors found that tight
schedules, high company-demanded productivity, and the incentives of the
per-mile pay method were associated with speeding. The authors also
estimated a logit model with a binary dependent variable indicating if a crash
had occurred in the past n years (hence drivers with less than “n” years of
experience were excluded from the sample). They hypothesized that crash
likelihood would be a function of carrier characteristics, driver characteristics,
and equipment features. They found that miles driven in the 12 months before
a crash and method of compensation (hourly vs. per-mile) were insignificant
(Beilock, Capelle Jr., and Page 1989). However, since firm characteristics were
based on current employer, and crash experience was based on the drivers’
overall experience over the past year, high industry turnover could have
prevented an accurate estimate of these effects.
7 See especially the NTSB investigation of the Victoria, Texas fatal bus crash.
http://dms.ntsb.gov/pubdms/search/projList.cfm?ntsbnum=HWY08mh011. See also their
investigations of the Sherman, Texas bus crash (fatal to seventeen people) and other crash
investigations: http://www.ntsb.gov/investigations/reports_highway.html
29
Another study examined the effects of a multicomponent incentive
system on the performance, safety, and satisfaction of 22 drivers working for a
private carrier. This case study claimed to find that the introduction of
performance-based pay incentives led to sustained productivity increases over
a long period, without accompanying increases in crashes or turnover or
decreases in workers’ satisfaction (LaMere et al. 1996). However, given the
random nature of truck crashes, the small sample may explain the lack of a
statistically significant increase in crashes. Even though the multiple baseline
design creates some econometric problems in attributing causality to the
intervention, the results reported are strong enough to suggest that the
incentive pay was an important factor in increased productivity. All drivers in
the study were paid by the hour and the incentives included a distance-driven
bonus. As a result, the carrier did not pass on earnings risk to drivers by
implementing the incentive pay system. In addition, the study provided very
limited information about driver characteristics (e.g., experience and tenure)
and driver exposure. This information may help to further explain the changes
(or lack thereof) in productivity and crashes.
In 1991, the US General Accounting Office (GAO) published the report
“Freight Trucking: Promising Approach for Predicting Carriers’ Safety Risks.”
The report documented the development of a model system of economic factors
and safety. Even though the GAO models driver quality as a function of
macroeconomic conditions of firms, driver compensation is the underlying
mechanism that makes this hypothesis operative. As firms face economic
hardship, they are unable to pay high compensation levels, and therefore the
quality of their work force decreases (General Accounting Office - U.S. Congress
1991). Similarly, the GAO hypothesizes that in the presence of unfavorable firm
financial conditions, drivers who are paid on a “rate basis ... can work at the
same pace and face income erosion or they can drive harder … to maintain
their incomes” (General Accounting Office - U.S. Congress 1991). The GAO
finds that as pay increases, the odds of engaging in a moving violation
decreases. However, for owner operators the odds of conviction decrease as pay
increases and then increase, forming a U-shaped curve (General Accounting
Office - U.S. Congress 1991).
Elements of GAO’s model were tested empirically using survey data from
the Regular Common Carriers Conference (RCCC) survey. The authors found
that compensation method was not a significant factor in determining the
probability of crash involvement for truck drivers who had experienced a crash
in the past 10 years (Beilock, Capelle Jr., and Page 1989). However, this study
had a selection bias because only drivers who had crashes were included in the
sample, making inferences about the driving population questionable. In a
subsequent study, Beilock found that compensation method (by the load, per
mile, per hour or fixed salary) was not significantly correlated with a driver’s
schedule tightness, but this study did not observe hours of service and speed,
and other factors (Beilock 1995).
30
These studies had significant flaws, however. There was little variation
in method of compensation in the sample (virtually all of the drivers were paid
by the mile), so the lack of significant results would be spurious. Second, a
reasonable assumption in the analysis is that no extended breaks were taken
before the interview because of the location of where the interviews were taking
place (Florida Peninsula, outbound). As a result, only cargo-loading (and not
weather or traffic, or cargo unloading) could actually explain any variations in
the schedules under different methods of pay. Furthermore, pay also can affect
the intensity of driving (speed), an effect not accounted for in this study. Braver
et al. did find that lower per-mile compensation levels were associated with
higher propensity to violate hours of service regulations, but they made no
explicit link to crashes (Braver et al. 1992). Hertz explicitly mentions
compensation method as a probable cause for the hours of service violations
found in her study. Per mile and per load compensation provide drivers “with
direct economic incentives to drive longer hours” (Hertz 1991).
A comprehensive study in Australia concluded that overall earnings had
significant negative influence on the number of driver convictions for moving
violations. The same study found strong evidence suggesting that owner-
operator compensation and company freight rates have a significant negative
influence on the propensity to speed (Hensher et al. 1991). In another
Australian study, using a set of structural equations, Golob and Hensher found
that rates of compensation significantly influence the propensity to speed, take
“stay-awake pills” (amphetamines), and to self-impose schedules; these
endogenous variables all contribute to safety problems for truck drivers (Golob
and Hensher 1994, 1995).
In addition to the violation of hours-of-service regulations, other factors
such as sleepiness, fatigue and speeding play an important role in driver
crashes. For example, a report on the causes and effects of sleepiness and
fatigue for motor carrier drivers in New York State concluded that pay method
was associated with driving more than 10 consecutive hours and taking fewer
than 8 hours off-duty (McCartt, Ann T., Hammer, and Fuller 1997a).8 Hensher
found strong evidence suggesting that owner operator compensation and
company freight rates have a significant influence on the propensity to speed.
The authors contend that “he negative relationship is stronger for owner
drivers as might be expected” (Hensher et al. 1991).
Besides being an important crash risk factor, speeding also correlates
with crash severity (Wasielewski 1984). Beilock suggested truck drivers speed
because of (a) pleasure or thrill, (b) they overestimate their abilities, and (c)
because of economic pressures, though without empirical evidence the
“pleasure” hypothesis remains conjectural. Assuming individuals are risk
averse, or at least risk neutral, there should be some payoff from increasing the
level of crash risk (Golob and Hensher 1995) associated with speeding (Beilock,
8 No multivariate analysis was included in the paper. It is unclear if the association found
between pay method and violations would hold after controlling for other relevant factors.
31
Capelle Jr., and Page 1989). Finally, research shows that overall earnings also
have a negative influence on average speeds (Hensher et al. 1991).
Other Issues in the Relationship between Driver Compensation and
Safety
Piece-rate compensation is a common form of performance-based pay
widely used in trucking. However, incentive mechanisms go well beyond piece
rates. Many firms have readily identified this and now offer pay bonuses for
maintaining a satisfactory safety record, having low fuel consumption, and
other characteristics of interest. It therefore is important to stress that the
incentive literature is replete with papers documenting varying degrees of
effectiveness of safety pay bonuses.
Wilde, considered to be the author of the risk homeostasis theory (a
fundamental concept in risk behavior analysis), has studied safety incentives
for the trucking industry (Wilde 1995). He claims that safety incentives are
“generally more effective than engineering improvement, personnel selection,
and other types of intervention, including disciplinary action”. This theory
would suggest that individual compensation tied to specific safety outcomes
might be the key to reducing crashes. His study provides solid evidence of the
success of safety incentives in other industries (mostly manufacturing), though
many of the studies assessing the effectiveness of safety incentives tend to
suffer from the econometric complications stemming from the longitudinal
character of the data. The author explicitly states, however, that he knows of
no controlled experiments addressing the safety and incentives issue (Wilde
1995).
Another study found a significant relation between the introduction of
safety incentives (e.g., surcharge and rebate system due to crash frequency)
and the reduction in the number of crashes (Kotz and Schaefer 1993). It is
unclear, however, if these differences observed are due to changes in manager
or worker behavior. Furthermore, there are other methodological questions of
concern (e.g., omitted variables correlated with predictors and the panel nature
of the data).
Besides the fundamental need to determine more precisely the
association between driver pay and driver safety, we have identified three areas
related to driver compensation and driver safety that warrant further detailed
study: (a) the interaction between compensation method and level, (b) the role
of pensions, and (c) the role of internal labor markets.
Regarding the interaction between compensation method and level, we
presented research suggesting that piece rates shift earnings risks to drivers.
Said differently, piece rates provide drivers with some degree of autonomy to
determine effort and intensity levels. It is reasonable to expect, therefore, that
the intensity and effort incentives afforded by piece rates vary according to the
different piece rate levels. For example, a driver paid low piece rates may have a
higher incentive to speed than a driver paid high piece rates. In order to reach
32
an earnings target, the driver on low piece rates might find it necessary to drive
more miles overall. In fact, some researchers have recently argued that workers
do exhibit a target level of earnings; as a result, workers earning below the
earnings target gain more satisfaction from additional pay than do those
earning above the target level (Drakopoulos and Theodossiou 1998). Incentives
may have a similarly varying effect at different piece-rate levels.
In contrast, the effects of incentives afforded by time rates are harder to
determine. On the one hand, a driver can speed in order to complete a task and
have more leisure time (or work more and earn extra pay). On the other hand,
a driver can drive or work slower than normal (i.e., shirk) and make extra
hourly pay, even though his time-on-task is monitored frequently. We have
found no other research about the potential interaction between compensation
method and compensation level.
Only Southern et al., in their survey of personnel managers, included
pensions as a compensation category. They find that vacation time and sick
time, pension fund contributions, and safety bonuses were not ranked as high
as pay as the most important factor in drivers’ choice of motor carriers for
employment (Southern, Rakowski, and Godwin 1989). A model that departs
from using only the traditional piece or hourly rate and includes pensions and
other bonuses may therefore be useful in painting a more accurate picture of
overall truck driver compensation levels. We found no other study in the
trucking industry that included the role of pensions on worker mobility and
worker satisfaction.
Internal labor markets are difficult to proxy with these data except by
looking at pay raises and retention as proxies for career ladders. Since drivers’
occupations are on the surface (and at our level of data analysis) homogeneous,
we are limited to this approach to internal labor markets.
Indirect Links between Driver Compensation and Driver Safety
Does the literature look at potential indirect effects? An examination of
available research shows sorting and effort-eliciting incentives for different
levels and methods of compensation. For example, through sorting, higher
compensation levels would attract a more qualified labor pool, which, in turn,
will exhibit safe behavior. Figure 1 shows the paths of direct and indirect
effects of compensation method and level on safety. This section evaluates
mediating variables that have been associated with both compensation and
safety for truck drivers, such as age, job satisfaction, turnover, and propensity
to engage in risky behavior (e.g., drive long hours, use illegal substances, and
speed), among others. These indirect links appear as dotted lines in Figure 1.
33
Figure 1. Direct and Indirect Effects — Compensation Method and Level
Indirect Effects, Compensation Level and Method
An important mediating variable is the link that exists between
compensation level and both job satisfaction and organizational commitment.
Previous research suggests that level of pay affects attitudes and perceptions
that affect behavior, including the propensity to have crashes. Results of a
controlled experiment suggest that neither the payment system nor the
incentive level directly affect pay satisfaction beyond their impacts on absolute
level of pay (Berger and Schwab 1980). As expected, other researchers have
established a link between job satisfaction (i.e., satisfaction with the employer)
and driver turnover (Richard, LeMay, and Taylor 1995).
Some researchers have found important differences in job satisfaction
between and within the truckload and the less-than-truckload segments of the
industry. Researchers divided TL drivers into short haul and long haul
occupations, and the differences reported correspond to the different job
characteristics. For example, long haul truckload drivers reported more
negative attitudes concerning issues such as benefits, income, and
advancement opportunities than did short haul drivers (McElroy et al. 1993).
Such results support other research showing substantial pay differentials
between regional and long-haul drivers; long-haul TL drivers are among the
lowest-paid U.S. workers (Belzer 2000). This might also be further evidence of
the importance of career ladders in some segments of the trucking industry, as
discussed previously.
Employee turnover becomes an issue because of low job satisfaction, but
it also is instrumental in determining the sorting effects caused by variations in
compensation levels. In fact, the sorting effect of efficiency wages or wage tilting
may be an indirect path that could result in increased safety. Some researchers
have found evidence that firms’ wage levels are associated positively with the
previous experience of new hires, the tenure of employees with the firm,
managers’ perceptions of employee productivity, and managers’ perceptions of
the ease of hiring qualified workers. Wage levels were negatively associated
with job vacancy rates and training time (Holzer 1990).
Compensation
Method and
Level
Mediating
Variables
Mediating
Variables
Safety
Outcomes
34
In a meta-analytic study, Cotton and Tuttle found that higher pay and
some socio-demographic variables were associated with lower turnover
likelihood. Demographic variables include age, tenure and number of
dependents (Cotton and Tuttle 1986). This finding is important because a
firm’s compensation policies might attract certain types of individuals who
might be more or less prone to quitting the job early. Cotton and Tuttle’s review
notes that 4 out of 5 papers assessing the link between individual performance
and turnover found that the relationship was negative and significant. LeMay et
al. found similar results in a truck driver study (LeMay, Taylor, and Turner
1993). In another trucking study, the driver’s sense of trust in the company
predicted actual turnover best. In the same study, trust, optimism and job
satisfaction had weak relationships with employee attitudes (Kalnbach and
Lantz 1997). Studies in other industries have shown that those who perceive
their jobs as stressful and those who have limited family responsibilities for
children tend to be prime candidates for turnover (Keller 1984).
Similar analyses have shown similar results for compensation method.
For example, one study used an experimental design to measure the
differences in employee satisfaction with pay for workers under time rates
compared with those under incentive payment systems. Results indicated that
neither the payment system nor incentive levels directly affect pay satisfaction
beyond their impacts on absolute level of pay (Berger and Schwab 1980).
The likelihood of using illegal drugs on the job also is an indirect effect of
compensation level. In the single study of this type for truck drivers, Hensher
et al. found that the pay level for owner operators is negatively associated with
the propensity to use illegal drugs. The higher the pay the less likely the owner
operator will use performance-enhancing drugs, particularly amphetamines
(Hensher et al. 1991; Hensher, Daniels, and Battellino 1992).
Indirect Effects, Driver Safety
If driver compensation influences the age distribution of the driver pool,
and the age of drivers correlates strongly with safe or unsafe behavior, then one
could argue that driver compensation and safety are linked via an age-
mediating variable. We describe in this section the “intermediate factors,” such
as age and tenure, and their association with driver safety.
Age
Considerable literature exists that links driver age with crash rates. For
example, younger and less experienced drivers have higher crash involvement.
The fatal crash involvement rates for drivers of large trucks decrease with
increasing driver age (National Highway Traffic Safety Administration - U.S.
Department of Transportation 1982). Younger drivers have six times the
frequency of crash involvement in comparison to the overall driver involvement
rate (Campbell 1991). In addition, research has shown that young truck
drivers, compared with older drivers, have significantly more traffic violations,
35
with a higher proportion of unsafe speed, reckless or careless driving, and
failure-to-yield violations (Blower 1996). In addition, Braver et al. found that
being a violator of hours-of-service regulations was significantly associated with
being a young driver, having a tendency to speed or drive longer when given
unrealistic schedules, and not knowing the hours-of-service rules (Braver et al.
1992).
Work experience
Research attempting to distinguish between age and experience has not
been very convincing. With respect to employee safety, worker experience
shows the same effect as the driver age variable, probably due to the high
collinearity between the two (Bloom and Milkovich 1995). Ayres attempts to
distinguish between the two concepts econometrically, and concludes that
experience and age make separate significant contributions to injury risk with
age as the most important predictor and experience the second most important
out of ten factors identified. Surprisingly, when both factors are in the same
equation the presence of each factor enhances the predictive power, but age
takes on a negative sign. Ayres explains this by claiming that this picks up a
tendency for more experienced drivers to acquire an “optimism bias” that, since
it is unwarranted, makes the driver feel overconfident and increases risk (Ayres
1996). While this may be true, econometric problems suggest this hypothesis
requires considerable more validation. Clearly, age and experience alone have a
positive affect on safety and incorrect statistical specification may have
introduced this paradoxical outcome. However, Lin, Jovanis and Yang studied
the experience of one large interstate carrier and found that while driving time
on the trip prior to a crash was the strongest predictor of a crash, drivers with
more than 10 years of experience had the lowest crash risk, although the
relationship was not linear between one and ten years of experience (Lin,
Jovanis, and Ynag 1993).
Fatigue
Despite its intuitive appeal, literature has shown no conclusive empirical
evidence linking driver compensation method and the onset of fatigue. Clearly,
more research is necessary in this area. An NTSB study of the factors that
affect fatigue in heavy truck crashes did observe pay structure (but not pay
level) as a variable affecting the onset of fatigue (National Transportation Safety
Board - U.S. Department of Transportation 1995). However, the aim of the
study was to examine the factors that affect driver fatigue, and not the
statistical incidence of it. This study introduced definite statistical biases
because it observed single-vehicle heavy truck crashes in which the driver
survived, and thus overestimated the incidence of fatigue substantially.
Nevertheless, it is safe to say that the report “raises questions about the
influence of pay policies on truck driver fatigue … and raises questions about a
link between method of compensation and fatigue-related accidents” (National
Transportation Safety Board - U.S. Department of Transportation 1995).
36
Hensher’s study in Australia tested the hypothesis linking driver fatigue
to the underlying economic conditions in the long distance trucking industry.
However, the experimental design did not allow the observation of fatigue per
se. Rather, Hensher assumed fatigue could not be observable directly. Instead,
Hensher used proxies for fatigue, such as the number of moving violation
convictions and number of crashes (Hensher et al. 1991), and questions
remain whether such proxies embody the phenomenon of interest. Even within
the industry, differences remain between drivers’ and companies’ perceptions
regarding the causes of fatigue, and strategies that should be used to manage
it (Arnold, Pauline K. and Hartley 1997; Arnold, Pauline K. et al. 1997).
The link between fatigue and driver safety, however, seems to be more
robust (Saccomanno, Frank F, Yu, and Shortreed 1995; Arnold, Pauline K. and
Hartley 1997; Chatterjee et al. 1994; Feyer et al. 1993; Golob and Hensher
1995; Wylie et al. 1996). Studies have shown increases in driving errors and
decreases in driver alertness due to fatigue (National Highway Traffic Safety
Administration 1982). A preliminary statistical link is established between
truck driver fatigue and crash rates, as a contributing factor (Saccomanno,
Frank F, Yu, and Shortreed 1995). Despite experimental design limitations, an
NTSB study found that fatigue and fatigue-drug interactions were involved in
more fatalities than alcohol and drug abuse alone (National Transportation
Safety Board - U.S. Department of Transportation 1990).
Turnover
High labor turnover rates have been linked to crash rates. For example,
the Bureau of Labor Statistics found that workers were approximately three
times more likely to be injured during the first month of employment than
during their ninth month of employment. In addition, it found that workers
under 25 years of age were 10 to 20 times more likely to sustain work injury
than older workers (Bureau of Labor Statistics - U.S. Department of Labor
1982). Several studies in the trucking industry have found a consistent positive
correlation between turnover and crash rates (Corsi and Fanara Jr. 1988;
LeMay, Taylor, and Turner 1993; Taylor and J & H Marsh & McLennan 1997).
The implications of these studies for future research on driver compensation
are important. Again, a correlation between driver turnover and accident rates
(at the firm level) is established, though the causal mechanisms remain
unclear. This correlation may be spurious, due to driver age, for example.
Younger drivers change jobs more frequently and have higher accident rates,
therefore accounting for the correlation.
In other firm-level studies, high turnover rates have been positively
correlated with injury rates and injury costs (Rinefort Jr. and Van Fleet 1998).
Again, in most instances these associations tell little about causation, though
plausible mechanisms outlining causality between turnover and crashes can be
devised easily.
37
Safety Climate
The safety culture of an organization is considered a subset of
organizational climate such as work practices, work style, training and
industrial hygiene. A poor safety climate is considered an antecedent of safety
outcomes such as crashes and unsafe behaviors. In a recent study of the
relationship between culture, turnover and driver safety, Taylor and McLennan
find a statistically significant correlation between intent-to-quit and the safety
culture of the organization (Taylor and J & H Marsh & McLennan 1997).9
Another study found a high correlation between traditional safety indices, such
as lost time and crash rates, and safety climate (Coyle, Sleeman, and Adams
1995).
At the individual level, driver stress affects performance significantly
(Matthews 1996), as does work pressure (Belzer 2009b). As with fatigue,
however, there appears to be no conclusive evidence linking compensation with
either safety culture or stress. It is intuitive to think that the performance
pressures induced by piece-rate systems, for example, have an effect on the
individual’s perception of stress and an organization’s safety climate. It may be
likely that a sorting mechanism underlies these phenomena. It may be simply
that data are lacking to test one way or the other. Individuals more able to
handle the stress of piece rate compensation schemes may opt for them while
others would find jobs that have different compensation systems (Rubin and
Perloff 1993), but the fact that the pay system for virtually every over-the-road
trucking job is piece-rate (either by the mile or a percentage of revenue) means
that few alternatives exist for those with the truck-driver skill set, and testing
for significant differences in the real world is almost impossible. Research does
link work stress with turnover (Keller 1984) and it is not difficult to imagine
that wage systems in trucking (including piece-work rates such as mileage pay
or percentage pay, or no explicit pay at all for non-driving time) would be
associated with work stress.
Driver Safety and Driver Crashes
Asalor et al. identify five primary root causes of crashes at the level of an
individual (Asalor, Onibere, and Ovuworie 1994):
1. environmental (e.g., the road and its surroundings);
2. vehicle (e.g., equipment failure);
3. driver;
4. pedestrian and other non-motorized users; and
5. “pure circumstance.”
Pure circumstance consists of being on the road at the wrong time and,
say, being struck by a passing vehicle. This is different from pure randomness,
9 See also TRB safety synthesis on the role of safety culture (Short et al. 2007).
38
however. If crash involvement for any given driver is purely random or
circumstantial, however, then crash involvement should not be an issue when
studying driver compensation policies. In fact, observing crash data that
contains a strong “pure circumstance” component to it introduces a standard
error bias.
Pure circumstance is a subset of pure randomness. Someone can get into
a crash for a number of reasons, such as environmental, vehicle and driver
factors. There is randomness in all of these. The fact that a driver’s tire blew
out because of a nail or the fact that he or she encountered black ice in his or
her lane has some randomness to it. Included in that randomness is “pure
circumstance” – the fact that the driver was at the wrong place at the wrong
time. A specific instance of pure circumstance comes from the fact that other
vehicles can hit you. Speaking personally, even though I did not encounter
black ice in my lane but my neighbor did, this occurrence resulted in a crash
between both of us. If pure circumstance is an important factor in crashes,
then observing multi-vehicle crashes may not be as efficient as observing
single-vehicle crashes for detecting the causes of the crash. This is because in
multi-vehicle crashes, some of the crashes are due to the pure circumstance of
being next to a vehicle that crashed into you. Instead, single vehicle crashes
will exhibit less (but still some) pure circumstance crashes than multi-vehicle
crashes, and as such there is less noise impeding the extracting of the causal
factors in single vehicle crashes. However, an individual driver’s ability to
avoid “pure circumstances” in which crashes occur – his ability to avoid risky
situations in which his vehicle is more likely to be struck by another vehicle or
an incautious driver – probably is a measure of his ability to drive more safely
in the same traffic pattern as others who have higher crash probabilities.
Pure circumstance must not exist in single vehicle crashes, except
insofar as an object falls from the sky and strikes the vehicle. A vehicle in a
multi-vehicle crash, however, may be there due to pure circumstance or due to
any of the first four categories listed. If pure circumstance were a factor, then
single vehicle crashes would be significantly different from multi-vehicle
crashes. The implication for future research is that additional information
about the crash (i.e., number of vehicles involved) might be desirable in order
to improve the explanatory and predictive power of the models.
Arguably some degree of human capital or incentive difference explains
these drivers’ safety records. Indeed, the studies by Belzer et al. all show that
individual characteristics of drivers associated with their compensation rates
predict greater propensity to avoid risk and thus greater safety on the job
(Belzer, Rodriguez, and Sedo 2002; Rodriguez et al. 2003; Rodriguez, Targa,
and Belzer 2006).
In addition to the use of subsets of crashes at the individual level,
researchers have used moving violation convictions as proxies for driver safety
behavior. The stochastic nature of crashes highlights the difficulty in predicting
them. As a result, researchers have consistently used driving convictions as
39
variables that are less vulnerable to randomness (Beilock, Capelle Jr., and Page
1989; Peck, McBirde, and Coppin 1971). Most researchers have found that
they generally can use moving violations to predict future crashes. These
results lead to the conclusion that drivers exhibit bad behavior, as measured
by moving violations, consistently over time (Ferreira 1972; Mitter and Vilardo
1984). This conclusion does not support the common belief that we can model
poor driver behavior as random walk (Poisson distribution or Poisson-related
model). The relevant variables probably have some of the same behavioral
elements involved in moving violations and are more stable and sensitive
measures of individual differences of driver behavior. Miller and Schuster,
however, found a positive relationship between previous violations and future
(or current) moving violation convictions but not with crashes (Miller and
Schuster 1983). Arguably “there is sufficient initial evidence to examine the
issue further, together with the relationship between employee status and
crashes” (Pearson and Ogden 1991).
Market Factors
In his extensive report on truck crash causation, Quinlan concludes that
Australia’s truck safety problems stem from competitive industry forces, and
particularly on pressures created by shippers who demand rapid and timely
service for a low price. This has created a “sweatshop” sort of environment in
Australia that is responsible for an alarming truck safety problem, including
long hours, high levels of chronic fatigue, and amphetamine abuse.
Regulations aimed at individual drivers are relatively ineffective because they
do not address underlying economic performance pressures on the industry.
Self-regulation in the absence of a market model, while laudable, also does not
work because it does not address the problems created by competitive market
forces. His inquiry recommends the establishment of an industry-wide “Code of
Practice” which would include coordination among regulatory agencies,
compulsory licensing of all participants in the logistics industry, the
replacement of logbooks with “Safe Driving Plans” signed and filed by motor
carriers and drivers, and minimum pay and conditions standards for all drivers
- a “safety rate” applicable to both employee and owner-operator drivers and
carriers (Quinlan 2001). Quinlan’s concept of a safety rate also has become
accepted as a matter of national policy in Australia (Quinlan, Wright, and
National Transport Commission 2008; Skulley 2009), although implementation
remains unclear and bogged down in process.
In his elaborate Trucking Industry Benchmarking Program, Belzer uses
cost-effective on-line data collection methods in an effort to collect data on both
direct and indirect operational factors, with which he hopes to predict motor
carrier safety. Based on the premise that cutthroat carriers cut corners to
attract business by having low operating costs, and assuming that this corner
cutting behavior includes practices that likely put the carrier at risk, Belzer
proposes to determine the extent to which marginal pricing in trucking, in the
absence of effective financial responsibility laws, might cause large and safe
40
carriers to subsidize unsafe carriers against their will, thereby creating a
market externality imposed on those carriers and the motoring public.
Economic theory suggests that carriers with few assets may be “damage proof”
because they can insure the value of their investment at a rate far lower than
that which the market would charge if insurance companies were allowed to
charge market rates for motor carrier insurance, representing their estimate of
carries’ true risk. If the cost of one fatal crash averages approximately US$3.5
million and federal regulations only require that carriers maintain $750,000 in
per-crash liability insurance—state laws allow carriers to insure themselves at
a prescribed minimum liability of $1 million or less—then it is quite possible
that this subsidy helps to drive down shipping rates as well as motor carrier
profits and driver pay rates. Belzer argues that self-regulation is possible only if
public policy forbids these subsidies and if motor carriers benchmark their
operational characteristics and practices (including compensation factors)
against each other (see http://www.ilir.umich.edu/TIBP/) as well as
Transportation Research Board presentations at
http://www.ndsu.edu/ndsu/trb/).
Research recently published demonstrates clearly the relationship
between market forces and motor carrier safety. Analyzing data collected from
J.B. Hunt, a large truckload carrier that elected to solve its driver supply
problem by raising wages substantially all at once, Belzer, Rodriguez, and Sedo
show that this carrier cut its turnover rate as well as its overall crash rate in
half in less than one year by paying an efficiency wage (Belzer, Rodriguez, and
Sedo 2002; Rodriguez et al. 2003; Rodriguez, Targa, and Belzer 2006). Indeed,
the firm reduced its monthly rate of major crashes four-fold, for unscheduled
over-the-road freight drivers. A duration model, predicting the probability that
each individual driver will have a crash in each succeeding month10 showed
that at the mean, for each ten percent in base mileage wage, the carrier
reduced the probability of crash for the average driver by 34 percent. In
addition, since some drivers received wage increases during this strategic
change in compensation policy, a ten percent increase in drivers’ base wages
produced a six percent lower probability of crash. Clearly the policy had the
desired effect.
IV. Conclusions
Economics – the competitive forces resulting from markets – strongly
influence the structure of industry as well as the structure of the labor markets
on which industries rely. This is a fundamental driving force in market
economies, where private companies compete for business and by selling goods
and services to customers subject to their preferences. Transportation is a
10 Duration models are a method of conducting survival analysis, appropriate to the particular
variables incorporated within a model. See methods section below for detail and explanations.
41
commodity within these markets because one unit of transportation is the
same as the next, subject to quality constraints with imperfect information.
Transportation service failures take the form of delays due to many
factors, including weather and equipment breakdown. Catastrophic failures, in
the form of vehicular crashes, are low probability high impact events the
predictability of which continues to stump analysts who know how to predict
crashes based on mechanical failures or precautionary failures, including
human error. The probabilistic prediction of commercial transportation
failures, however, has eluded analysts who continue to restrict their focus to
the equipment or human factors without looking systematically at the
economic environment in which the commercial activity – and the failure –
occurs.
Most critically, analysts fail to take into consideration that unlike
personal travel, commercial vehicle transportation – whether by marine, air,
rail, or highway – constitutes a derived demand industry that responds to the
laws of economics as surely as it responds to law and regulation. In other
words, while truck drivers respond to laws and regulations governing their
operations, such laws vary by time and place, while economic laws do not vary.
Truckers and trucking companies respond to the market demands of their
customers, and compete based on satisfying those customers’ demands for
both price and service. Governments regulate truckers’ equipment, practices
and behaviors to create boundaries around that competition, but they can do
so imperfectly. Since markets are systems of reciprocal demands set in a social
context, the context itself requires systematic regulation that acknowledges the
markets that frame the system. In other words, we must embed systematic
truck safety regulation in the context of market systems.
Trucking is a labor-intensive industry, so we cannot effectively regulate
trucking industry safety without addressing the fact that truck driver
compensation is a major factor underlying the price of service that underlies
this market. If freight transportation is a derived demand industry and if price
and service are the dominant factors motivating competitive carriers, then we
must deal with compensation factors if we are going to have any effect on
motor carrier safety.
These studies show that higher driver pay is associated with safer
operations. Clearly the more drivers are paid, and the more they are paid for
their non-driving time, the less likely they are to have crashes. Part of this
effect is due to labor market sorting: carriers that pay more money can afford
to be more choosy, which allows them to select drivers with superior
unobserved (to us) human capital characteristics. Part of this effect also is due
to incentives: drivers who earn more money are motivated to protect their
records and, if they have them, their retirement plans. Carriers that pay
drivers more money do so because the value of their service is higher to the
customer, and the generally higher value is associated with greater service
demands and necessarily higher value of the freight.
42
These studies also show that market competition, an extremely powerful
force in a world of unregulated economic competition, has put supply chain
power in the hands of the shippers and consignees who determine rates and
conditions under which freight services are allocated. The development of the
supply chain approach to freight transport has placed the consumer in the
most powerful position, as the consumer drives transactions in a world
governed by welfare economics. The shippers and consignees are the
consumers within the supply chain and represent end consumers.
The question is whether all the costs of transport are incorporated within
the supply chain. Does the market governing supply chain externalize costs to
society, creating inefficient market signals within supply chain transactions?
Evidence presented here suggests that not only does the market incent
inefficient use of freight transport resources, creating sprawl and
environmental consequences, but it incents safety and health consequences
the cost of which are borne by commercial motor vehicle drivers as well as the
motoring public. These consequences represent a market failure that calls for
regulatory solutions designed to incorporate all costs and benefits into an
efficient market. An efficient market can therefore not only increase
macroeconomic efficiency but spin off the equity that is the promise of the
utilitarian ideal.
V. Policy Implications
Engage the U.S. Department of Labor as well as FMCSA
o Get government regulators out of their silos. FMCSA and the Department of
Labor should cooperate to regulate the economic conditions that lead to
safety problems. The DOL has the authority to regulate compensation and
should do so.
The FMCSA should not have sole responsibility for CMV safety. While
safety regulation is an important DOT function, safety is everybody’s business.
Once we recognize that safety problems have economic origins, and that these
economic origins stem substantially from the effects of competition on the labor
market, it becomes apparent that the Department of Labor needs to share
responsibility. The silos of the Federal Government do not help to solve
problems when they create artificial barriers for public policy.
FMCSA believes it does not have the authority to regulate compensation,
even though it has commissioned research showing that competitive forces,
including compensation and industry segment (a proxy for the price carriers
charge to cargo owners, which eventually leads to driver compensation levels),
play a major role in safety performance. The Department of Labor likewise
believes it must take a hands-off attitude toward trucking, which originally was
regulated by a Congressional agency – the Interstate Commerce Commission –
43
that has not existed for more than fifteen years. This analysis shows that we
will not make lasting progress in safety without reconciling this turf question.
Regulations enforcing the FLSA should require explicit pay for implicit as
well as explicit work. While it’s fine to say that drivers must at least earn the
minimum wage, many earn less than the minimum wage for all time employed,
and most earn nothing explicitly for the hours they spend doing non-driving
labor. Research cited here suggests that the average intercity driver probably
works about 25% more hours than he logs, because he simply does not log
unpaid non-driving labor time, and surveys show that on average 25% of
drivers’ work time involves non-driving labor. If carriers and cargo owners had
to pay drivers for all of their time, the amount of time spent in doing non-
driving labor would decline accordingly; cargo owners would no longer benefit
from the moral hazard of playing with somebody else’s time – or money. This
moral hazard causes economic deadweight loss for society, as cargo owners
and their agents demand more freight services – including service that they
value at a very low rate – than the market would bear absent this moral
hazard.
Carriers must charge, and cargo owners must pay, for all services they
receive. It should be illegal to decline to collect such fees, or to refuse to pay
documented charges. These fees include various “ancillary” charges such as
waiting time (waiting to get loaded or unloaded), inside delivery, stacking and
restacking freight inside food warehouses, and “demurrage” (excessive delay
time). Shippers can order a truck early because they have the leverage to
require it and receivers can refuse to unload a truck when it arrives because
they aren’t ready for the freight (or because the driver missed the time window).
This causes drivers to engage in risky behavior to make appointments and they
will not log unpaid time, extending their workday and workweek by working “off
the clock”, again demonstrating the interaction between competitive forces and
safety and health risk.
In sum, while “safety culture” of the firm is something that FMCSA can
address, and it can issue regulations on equipment and driver training,
behavior, and qualifications, if economic forces require that safety culture be
superimposed on a no-holds-barred competitive environment, the regulator will
be fighting a continuous rear-guard battle against the iron law of competition.
If the fundamental exigencies of markets work at all, then cargo owners’ need
for lowest price will lead to a race to the bottom and safety will suffer. Because
economic forces are involved, economic solutions must be considered.
Implement chain of responsibility regulations
o Implement Chain of Responsibility regulations like those enacted by the
Australian Parliament to create a level playing field in a deregulated
environment.
44
Mitigation of the negative effects of competition requires that everyone in
the supply chain – everyone in the chain of custody – take joint responsibility
for safety outcomes. If cargo owners share the responsibility for the safe
transportation of goods and people, they will have an incentive to work together
with brokers and transportation providers to insist on socially responsible
contracting practices, including a willingness to pay reasonable rates for the
service. Following an inquiry on truck safety that determined that economic
forces underlie commercial motor vehicle safety (Quinlan 2001), Australia
implemented a “chain of responsibility” policy, in cooperation with the trucking
industry and all levels of government (2004)11. On the principle that all
participants in the chain of custody need to participate in developing and
implementing a safety culture, government safety officials have cooperated with
the industry to develop a safety accreditation scheme designed to engage the
industry in continuous improvement with respect to safety (Baas and
Taramoeroa 2008).
In Australia the government has gone so far as to announce a “safe
rates” policy setting a minimum compensation package for truckers (Quinlan,
Wright, and National Transport Commission 2008), which was passed the
House on March 12, 2012 and the Senate on March 20, 2012.12 Fair Work
Australia has set up an industrial tribunal that begins work July 1 to establish
a minimum national compensation scale for all truckers. It has widespread
political as well as scientific support.
Carriers, drivers, third-party logistics providers, brokers, and cargo
owners must be responsible for the supply chain in its entirety. The
fragmentation of economic and legal responsibility for freight transport imposes
hidden costs on the transportation system by imposing hidden costs on
society. These costs appear in the form of safety and health burdens absorbed
disproportionately by CMV drivers for whom the excessive work hours and
safety and health burdens impose risks, and for motorists and others on the
public roadway as well as health burdens suffered by the public generally by
excessive low-cost trucking. It leads to widespread subcontracting as well,
which shifts risk burdens to those least able to support them, shifting risk
from the service providers to society, with attendant efficiency losses.
Currently the largest carriers, with the greatest visibility and assets to
protect, tend to be the deep pockets that attract lawsuits. Our legal standards,
which tends to hold parties responsible for damages according to the depth of
their pockets, creates some inefficient incentives. The FMCSA only requires
that carriers carry insurance for up to $750,000 per incident, even though
single incidents can cost millions of dollars, and this unrealistically low level
subsidizes unsafe carriers that can charge rates reflective of their inadequate
11 http://www.ntc.gov.au/newsdetail.aspx?newsid=149;
http://www.ntc.gov.au/viewpage.aspx?documentid=01419 (accessed on July 9, 2012).
12 http://www.ministers.deewr.gov.au/shorten/safer-roads-all-australians (accessed on July 9,
2012). To locate the full Hansard, search http://www.aph.gov.au/hansard.
45
coverage while society bears the cost of this risk. In addition, motor carrier
risk is hard to assess, and though the chance of a major loss is small, the cost
could be great. Because low probability, high impact events are so hard to rate
they can be hard to insure, and these carriers may be able to obtain insurance
from assigned risk pools that, at least in some states, may charge below-
market rates. Large motor carriers, on the other hand, which are substantially
self-insured, pay the full cost of insuring against losses and may pay a
premium over less safe carriers.
Australian policy makers have found that although “chain of
responsibility” is hard to define and implement, it has been an effective way to
get everyone’s attention. In some cases where a willful pattern of violations has
been identified, such as a case in New South Wales involving systematic
overloading of trucks by grain shippers, criminal charges have been made, and
industry-wide compliance occurred quickly.13
Subcontracting
o Tighten regulations on subcontracting that balances the power between
contractors and trucking companies, as Australians have done. Court
rulings 40 years ago usurped legislative authority, disallowing traditional
cooperation among owner-drivers to negotiate with carriers. This would give
owner-drivers a fair shake.
Widespread subcontracting, and arguably misclassification of workers as
contractors in an attempt to evade employment and labor law as well as escape
other burdens of having employees, has undermined public policy relative to
employment and undermined true small business truckers as well.
Independent businesses owners do not have to pay themselves a minimum
wage, much less a living wage, removing the floor from the labor market
entirely. When employees with no bargaining power are classified as business
owners, they make a mockery of small business. As discussed in Belzer and
Swan (2011), an intensive study of owner-drivers showed that the average
owner-driver of one truck in interstate commerce, which he drives himself,
earns only $21,267 in wages and profits combined. Since we know from other
surveys that these drivers work at least 3,000 hours per year, their average
earnings are slightly greater than $7 per hour. Since the median is almost
identical to the mean, half earn less than that. Again, with pay a strong
predictor of safety, economic pressures may account for most of their safety
risk, and their risk as well as the risk to other highway users is substantial.
13 Philip Halton, Assistant Director, National Transport Policy, Licensing, Registration &
Freight. Roads and Traffic Authority of New South Wales. Personal communication and talk at
the University of Michigan Transportation Research Institute conference in June 2009. Haltom
presentation on “Compliance Issues (P10-1187)” also given at Transportation Research Board
Annual Meeting on Sunday, January 10, 2010, in the session: “OECD-JTRC International
Study on Truck Transport Safety, Productivity, and Sustainability: Final Results”.
46
Subcontracting (or worker misclassification) has increased in recent
years, with thousands of workers essentially buying their jobs. They own the
equipment and the risk but motor carriers, under whose authority they
operate, control them just like employees, with many working under conditions
that resemble debt peonage. Many of these subcontract to other drivers who,
though they do not own the equipment they drive, also become subcontractors.
This individualization of work, now also widespread in the construction
industry (especially residential), completely changes the employment dynamic,
making labor and employment law enforcement, including regulations
protecting worker safety and health as well as tax collection, virtually
unenforceable. This creates a dangerous climate for safety and puts both the
drivers and the public at great risk.
While these are just three recommendations that arise from this research
stream, these three changes would have a profound impact on the economics
of safety and health in the U.S. commercial carrier industry. Implications for
trucking are obvious, but the same kinds of reform would result in safer
airlines and commercial motor coach bus industries as well.
47
V. Bibliography
2004. "Chain of Responsibility". Australia: Information Bulletin. Australia
National Transport Commission. February 2004. pp. 7.
Akerlof, George A., Andrew K. Rose, and Janet L. Yellen. 1988. "Job Switching
and Job Satisfaction in the U.S. Labor Market". In Brookings Papers on
Economic Activity. Washington: Brookings Institution.
Akerlof, George A., and Janet L. Yellen. 1988. "Fairness and Unemployment."
American Economic Review, Papers and Proceedings Vol.78, No.2:pp 44-
49.
———. 1990. "The Fair Wage-Effort Hypothesis and Unemployment." Quarterly
Journal of Economics Vol.105:pp 1003-16.
Allen, Steven G., Robert L. Clark, and Ann A. McDermed. 1993. "Pensions,
Bonding, and Lifetime Jobs." Journal of Human Resources Vol.28,
No.3:pp 463-81.
Arnold, Hugh J., and Daniel C Feldman. 1982. "A Multivariate Analysis of the
Determinant of Job Turnover." Journal of Applied Psychology Vol.67,
No.3:pp 350-60.
Arnold, Pauline K., and Laurence R. Hartley. 1997. "Hours of Service for Truck
Drivers: Recent Research on Fatigue and a Proposed Approach to Dealing
with the Problem". Paper read at International Large Truck Safety
Symposium, at Knoxville, TN.
Arnold, Pauline K., Laurence R. Hartley, Angela Corry, David Hochstadt, Fulvio
Penna, and Anne Marie Feyer. 1997. "Hours of Work and Perceptions of
Fatigue Among Truck Drivers." Accident Analysis and Prevention Vol.29,
No.4:pp 471-77.
Arvin, Mak B. 1991. "A Role for Private Pensions in Labour Contracts." Oxford
Economic Papers Vol.43, No.1:pp 99-115.
Asalor, J.O., E.A. Onibere, and G.C. Ovuworie. 1994. "Pure Circumstance in
Road Traffic Accidents." Journal of Safety Research Vol.25, No.1:pp 47-
60.
Ayres, T J. 1996. "Operator Experience and Accident Risk". Paper read at 40th
Human Factors and Ergonomics Society Meeting, at Philadelphia, PA.
Baas, Peter, and Neon Taramoeroa. 2008. "Analysis of the Safety Benefits of
Heavy Vehicle Accreditation Schemes". Sydney. Austroads Incorporated.
Austroads Project No. SS1057. May. pp. v, 52.
Becker, Gary S. 1975. Human Capital. 2nd. ed. New York: Columbia University
Press.
Beilock, Richard. 1995. "Schedule-Induced Hours-of-Service and Speed Limit
Violations Among Tractor Trailer Drivers." Accident Analysis and
Prevention Vol.27, No.1:pp 33-42.
Beilock, Richard, Russell B. Capelle Jr., and Edith Page. 1989. "Speed and
Training Factors Associated with Heavy Truck Accidents." Transportation
Quarterly Vol.43, No.4:pp 571-89.
48
Belman, Dale L., and Kristen A. Monaco. 2001. "The Effects of Deregulation,
De-Unionization, Technology, and Human Capital on the Work and Work
Lives of Truck Drivers." Industrial and Labor Relations Review Vol.54,
No.2A:pp 502-24.
Belman, Dale L., Kristen A. Monaco, and Taggert J. Brooks. 2004. Sailors of
the Concrete Sea: A Portrait of Truck Drivers' Work and Lives: Michigan
State University Press.
Belzer, Michael H. 1994a. "The Motor Carrier Industry: Truckers and
Teamsters Under Siege". In Contemporary Collective Bargaining in the
Private Sector, edited by P. B. Voos. Madison, Wisconsin: Industrial
Relations Research Association.
———. 1994b. "Paying the Toll: Economic Deregulation of the Trucking
Industry". Washington, DC. Economic Policy Institute. pp. 84.
———. 1995a. "Collective Bargaining After Deregulation: Do the Teamsters Still
Count?" Industrial and Labor Relations Review Vol.48, No.4:pp 636-55.
———. 1995b. "The Trucking Industry: Structure, Work Organization,
Productivity, and Labor Market". Washington, DC. Office of Technology
Assessment. July 1, 1995. pp. 1-47.
———. 2000. Sweatshops on Wheels: Winners and Losers in Trucking
Deregulation. Oxford, UK and New York, NY: Oxford University Press.
———. 2006. "OOIDA 2003-2004 Cost of Operations Survey: Report of
Results". Detroit, MI. Wayne State University. April 5, 2006. pp. 50.
———. 2008. Truck Driver Hours of Service, Interim Final Rule; 72 FR 71247,
December 17, 2007 [comments]. Regulations.gov,
http://www.regulations.gov/fdmspublic/custom/jsp/search/searchresu
lt/docketDetail.jsp#.
———. 2009a. "North American Motorcoach Bus Industry". Detroit. Wayne
State University. pp. 68 plus appendices.
———. 2009b. "Report of Analysis: Truck Crashes and Work-Related Factors
Associated with Drivers and Motor Carriers". Sound Science, Inc. April
28. pp. 275; includes appendices.
———. 2010a. "Crisis in the North American Motorcoach Bus Industry: Threats
and Opportunities Created by Growth in Intercity Bus Traffic". Detroit:
Report. Wayne State University. pp. 59; appendices.
———. 2010b. "The Low Road: Fiat/Chrysler’s New Auto Transport Model
Threatens Highway Safety and Economic Recovery". Ann Arbor, MI.
Sound Science, Inc. pp. 14.
———. 2011. "Labor and Human Resources". In Intermodal Freight
Transportation: Moving Freight in a Global Economy, edited by L. Hoel,
G. Giuliano and M. Meyer. Washington: Eno Foundation for
Transportation.
Belzer, Michael H., Kenneth L. Campbell, Stephen V. Burks, Dale Ballou,
George Fulton, Donald Grimes, and Kristen A. Monaco. 1999. "Hours of
Service Impact Assessment". Ann Arbor, MI. University of Michigan
Transportation Research Institute and Federal Highway Administration,
Office of Motor Carriers and Highway Safety. March 5, 1999. pp. 65.
49
Belzer, Michael H., and Susan Christopherson. 2008. "Freight Transportation
and Economic Development: Who Pays? Who Profits?" Urban and
Regional Policy and Its Effects:pp 54.
Belzer, Michael H., George A. Fulton, Donald R. Grimes, Gregory M. Saltzman,
Stanley Sedo, and Lucie G. Schmidt. 2002. "Proposed Changes in Motor
Carrier Hours of Service Regulations: An Assessment". Ann Arbor, MI.
University of Michigan Trucking Industry Program. March 6, 2002. pp. 1-
73.
Belzer, Michael H., Daniel A. Rodriguez, and Stanley A. Sedo. 2002. "Paying for
Safety: An Economic Analysis of the Effect of Compensation on Truck
Driver Safety". Washington, DC. United States Department of
Transportation, Federal Motor Carrier Safety Administration. September
10, 2002. pp.
Belzer, Michael H., and Peter F. Swan. 2011. "Supply Chain Security: Agency
Theory and Port Drayage Drivers." The Economic and Labour Relations
Review Vol.22, No.1:pp 23.
Berger, Chris J. , and Donald P. Schwab. 1980. "Pay Incentives and Pay
Satisfaction." Industrial Relations Vol.19, No.2:pp 206-11.
Berkes, Howard. 2011. Report Blasts Massey For 'Deviance' In Safety Culture.
In Mine Safety In America. United States: National Public Radio.
Blevins, M., and Garland Chow. 1988. "Truck Safety and Financial Distress: A
Preliminary Analysis." Journal of the Transportation Research Forum
Vol.29, No.1:pp 18-23.
Blinder, Alan S. 1990. Paying for Productivity. Washington, D.C.: Brookings
Institution.
Bloom, Matthew C., and George T. Milkovich. 1995. "The Relationship Between
Risk, Performance-Based Pay, and Organizational Performance". Paper
read at Academy of Management Annual Meeting, Business Policy &
Strategy Division.
Blower, Daniel. 1996. "The accident experience of younger truck drivers". Ann
Arbor, MI. University of Michigan Transportation Research
Institute/Great Lakes Center for Trucking and Transit Research. pp. 55.
Braver, Elisa R., C. W. Preusser, D. F. Preusser, H. M. Baum, Richard Beilock,
and R Ulmer. 1992. "Who Violates Work Hour Rules? A Survey of
Tractor-Trailer Drivers". Arlington, VA. Insurance Institute for Highway
Safety. pp. 26.
Brown, Charles. 1990. "Firms' Choice of Method of Pay." Industrial and Labor
Relations Review Vol.43, No.3:pp 165-S.
———. 1992. "Wage Levels and Method of Pay." Rand Journal of Economics
Vol.23, No.3:pp 366-75.
Bruning, Edward R. 1989. "The Relationship Between Profitability and Safety
Performance in Trucking firms." Transportation Journal Vol.28,
No.Spring:pp 40-49.
Burawoy, Michael. 1979. Manufacturing Consent: Changes in the Labor
Process Under Monopoly Capitalism. Chicago: University of Chicago
Press.
50
Bureau of Labor Statistics - U.S. Department of Labor. 1982. "Age and
Experience Count". Washington, DC. Bureau of Labor Statistics. pp.
Burks, Stephen V. 1999. "The Origins of Parallel Segmented Labor and Product
Markets: A Reciprocity-Based Agency Model with an Application to Motor
Freight". Ph.D. dissertation, Economics Department, University of
Massachusetts, Amherst.
Burks, Stephen V., Michael H. Belzer, KWAN Quon, Stephanie G. Pratt, and
Sandra Shackelford. 2010. Trucking 101: An Industry Primer. Edited by
Transportation Research Board Committee on Trucking Industry
Research (AT060), TRB Transportation Research Circular. Washington:
Transportation Research Board.
Cadotte, E. R., H. L. Sink, and Arun Chatterjee. 1997. "Human Factors and
Their Influence on Accidents Involving Long-Haul, Tractor Trailers in the
United States". Paper read at International Large Truck Safety
Symposium, at Knoxville, TN.
Campbell, Kenneth. 1991. "Fatal Accident Involvement Rates by Driver Age for
Large Trucks." Accident Analysis and Prevention Vol.23, No.4:pp 287-95.
Cappelli, Peter, and Keith Chauvin. 1991. "An Interplant Test of the Efficiency
Wage Hypothesis." The Quarterly Journal of Economics Vol.106, No.3:pp
769-87.
Chatterjee, Arun, UNC Institute for Transportation Research and Education,
University of Tennessee Knoxville Transportation Center, University
Transportation Centers Program (U.S.), and Southeastern Transportation
Center (U.S.). 1994. "Driver-related factors involved with truck
accidents". Raleigh, NC. University of North Carolina Institute for
Transportation Research and Education. January. pp. viii, 72 leaves.
Chen, Paul. 1992. "Industry wage differentials, efficiency wages, and method of
pay". Uppsala, Sweden. Uppsala University Department of Economics.
Issue. pp. 27.
Chinoy, Ely. 1965. Automobile Workers and the American Dream. 1, with an
introduction by David Reisman ed. Boston: Beacon Press, by
arrangement with Random House. Original edition, Random House,
1955.
Chow, Garland. 1989. "Deregulation, financial distress and safety in the
general freight trucking industry". In Transportation Safety in an Age of
Deregulation, edited by L. N. Moses and I. Savage. New York, NY: Oxford
University Press.
Chow, Garland, Mark Vasina, Bahram Adrangi, and Richard D. Gritta. 1987.
"The definition and measurement of financial fitness in the trucking
industry." Journal of the Transportation Forum Vol.28, No.1:pp 319-24.
Christopherson, Susan, and Michael H. Belzer. 2009. "The Next Move:
Metropolitan Regions and the Transformation of the Freight Transport
and Distribution System". In Urban and Regional Policy and Its Effects,
edited by N. Pindus, H. Wial and H. Wolman. Washington: Brookings
Institution Press.
51
Corsi, Thomas M., and Philip Fanara Jr. 1989. "Effects of new entrants on
motor carrier safety". In Transportation Safety in an Age of Deregulation,
edited by L. N. Moses and I. Savage. New York City: Oxford University
Press.
Corsi, Thomas M., and Philip Fanara Jr. 1988. "Driver Management Policies
and Motor Carrier Safety." Logistics and Transportation Review Vol.24,
No.2:pp 153-63.
Corsi, Thomas M., Philip Fanara Jr., and Judith L. Jarrell. 1988. "Safety
Performance of Pre-MCA Motor Carriers: 1977 Versus 1984."
Transportation Journal Vol.27:pp 30-36.
Corsi, Thomas M., Philip Fanara Jr., and Merrill J. Roberts. 1984. "Linkages
between Motor Carrier Accidents and Safety Regulation." Logistics and
Transportation Review Vol.20, No.2:pp 147-64.
Cotton, John L. , and Jeffrey M. Tuttle. 1986. "Employee Turnover: A Meta-
Analysis and Review with Implications for Research." Academy of
Management Review Vol.11, No.1:pp 55-70.
Council, Forrest M., Michael H. Belzer, John R. Billing, Kenneth L. Campbell,
James W. Dally, Anne T. McCartt, Hugh W. McGee, A. James McKnight,
Jack Stuster, Steven J. Vaughn, and Frank R. Wilson. 2003. "Letter
Report to Federal Motor Carrier Safety Administration Administrator
Annette M. Sandberg". Washington, DC. Transportation Research Board.
September 4, 2003. pp. 66.
Coyle, Ian R., David S. Sleeman, and Neil Adams. 1995. "Safety Climate."
Journal of Safety Research Vol.26, No.4:pp 247-54.
Crooks, Ed. 2011. "Oil and water: The causes and consequences of the BP
catastrophe". Review of Deep Water: The Gulf Oil Disaster and the Future
of Offshore Drilling, by the National Commission on the BP Deepwater
Horizon Oil Spill; $39 or free online at www.oilspillcommission.gov; In
Too Deep: BP and the Drilling Race that Took it Down, by Stanley Reed and
Alison Fitzgerald; John Wiley £16.99, 248 pages. Financial Times (USA),
Jan 29, W16.
Doeringer, Peter B., and Michael J. Piore. 1971. Internal Labor Markets &
Manpower Analysis. Lexington, MA: D.C. Heath and Company.
Drakopoulos, S.A, and L. Theodossiou. 1998. "Job Satisfaction and Target
Earnings." Journal of Economic Psychology Vol.18, No.6:pp 693-704.
Edwards, Richard C., Michael Reich, and David M. Gordon. 1975. Labor
Market Segmentation. Lexington, Massachusetts: D. C. Heath and
Company.
Even, William E., and David A. Macpherson. 1996a. "Employer Size and Labor
Turnover: The Role of Pensions." Industrial and Labor Relations Review
Vol.49, No.4:pp 707-28.
———. 1996b. "Why did male pension coverage decline in the 1980s."
Industrial and Labor Relations Review Vol.47:pp 439-.
Feeny, J. 1995. "Driver turnover: the leading cause of truck accident
frequency". Gladstone, MO. Motor Carrier Safety Services, Inc. pp. 18p.
52
Fernie, Sue, and David Metcalf. 1996. It's Not What You Pay; It's the Way you
Pay It and That's What Gets Results: Jockey's Pay and Performance. In
International Industrial Relations Association Study Group 9. New
Orleans.
Ferreira, Jr. Joseph. 1972. "Driver Accident Models and Their Use in Policy
Evaluation". In Analysis of Public Systems, edited by K. Drake, and
Morse. Cambridge, MA: MIT Press.
Feyer, Anne Marie, Ann M. Williamson, R. Jenkin, and T. Higgins. 1993.
"Strategies to combat fatigue in the long distance road transport industry
- bus and coach perspective". Canberra, Australia. Australian National
Occupational Health and Safety Commission, Sydney. pp. 218.
General Accounting Office - U.S. Congress. 1991. "Freight Trucking: Promising
Approach for Predicting Carriers' Safety Risks". Washington, D.C. United
States Congress, General Accounting Office. April. pp. 92.
Gibbons, Robert. 1987. "Piece-Rate Incentive Schemes." Journal of Labor
Economics Vol.5, No.4, Part 1:pp 413-29.
Global Insight Inc. 2005. "The U.S. Truck Driver Shortage: Analysis and
Forecasts". American Trucking Association. May 2005. pp. 1-39.
Golbe, D. L. 1986. "Safety and profits in the airline industry." Vol.24, No.3:pp
305-18.
Golob, Thomas F., and David Hensher. 1994. "Driver Behavior of Long Distance
Truck Drivers: The Effects of Schedule Compliance on Drug Use and
Speeding Citations". Irvine. University of California, Institute for
Transportation Studies. UPI-ITS-WP-94-4. May. pp. 33.
———. 1995. "Driver Behavior of Long Distance Truck Drivers: The Effects of
Schedule Compliance on Drug Use and Speeding Citations". Paper read
at Transportation Research Board 74th Annual Meeting, January 22-28,
1995, at Washington, DC.
Gordon, David M. 1972. Theories of Poverty and Underemployment; Orthodox,
Radical, and Dual Labor Market Perspectives. Lexington, MA: D.C. Heath
and Company.
Gordon, David M., Richard Edwards, and Michael Reich. 1982. Segmented
Work, Divided Workers: Cambridge University Press.
Governor’s Independent Investigation Panel. 2011. "Upper Big Branch: The
April 5, 2010, explosion: a failure of basic coal mine safety practices".
Wheeling, WV. National Technology Transfer Center, Wheeling Jesuit
University. May. pp. 120; figures; 3 appendices.
Groshen, Erica L, and Alan B Krueger. 1990. "The Structure of Supervision and
Pay in Hospitals." Industrial & Labor Relations Review Vol.43, No.3:pp
134S-46S.
Gustman, Alan L., and Thomas L. Steinmeier. 1993. "Pension Portability and
Labor Mobility: Evidence from the Survey of Income and Program
Participation." Journal of Public Economics Vol.50, No.3:pp 299-323.
———. 1994. "The Role of Pensions in the Labor Market: A Survey of the
Literature." Industrial and Labor Relations Review Vol.47, No.3:pp 417-
38.
53
Hensher, David A., Helen C. Batellino, Julie L. Gee, and Rhonda F. Daniels.
1991. "Long Distance Truck Drivers On-road Performance and Economic
Reward". Sydney, NSW, Australia: Research and Analysis Report.
Instirute of Transport Studies, Graduate School of Management & Public
Policy. December. pp. 190, including appendix.
Hensher, David A., Helen C. Batellino, and Julie Young. 1989. "Long Distance
Truck Drivers - A Pilot Survey: On-road Performance and Economic
Reward". Dept. of Transport and Communications Federal Office of Road
Safety. November. pp. 91.
Hensher, David A., Rhonda F. Daniels, and Helen C. Battellino. 1992. "Safety
and Productivity in the Long Distance Trucking Industry". Paper read at
Proceedings, 16TH ARRB Conference, November 1992, at Perth, Western
Australia.
Hertz, R. P. 1991. "Hours of service violations among tractor-trailer drivers."
Accident Analysis and Prevention Vol.23, No.1:pp 29-36.
Herzberg, Frederick. 1966. Work and the Nature of Man. New York: New
American Library.
Hofmann, David A., Rick Jacobs, and Frank Landy. 1995. "High Reliability
Process Industries: Individual, Micro, and Macro Organizational
Influences on Safety Performance." Journal of Safety Research Vol.26,
No.3:pp 131-49.
Holzer, Harry J. 1990. "Wages, Employer Costs, and Employee Performance in
the Firm." Industrial & Labor Relations Review Vol.43, No.3:pp 147S-
64S.
Horrace, William C., and Thomas P. Keane. 2004. "Ranking and selection of
motor carrier safety performance by commodity." Accident Analysis &
Prevention Vol.36, No.6:pp 953-60.
Horrace, William C., Thomas P. Keane, and Kristine N. Braaten. 2002. "Motor
Carrier Industry Profile Study: Statistical Inference of Safety Performance
Measures". U.S. Department of Transportation. February 14, 2002. pp.
1-30.
Hunter, Natalie J., and Stephen L. Mangum. 1995. "Economic Regulation,
Employment Relations, and Accident Rates in the U.S. Motor Carrier
Industry." Labor Studies Journal:pp 48-63.
Ippolito, Richard A. 1991. "Encouraging Long-term Tenure: Wage Tilt or
Pensions?" Industrial and Labor Relations Review Vol.44, No.3:pp 520-
35.
John A. Volpe National Transportation Systems Center. 1998. "New Entrant
Safety Research: Final Report". Cambridge, MA. Prepared by John A.
Volpe National Transportation Systems Center for the Federal Highway
Administration Office of Motor Carriers. April 1998. pp. 1-45.
Kalnbach, L. R., and Brenda M. Lantz. 1997. "The Effects of Optimism and
Willingness to Trust on Work-Related Attitudes and Behaviors: An
Application to the Commercial Vehicle Industry". Fargo, ND. Mountain-
Plains Consortium, North Dakota State University. July, 1997. pp. 32.
54
Keller, Robert T. 1984. "The Role of Performance and Absenteeism in the
Prediction of Turnover." Academy of Management Journal Vol.127,
No.1:pp 176-83.
Keselman, Gerald, Michael T. Wood, and Eileen Hagen. 1974. "Relationship
Between Performance and Satisfaction Under Contingent and
Noncontingent Reward Systems." Journal of Applied Psychology Vol.59,
No.3:pp 374-76.
Kotz, Hein, and Hans-Bernd Schaefer. 1993. "Economic incentives to accident
prevention: An empirical study of the German sugar industry."
International Review of Law and Economics Vol.13:pp 19-33.
LaMere, Jeanne M., Alyce M. Dickinson, Michele Henry, Henry Gordon, and
Alan Poling. 1996. "Effects of a Multicomponent Monetary Incentive
Program on the Performance of Truck Drivers." Behavior Modification
Vol.20, No.4:pp 385-405.
Lazear, Edward P. 1979. "Why is There Mandatory Retirement?" Journal of
Political Economy Vol.87, No.6:pp 1261-84.
———. 1990. "Pensions and Deferred Benefits as Strategic Compensation."
Industrial Relations Vol.29, No.2:pp 263-80.
———. 1995. Personnel Economics. Cambridge, MA: MIT Press.
LeMay, Stephen A., G. Stephen Taylor, and Gregory B. Turner. 1993. "Driver
Turnover and Management Policy: A Survey of Truckload Irregular Route
Motor Carriers." Transportation Journal Vol.33, No.2 - Winter:pp 15-21.
Leonard, Jonathon S. 1987. "Carrots and Sticks: Pay, Supervision, and
Turnover." Journal of Labor Economics Vol.5, No.4, part 2:pp S136-S52.
Levine, David I. 1992. "Can Wage Increases Pay for Themselves? Tests with a
Production Function." Economic Journal Vol.102, No.414:pp 1102-15.
Lin, T. D, P. Jovanis, and C. Z. Ynag. 1993. "Modeling the Safety of Truck
Driver Service Hours Using Time-Dependent Logistic Regression."
No.1407:pp 1-10.
Loeb, Peter D., Wayne K. Talley, and Thomas J. Zlatoper. 1994. Causes and
deterrents of transportation accidents: An analysis by mode. Westport,
Conn: Greenwood.
Maslow, Abraham. 1954. Motivation and personality. New York: Harper and
Row.
Mason Jr., J. M., M. L. Patten, C. M. Plummer Jr., and R. J. Micsky. 1991.
"Safer Truck Travel". University Park, PA. Pennsylvania Transportation
Institute, Pennsylvania State University; Kelly Michener, Incorporated;
Pennsylvania Department of Transportation. November 1991. pp.
Matthews, Gerald. 1996. "Individual differences in driver stress and
performance". Paper read at 40th Human Factors and Ergonomics
Society Meeting, at Philadelphia, PA.
Mayhew, Claire, and Michael Quinlan. 1997. "Trucking Tragedies: Why
Occupational Health and Safety Outcomes are Worse for Subcontract
Workers in the Road Transport Industry". Sydney: Working Paper.
University of New South Wales. ISSN 1325-8028. pp. 31.
55
———. 2006. "Economic pressure, multi-tiered subcontracting and
occupational health and safety in Australian long-haul trucking."
Employee Relations Vol.28, No.3:pp 212-29.
Mayhew, Claire, Michael Quinlan, and Rande Ferris. 1997. "The effects of
subcontracting/outsourcing on occupational health and safety: Survey
evidence from four Australian industries." Safety Science Vol.25, No.1-
3:pp 163-78.
McCartt, Ann T., M. C. Hammer, and S. Z. Fuller. 1997a. "Work and
Sleep/Rest Factors Associated with Driving While Drowsy Experiences
Among Long-Distance Truck Drivers". Paper read at 41st Annual
Conference of the Association for the Advancement of Automotive
Medicine, November 10-11, 1997, at Orlando, FL.
McCartt, Anne, Mark C. Hammer, and Sandra F. Fuller. 1997b.
"Understanding and Managing Fatigued Driving: Study of Long-Distance
Truck Drivers in New York State". Paper read at International Large
Truck Safety Symposium, at Knoxville, TN.
McElroy, James C., Julene M. Rodriguez, Gene C. Griffin, Paula C. Marrow,
and Michael G. Wilson. 1993. "Career Stage, Time Spent on the Road,
and Truckload Driver Attitudes." Transportation Journal Vol.32, No.1:pp
5-14.
Milgrom, Paul R., and John Roberts. 1992. Economics, Organization and
Management. Englewood Cliffs: Prentice Hall.
Milkovich, George T., and Jerry M. Newman. 1993. Compensation. 4 ed.
Boston, MA: Irwin.
Milkovich, George T., Alexandra Wigdor, Renae F. Broderick, Anne S. Mavor,
James Baron, Michael Beer, Charles Brown, Thelma A. Crivens, Charles
Fiero, Patricia Ingraham, William Morrill, Kevin Murphy, James Perry,
Walter H. Read, and Carolyn J. Sax. 1991. Pay for Performance:
Evaluating Performance Appraisal and Merit Pay. Edited by N. R. C.
Committee on Performance Appraisal for Merit Pay. Washington, D.C.:
National Academy of Sciences.
Miller, T. M., and Donald Schuster. 1983. "Long-Term Predictability of Driver
Behavior." Accident Analysis and Prevention Vol.15, No.1:pp 11-22.
Mitter, Eric, and Frank J. Vilardo. 1984. "Driver Data as a Predictor of Crash
Involvement: A Comparative Analysis of Heavy Truck Operators Versus
Other Drivers". Paper read at 28th Annual Proceedings American
Association for Automotive Medicine, October 8-10, at Denver, CO.
Mixon, Franklin G., Jr., and K. P. Upadhyaya. 1995. "An Empirical Analysis of
Firm Size and Vertical Integration in the Motor Carrier Industry."
Logistics and Transportation Review Vol.31, No.1:pp 21-30.
Moses, Leon N., and Ian Savage. 1992. "The effectiveness of motor carrier
safety audits." Accident Analysis and Prevention Vol.24, No.NA:pp 479-
96.
———. 1994. "The effect of firm characteristics on truck accidents." Accident
Analysis and Prevention Vol.26, No.2:pp 173-79.
56
National Commission on the BP Deepwater Horizon Oil Spill and Offshore
Drilling. 2011. "The Macondo Well and the Blowout". Washington:
National Commission on the BP Deepwater Horizon Oil Spill and
Offshore Drilling.
National Highway Traffic Safety Administration - U.S. Department of
Transportation. 1982. "Large-Truck Accident Causation". Washington,
DC. United States Department of Transportation, Research and
Development, National Center for Statistics and Analysis, NHTSA. pp.
National Transportation Safety Board - U.S. Department of Transportation.
1988. "Safety Study: Case summaries of 189 Heavy Truck Accident
Investigations". Washington, DC. U.S. Department of Transportation. pp.
———. 1990. Safety study: fatigue, alcohol, other drugs, and medical factors
in fatal-to-the-driver heavy truck crashes. 2 vols. Washington D.C.:
National Transportation Safety Board.
———. 1995. "Factors That Affect Fatigue in Heavy Truck Accidents".
Washington. U. S. Department of Transportation, National
Transportation Safety Board. Volume 1: NTSB Number: SS-95/01; NTIS
Number: PB95-917001. Volume 2: NTSB Number: SS-95/02; NTIS
Number: PB95-917002. January. pp. vii; 227.
———. 2010. "Loss of Control on Approach, Colgan Air, Inc., Operating as
Continental Connection Fight 3407; Bombardier DHC-8-400, N200WQ;
Clarence Center, New York; February 12, 2009". Washington, DC:
Aircraft Accident Report. National Transportation Safety Board. NTSB
Number AAR-10/01; NTIS Number PB2010-910401. February 2. pp.
285.
Office of Technology Assessment - U.S. Congress. 1988. "Gearing up for Safety:
Motor Carrier Safety in a Competitive Environment". Washington, DC.
U.S. Government Printing Office. OTA-SET-382. September. pp. vi; 186.
Osterman, Paul. 1978. "An empirical study of labor market segmentation."
Industrial and Labor Relations Review Vol.28, No.1:pp 508-.
Ouellet, Lawrence J. 1994. Pedal to the Metal: The Work Lives of Truckers.
Philadelphia: Temple University Press.
Paarsch, Harry J., and Bruce S. Shearer. 1997. "Fixed Wages, Piece Rates, and
Intertemporal Productivity: A Study of Tree Planters in British
Columbia". Montreal. Université Laval - Département d'économique.
January. pp. 39.
Pearson, R. A., and K. W Ogden. 1991. "Review of issues relating to drivers,
and enforcement: Australian truck safety study task 3". Melbourne,
Australia. Published by Australian Road Research Board. pp. 46.
Peck, R. C., R. S. McBirde, and S. Coppin. 1971. "The Distribution and
Prediction of Driver Accident Frequencies." Vol.2, No.NA:pp 243-99.
Piore, Michael J. 1973. "Fragments of a "Sociological" Theory of Wages". Paper
read at Proceedings of the Twenty-Fifth Anniversary Meeting, December
28-29, 1972, at Madison.
57
Quinlan, Michael. 2001. "Report of Inquiry into Safety in the Long Haul
Trucking Industry". Sydney, NSW, Australia. Motor Accidents Authority
of New South Wales. November. pp. 356-498.
———. 2007. "Report on Occupational Health and Safety at the Beaconsfield
Joint Venture Gold Mine, Tasmania Up to and Including the Time of the
Rock Fall Incident at the 925 Level of the Mine that Occurred at Around
9:23 PM on 25 April 2006, Resulting in the Death of Larry Paul Knight
and the Entrapment of Todd Andrew Russell and Brant George Webb".
School of Organisational Behavior and Management, University of New
South Wales. alx.016.002.0011 and alx.016.002.0012. 30 August. pp.
422.
Quinlan, Michael, and Philip Bohle. 2002. "Precarious Employment and work
related stress: A review of Australian and international research". Paper
read at Occupational Stress Workshop – Academy of the Social Sciences
in Australia November 28-29, 2002, at Adelaide, Australia.
Quinlan, Michael, Claire Mayhew, and Richard Johnstone. 2006. "Trucking
Tragedies: The Hidden Disaster of Mass Death in the Long-Haul Road
Transport Industry". In Working Disasters: The Politics of Recognition
and Response, edited by E. Tucker: Baywood Publishing.
Quinlan, Michael, Lance Wright, and National Transport Commission. 2008.
"Safe Payments: Addressing the Underlying Causes of Unsafe Practices in
the Road Transport Industry". Melbourne. National Transport
Commission. October. pp. 78.
Reich, Michael, David M. Gordon, and Richard C. Edwards. 1973. "A Theory of
Labor Market Segmentation". Paper read at Proceedings of the Twenty-
Fifth Anniversary Meeting of the Industrial Relations Research
Association, December 28-29, 1972, at Madison.
Rice, R. W., S. M. Philips, and D. B. McFarlin. 1990. "Multiple Discrepancies
and Pay Satisfaction." Journal of Applied Psychology Vol.75, No.4:pp
386-93.
Richard, Michael D., Stephen A. LeMay, and G. Stephen Taylor. 1995. "A
Factor-Analytic Logit Approach to Truck Driver Turnover." Journal of
Business Logistics Vol.16, No.1:pp 281-98.
Rinefort Jr., Foster C., and David D. Van Fleet. 1998. "Work Injuries and
Employee Turnover." American Business Review Vol.16, No.2:pp 9-13.
Rodrigue, Jean-Paul. 2006. Intermodal Transportation and Integrated
Transport Systems: Spaces, Networks and Flows. In Flowpolis: The Form
of Nodal Space. Las Palmas de Gran Canaria, Spain.
Rodriguez, Daniel A., Marta Rocha, and Michael H. Belzer. 2004. "The Effects
of Trucking Firm Financial Performance on Driver Safety". In
Transportation Labor Issues and Regulatory Reform, edited by J. H.
Peoples and W. K. Talley. Rotterdam, The Netherlands: Elsevier Science
Publishers.
Rodriguez, Daniel A., Marta Rocha, Asad J. Khattak, and Michael H. Belzer.
2003. "Effects of Truck Driver Wages and Working Conditions on
Highway Safety: Case Study." Transportation Research Record
58
Vol.Freight Policy, Economics, and Logistics; Truck Transportation,
No.1833:pp 95-102.
Rodriguez, Daniel A., Felipe Targa, and Michael H. Belzer. 2006. "Pay
Incentives and Truck Driver Safety: A Case Study." Industrial and Labor
Relations Review Vol.59, No.2:pp 205-25.
Rubery, J. 1984. "The Origins of Labor Market Segmentation." Cambridge
Journal of Economics Vol.8:pp 17-36.
Rubery, Jill. 1997. "Wages and the Labour Market." British Journal of
Industrial Relations Vol.35, No.3:pp 337-66.
Rubin, Kate, and Jeffrey M. Perloff. 1993. "Who Works for Piece Rates and
Why." American Journal of Agriculture Economics Vol.75, No.4:pp 1036-
43.
Sabel, Charles F. 1979. "Marginal Workers in Industrial Society". In
Unemployment and Inflation, edited by M. J. Piore. Armonk, NY: M. E.
Sharpe.
Saccomanno, F. F., L. V. Craig, and J. H. Shortreed. 1997. "Multi-Stakeholder
Perspectives on Truck Safety: Results of a Conference on Truck Safety -
Perceptions and Reality". Paper read at 76th Transportation Research
Board Annual Meeting, at Washington, DC.
Saccomanno, Frank F, M. Yu, and J. H. Shortreed. 1995. "Effect of Driver
Fatigue on Truck Accident Rates". Paper read at Urban transport VIII :
Urban transport and the environment in the 21st century, at Seville,
Spain.
Salop, Joanne, and Steven Salop. 1976. "Self-Selection and Turnover in the
Labor Market." Quarterly Journal of Economics Vol.90, No.4:pp 619-28.
Seiler, Eric. 1984. "Piece Rate vs. Time Rate: The Effect of Incentives on
Earnings." Review of Economics and Statistics Vol.66, No.3:pp 363-76.
Shearer, Bruce. 1996. "Piece-Rates, Principal-Agent Models, and Productivity
Profiles; Parametric and Semi-Parametric Evidence from Payroll
Records." The Journal of Human Resources Vol.31, No.2:pp 275-303.
Short, Jeffrey, Linda Boyle, Sandra Shackelford, Bob Inderbitzen, and Gene S.
Bergoffen. 2007. "The Role of Safety Culture in Preventing Commercial
Motor Vehicle Crashes ". Washington, D.C. Transportation Research
Board. pp. 38.
Skulley, Mark. 2009. "Truckie `safe rates' aim to cut road toll". Australian
Financial Review, May 6.
Soguel, Nils C. 1995. Contingent Valuation, Transport Safety and the Value of
Life: Introduction.
Southern, R. Neil, James P. Rakowski, and Lynn R. Godwin. 1989. "Motor
Carrier Road Driver Recruitment in a Time of Shortages." Transportation
Journal Vol.28, No.4:pp 42-48.
Stephenson Jr., Frederick J., and R. J. Fox. 1996. "Driver Retention Solutions:
Strategies for For-Hire Truckload (TL) Employee Drivers." Transportation
Journal Vol.35, No.4:pp 12-25.
Taylor, Stephen G., and J & H Marsh & McLennan. 1997. "An Analysis of the
Relationship Between Culture, Intent-to-Quit and Driver Accidents".
59
Paper read at International Large Truck Safety Symposium, at Knoxville,
Tennessee.
U.S. Department of Transportation, National Highway and Transportation
Safety Administration, and Robert M. Clarke. 1987. "Heavy Truck Safety
Study". Washington, DC. U.S. Department of Transportation. DOT HS
807-109. pp.
Urbina, Ian. 2010. "In Gulf, It Was Unclear Who Was in Charge of Oil Rig". New
York Times, June 5.
Vietorisz, Thomas, and Bennet Harrison. 1973. "Labor Market Segmentation:
The Endogenous Origin of Barriers to Mobility". Paper read at
Proceedings of the Twenty-Fifth Anniversary Meeting of the Industrial
Relations Research Association, December 28-29, 1972.
Wasielewski, P. 1984. "Speed As A Measure of Driver Risk: Observed Speeds
versus Driver and Vehicle Characteristics." Accident Analysis and
Prevention Vol.16, No.2:pp 89-104.
Weiss, Andrew. 1990. Efficiency Wages. Princeton, NJ: Princeton University
Press.
Wilde, Gerald. 1995. "Improving Trucking Safety and Profitability Through
Safety Incentive Schemes". Queen's University, Australia. pp.
Williamson, Ann M., Anne-Marie Feyer, Christine Courmarelos, and Anthony
Jenkins. 1992. "Stage 1: The industry perspective". Canberra, Australia.
Federal Office of Road Safety. pp.
Williamson, Oliver E. 1975. Markets and Hierarchies: Analysis and Antitrust
Implications: A study in the economics of internal organization. New
York: Free Press.
Wylie, C. D., T. Paul Schultz, James C. Miller, Merrill M. Mitler, and R. R.
Mackie. 1996. "Commercial Motor Vehicle Driver Fatigue and Alertness
Study Project Report". Goleta, CA. Essex Corporation, Federal Highway
Administration (US Department of Transportation), Trucking Research
Institute (American Trucking Associations) and Transportation
Development Centre, Safety, and Security (Transport Canada). pp.
Yellen, Janet L. 1984. "Efficiency Wage Models of Unemployment." American
Economic Review Vol.74, No.2:pp 200-05.
Young, Rick. 2010. Flying Cheap. USA: PBS.
http://www.pbs.org/wgbh/pages/frontline/flyingcheap/view/
