Close Window

Preliminary Regulatory Evaluation and Regulatory Flexibility Act Analysis
Hours of Service NPRM
Federal Motor Carrier Safety Administration
April 2000

Executive Summary

The Federal regulations governing the number of hours commercial motor vehicle (CMV) drivers may
drive their vehicles have remained substantially unchanged for six decades. The Federal Motor Carrier
Safety Administration (FMCSA) is considering revising these regulations to take advantage of
improved understanding of the physiology of sleep and its relationship to human alertness. Because of
the potentially large safety and economic impacts of the options under consideration, the FMCSA has
completed this Regulatory Evaluation.

The FMCSA examined 5 options. Under option 1, drivers would have to be off-duty for at least 12
consecutive hours, and could be on-duty the remaining 12 hours. There would be no distinction
between on-duty driving time and on-duty non-driving time. Drivers would be encouraged to begin
work at approximately the same time each day, and would be required to have at least 58 consecutive
hours off-duty per work week.

Under option 2, most drivers would face the same requirements as under proposal 1. Long-haul and
regional drivers could work and drive up to 12 hours, and would need a minimum of 10 consecutive
hours off-duty. The 2 additional off-duty hours could be taken during the on-duty period or added to
the consecutive off-duty period Long-haul and regional drivers would also be allowed to use a two
week schedule for determining “weekend” off-duty time, with one long and one short weekend. Worktruck
and bus drivers would need a minimum of 11 hours off-duty, 9 of which must be consecutive.
These drivers must stop driving within 15 hours of starting work, and could only drive 5 hours per day.
They would be limited to 25 hours of driving per week.

Option 3 is the same as proposal 2, with the added provision that drivers would not be allowed to drive
between the hours of midnight and 6 AM more than 18 hours per week. Option 4 is the same as
proposal 2, except that all long-haul drivers would be required to use an electronic on-board recorder
(EOBR). Option 5 is the same as proposal 4, except both long-haul and regional drivers must use an
EOBR.

The FMCSA estimates that fatigue is directly or indirectly involved in 15 percent of all fatal and injury
crashes involving a large truck, contributing to 755 fatalities and 19,705 injuries annually. We assume
that options 1 and 2 would reduce these crashes by 5 percent annually, option 3 by 7.5 percent, option
4 by 20 percent for long-haul drivers and 5 percent for other drivers, and option 5 by 20 percent for
long-haul and regional drivers and 5 percent for all other drivers. We did not include the benefit of a
reduction in property damage only crashes.

All options would eliminate the record of duty status (RODS), reducing paperwork by between 32 and
37 million hours annually. The removal of this paperwork requirement accounts for between 70 and 50
percent of the overall benefit of these options.

Table ES-1 shows the costs, benefits, and crashes prevented for the 5 options. All dollar figures are
for 10 years, at a 7 percent discount rate. Options 1 and 2 have the lowest benefits, with costs slightly
higher than option 3, resulting in fairly high low benefits of over $1.7 billion. Option 3 has the lowest
costs, at $2.6 billion, and net benefits of $2.4 billion. Options 4 and 5 have high costs and benefits,
with net benefits of $2.3 and $3.4 billion, respectively.

Table ES-1
Costs and Benefits Proposals

Because the elimination of the RODS accounts for such a large percentage of total benefits, we also
calculated the net costs and benefits independent of the paperwork benefits. Removing this benefit
lessens the net benefits of all proposals, with all options except number 5 shifting from positive net
benefits to negative, as demonstrated in Table ES-2

Table ES-2
Costs and Benefits Excluding Paperwork Benefits

Because paperwork constitutes such a large percent of the benefits, the options are not particularly
ES-2 sensitive to changes in the estimated value of key parameters. Sizeable changes in the assumed percent
of crashes which are fatigue related, or in the percent of crashes that will be prevented, have only a
minor impact on the results. Table ES-3 shows the impact of assuming that 7.5 percent of all crashes
are fatigue-related, rather than the 15 percent used throughout this analysis. This change lowers the net
benefits of all options, with option 4 suffering the largest decline.

Table ES-3
Impact of 7.5 Percent Baseline Fatigue Related Crash Rate

While there will be some administrative costs required to comply with these options, the FMCSA did
not include these costs in this analysis. These costs are difficult to reliably predict, and are generally not
significant. The FMCSA also did not estimate the number of property-damage-only crashes prevented
by these proposals. Including these extra costs or benefits would not substantially alter the overall
results. Additionally, property-damage-only crash information is not generally considered reliable.

The Agency’s estimate of both the baseline percent of fatigue-related crashes and the percent reduction
in these crashes caused by the options are on the low end of the likely range. The FMCSA chose
these conservative values to avoid biasing the results towards a particular outcome. If significantly more
than 15 percent of crashes are fatigue related, then the benefits of any option would be significantly
higher.

Chapter 1
Background

Since 1938, the Federal government has regulated the number of hours that commercial motor vehicle
(CMV) drivers can work and drive. While there have been many revisions over the last six decades,
the essence of the regulations has changed very little. Many drivers and other industry observers
believe the current hours of service (HOS) regulations are outdated, and do not adequately protect
either the motoring public or CMV drivers. The Federal Motor Carrier Safety Administration
(FMCSA) is therefore considering changes to the HOS regulations. Because the changes may have a
significant economic and safety impact, the FMCSA has prepared this regulatory evaluation and
regulatory flexibility analysis, in accordance with Executive Order 12866 and Department of
Transportation (DOT) policy.

Chapter 1 of this report discusses the background to this proposal, including a brief description of the
current regulations and some shortcomings with these regulations. Also included is a description of the
options analyzed for this proposal. Chapter 2 provides baseline vehicle and driver information, with
detailed estimates of driver operational characteristics. Chapter 3 presents the FMCSA’s estimate of
the number of fatigue related truck crashes, including a description of some relevant characteristics of
these crashes. Chapter 4 details the benefits of the proposals, which consist of a reduction in fatigue
related crashes and elimination of a significant paperwork burden. Chapter 5 analyzes the costs of the
five options. Finally, chapter 6 compares the costs and benefits of the options, and presents some
sensitivity analysis.

Background

Federal regulations divide drivers’ time into four categories: on-duty driving, on-duty not driving, offduty,
and sleeper berth use. Driving time includes all time spent at the wheel of a CMV in operation. It
includes much of the time the driver is waiting, while in control of the CMV, to be loaded and unloaded
as well as time the vehicle is in motion. On-duty not driving time includes all time the driver is required
to be ready to work but is not actually operating the controls of a motor vehicle. It includes, among
other things, time inspecting the vehicle, time waiting for an assignment, time loading or unloading the
truck, time repairing a truck, and any time the driver is engaged in an activity for which he will be
compensated. Off-duty time includes time the driver is not on-duty. Sleeper berth time includes only
resting and sleeping in the sleeper berth.

Current Federal HOS regulations limit driving and on-duty time, for any one period and cumulatively.
After a minimum 8-hour off-duty period, drivers may not drive more than 10 hours, nor may they drive
after having been on-duty (including both driving and non-driving time) more than 15 hours. After 10
hours of driving, or being on-duty for 15 hours, a driver may not drive again until he has a minimum of 8
hours off-duty. Drivers who work for companies which operate CMVs 7 days a week may not drive
after having been be on duty more than 70 hours in the previous 8 days. If the company does not
operate CMVs every day, the driver may not may not drive after having been on-duty more than 60
hours in the previous 7 days. Drivers may take an exception to the required 8 hours off-duty period by
using a specification sleeper berth and splitting consecutive off-duty time into two periods.

Drivers are required to keep a record of their time in a record of duty status (RODS), commonly
known as a logbook. The RODS must show the driver’s status in 15 minute increments, as well as
certain other information. Drivers must keep a current copy of their RODS in their possession while
driving. Drivers who operate within a 100 air-mile radius of their normal work reporting location are
not required to maintain a RODS, but their employers must maintain a time record.

Drivers in the States of Alaska and Hawaii, as well as drivers of oil field equipment, utility service
vehicles, construction material and equipment vehicles, farmers, and ground water well drilling
equipment, are granted various exemptions to some of the HOS and RODS provisions. The HOS
regulations may be found at 49 CFR part 395, with relevant definitions in $395.2.

Much of the discussion in this evaluation uses the term fatigue. Fatigue has become a shorthand phrase
denoting sleepiness, drowsiness, and low-alertness. However, the human-factors researcher Ivan
Brown defines fatigue as the decreased capability of doing physical or mental work, or the subjective
state in which one can no longer perform a task effectively. Fatigue can not be measured. What can
and has been measured is how alert a person is, how they perform on tasks requiring sustained
attention, hand-eye coordination, and responding to changes in the environment. Researchers have also
measured changes in how the body functions, including brain wave patterns, and eyelid position, that
relate to how alert or drowsy a person is. Because of its more common usage, this document generally
uses the term fatigue.

There are two distinct but related safety issues concerning the existing HOS regulations. First,
compliance with the regulations does not necessarily ensure that drivers are adequately rested and alert
Drivers may comply fully with the regulations and still suffer from severe fatigue. Second, many drivers
do not comply with these regulations. Numerous surveys and much anecdotal evidence indicate that
many drivers violate the HOS regulations. These non-compliant drivers are especially susceptible to
fatigue.

Moreover, a driver may be fully compliant with HOS regulations and not be sufficiently rested.
Researchers understanding of sleep has improved dramatically since the original HOS rules were
written. Researchers have examined the biology and physiology of sleep, the mechanics of circadian
rhythms, differences between daytime and nighttime rest, and the effects of cumulative sleep
deprivation, among other things. There is evidence that the existing HOS regulations do not comport
well with this scientific understanding of the role and structure of sleep, and, when used to set minimum
off-duty periods rather than maximum driving periods, might actually promote drivers’ fatigue.
According to an expert panel of sleep researchers and transportation safety experts convened by the
FHWA, “...the hours of service regulations are sorely inconsistent with the best available information
today.” (Expert Panel Report).

The FHWA convened an expert panel to review the current state of knowledge about sleep and
fatigue, and to examine HOS options presented by the Agency. The expert panel included eight
researchers expert in traffic safety, human factors, and sleep medicine. Panel members, who work in
academia, government, safety organizations, and as private consultants, were provided summaries of
over 80 (mostly peer reviewed) research reports compiled by the FHWA. The panel was asked to
evaluate the current regulations and various proposals in light of the scientific understanding of sleep and
alertness. This section describes some of the findings of the Expert Panel, particularly the discussion of
the inadequacies of the present regulations. The complete Expert Panel Report may be found in the
public docket.

One major concern of the panel was the absence of consideration of a 24 hour cycle in the HOS
regulations. Human evolution responded to the natural light cycle, and human biology continues to
exhibit strong cyclical effects. Human metabolism, and thus alertness, shows daily 24-hour patterns,
including primary and secondary peaks and troughs. These peaks and troughs appear independent of
the effects of light, indicating that they are deeply embedded in the human physiology. Since 1962,
HOS regulations have had no 24 hour component. A driver could conceivably drive from midnight to
10 AM, rest from 10 AM to 6 PM, then resume driving until 4 AM the following day. This shortens
the day by up to 6 hours, equivalent to an east-to-west transatlantic flight in terms of “jet lag”.

Another concern of the Panel was the difference between daytime and nighttime driving. The Panel’s
report noted several problems associated with nighttime driving. First, as demonstrated by the Driver
Fatigue and Alertness Study (Wylie et al), the strongest and most consistent factor influencing fatigue
and alertness was time of day. Night driving was associated with a higher level of observed
drowsiness, poorer lane-tracking, and degradation of mental performance. In addition, the Panel noted
evidence suggesting that daytime sleep is not as restorative as nighttime sleep, with both fewer hours
generally spent sleeping and a lower quality of sleep. Drivers generally agree that nighttime sleep is
superior to daytime sleep (Abrams et al.). The result is that overall alertness and performance is lower
in the nighttime than in the day, and crash risk is correspondingly higher. The Panel’s report cites
evidence suggesting that nighttime driving is associated with as much as a 4-fold or more increase in
fatigue-related crashes. The existing regulations make no distinction between day and nighttime driving.

The Panel noted the importance of continuous time off-duty. They reported that sleep obtained in
discontinuous segments is not as restorative as continuous sleep. The Expert Panel also cited studies
which demonstrate that longer periods of off-duty time are associated with longer periods of sleep. The
current regulations require that drivers have at least 8 continuous hours off-duty before returning to
duty The Expert Panel criticized this requirement as inadequate, because it does not allow drivers time
to travel to a resting place or to take care of personal needs, and because 8 hours off-duty time
generally does not translate into 8 hours of sleep. Wylie showed that people who are off-duty for 8
hours generally only obtain about 5 hours of sleep (Wylie et al).

The Panel also asserted that there should be no differentiation between driving time and on-duty notdriving
time. They cited several studies which show that performance of tasks declines with increased
time on duty, regardless of how on-duty time is spent. The panel believes that all on-duty time should
be treated the same, as the effect on driver safety is similar.

The Panel agreed that limits on cumulative on-duty time were required, because drivers might not get
adequate rest, even if they have sufficient opportunity to rest on a daily basis. Like all workers, drivers
have non-work responsibilities which may lessen the amount of time devoted to rest on a specific day,
and it is often difficult to catch up on missing sleep. According to the panel, this can result in cumulative
fatigue.

Related to a limit on cumulative on-duty time is the need for adequate recovery time. Because sleep
loss is cumulative, it cannot be remedied by a single night’s rest. The Driver Fatigue and Alertness
study and other research demonstrated the prevalence of cumulative sleep deprivation in the motor
carrier industry, with average sleep lengths of just under 4 to 5.5 hours. The Expert Panel also cited
research which shows that nighttime sleep is more efficient than daytime sleep, with more time in bed
spent sleeping. Therefore, the Panel recommended that any new regulations must allow for at least one
off-duty period every seven days which encompasses two midnight to 6 AM periods.

The panel commented on the need for drivers to have advance knowledge of their work schedule, and
on the specific issues related to sleeper berth use and split shift driving. As noted above, the Expert
Panel report is available for review in the public docket.

While drivers who abide by the HOS rules may be fatigued, the situation of drivers who are not in
compliance is undeniably worse. Whatever the limitations of, for example, 5 to 6 hours of interrupted
sleep, it is clearly more restorative than no sleep. Unfortunately, many drivers violate the HOS
regulations. The Insurance Institute for Highway Safety (IIHS) interviewed over 1,200 drivers at truck
stops, truck inspections stations, and agricultural inspection stations in the early 1990s. Based on the
drivers’ responses, the authors classified 73% of the drivers as hours-of-service violators (Braver et
al.). A 1995 survey of over the road drivers in New York State found that over one-third reported
driving more than 60 hours in a typical week, and a similar proportion reported working 70 hours in a
typical week. More than 40 percent of respondents reported always/often/sometimes driving more
than ten consecutive hours without 8 hours off-duty (McCartt et al.). A more recent survey performed
by the University of Michigan’s Trucking Industry Program (UMTIP) corroborated this high violation
rate. UMTIP surveyed over 800 mostly over-the-road drivers at a number of truck stops in the
midwest in 1997. Only 16% of drivers surveyed reported that log books were generally accurate, and
56% stated that they had worked more hours than recorded in their RODS in the last 30 days. The
average driver worked over 64.3 hours in the previous 7 days. Twenty five percent of drivers reported
working at least 75 hours in the last 7 days, and 10% of drivers reported working more than 90 hours.
The UMTIP survey is described at greater length in Chapter 2.

Thus, drivers who comply with the HOS regulations may still not be adequately rested, and the
significant percentage who do not comply are assuredly not rested.

Other organizations have also indicated their concern over driver fatigue, and their concomitant belief
that the present hours of service regulations do not adequately ensure that drivers are rested. Driver
fatigue was voted the number one safety concern of the 1995 Truck and Bus Safety Summit, a meeting
of over 200 drivers, motor carrier representatives, government officials, and safety advocates. The
National Transportation Safety Board has asked the FHWA to investigate and promulgate new
regulations to combat driver fatigue, and Congress has mandated that the FHWA issue a rule “dealing
with a variety of fatigue-related issues” in the ICC Termination Act of 1995.

Proposals

Improved understanding of the mechanics of fatigue from sleep research, intense interest expressed by
Congress and others, and the perceived widespread violation of the existing hours-of-service
regulations prompted the FMCSA to consider changing them. As was noted above, the underlying
goal of the current regulations is to prevent drivers from driving more than a certain number of hours per
day and week. The lodestar of the proposals under consideration is to assure that drivers have an
opportunity to obtain sufficient rest. The options are described below.

Option 1 12/12

Under this proposal, all drivers must have a minimum of 12 consecutive hours off-duty in every 24 hour
period, and may work for the other 12 hours. Rest and meal breaks within the 12 hour work period
would not count as off-duty time. Drivers must also have at least 58 hours of consecutive off-duty time
every week, including a minimum of two midnight to 6 AM periods. This would allow for more than 2
full days off to obtain quality, uninterrupted, restorative sleep.

Motor carriers would be encouraged to start drivers work days approximately the same time each day,
with some leeway. Motor carriers wouId be encouraged to limit drivers backwards rotating their shifts
to no more than 1 hour a day, with no limits on forward shifting. A driver who works from 6 AM to 3
PM Monday, and then takes the required 12 hours off, should not begin work again until 5 AM
Tuesday, even though he accumulates 12 consecutive hours off-duty by 3 AM. In this case, the driver
would be encouraged to take 14 consecutive hours off-duty, since the recommendation on backward
shifting schedules would prevent him from starting before 5 AM.

Although backward rotation is disruptive of drivers circadian rhythm, as noted previously, the FMCSA
does not currently propose to ban the practice. The Agency does not believe the scientific community
has reached a consensus on the magnitude of the harm to drivers from a backwards rotating schedule.
While credible research agrees on the undesirability of this practice, it is not clear how large a role
backwards rotating schedules, in isolation, play in fatigue-related crashes.

The FMCSA considered mandating the regularity practices outlined above, but the costs were found to
be unexpectedly high. Given the high costs and the uncertainty over the importance of irregularity, the
FMCSA refrained from requiring regularity. Nonetheless, the Agency does encourage drivers and
motor carriers to ensure that driving begins at approximately the same time each day.

All the options change the definition of on-duty and off-duty time to be consistent with Department of
Labor (DOL) regulations governing the minimum wage, since most employed drivers are subject to the
minimum wage provisions of the Fair Labor Standards Act (FLSA). By adopting the DOL’s
definitions, the FMCSA will be able to rely largely on DOL-required paperwork to enforce the HOS
regulations.

Because the FMCSA proposes to rely primarily on DOL-required paperwork, drivers would no longer
have to prepare an RODS. Instead, on-duty and off-duty time would be monitored by a DOLrequired
time card or sheet, which includes the time a driver checks into and out of work. Time cards
would be kept at the drivers normal work reporting station, and most drivers would not have to carry
their time card while driving. Because of their extended absence from their normal work reporting
stations, long haul and regional drivers would be required to have a current time card, with annotations
of the locations where they change their duty status, on their vehicles while driving.

Option 2

For most drivers, option 2 is the same as option 1. The majority of drivers must have 12 consecutive
hours off-duty However, this option provides alternatives for two groups of drivers: (1) long haul and
regional drivers, and (2) split shift drivers and work truck and bus drivers.

Long-haul and regional drivers must have a minimum of 12 hours off-duty, 10 of which must be
consecutive. These drivers would be allowed to drive for up to 12 hours. Two of the off-duty hours
could be either contiguous with the 10 consecutive off-duty hours, or taken during the 12 hours of
driving.

Long-haul and regional drivers would also have a choice in calculating weekly driving time. They could
follow the procedure outlined in Option 1, whereby they would need 58 consecutive off-duty hours,
including 2 consecutive midnight to 6 AM periods, at the end of the work week. Alternately, they
could use a two week schedule. Under the two week schedule, drivers would have one short and one
long “weekend”. For the short weekend, drivers would need 32 hours of consecutive off-duty time,
including 2 consecutive midnight to 6 AM periods. After the short weekend, the driver could drive up
to an additional 48 hours over the next 4 days, but then would require 82 consecutive hours off-duty,
including 2 consecutive midnight to 6 AM periods, before returning to work.

Work truck and bus drivers are those who spend less than one third of their on-duty time driving a
truck or bus. This category excludes drivers for for-hire motor carriers of passengers. These drivers
may only drive 5 hours per day, with a weekly maximum of 30 hours. They must have at least 9
consecutive off-duty hours per day, and 2 additional off-duty hours which may be either contiguous
with the 9 hour off-duty period, or taken during the driver’s work period. Work truck and bus drivers
may work up to 13 hours a day, and they are not allowed to drive 15 hours after beginning work.
Work truck and bus drivers must also have a minimum of two midnight to 6 AM off-duty periods per
week.

Option 3 - Option 2 with exemptions and nighttime differential

This is the same as option 2, except all drivers would be limited to a maximum of 18 hours of driving
per week between the hours of midnight and 6 AM. The purpose of this option is to limit the number of
hours driving in the middle of the night, because of abundant evidence which indicates that this is the
most fatiguing period for drivers.

Option 4 - Long-Haul Drivers must use an EOBR

This is the same as option 2, except long-haul drivers must use an electronic on-board recorder
(EOBR). The EOBR will allow companies and enforcement officers to more accurately monitor
drivers compliance with the hours-of-service regulations.

Long haul carriers which employ fewer than 20 drivers must use an EOBR within 4 years of
promulgation of a final rule. Carriers with 21 to 49 drivers would have 3 years to comply, and those
with more 50 drivers would need an EOBR within 2 years.

Local and work truck drivers will be allowed to use the DOL time card. Because of their extended
absence from their reporting location, regional drivers will be required to continue to use the existing
RODS. Likewise, long haul carriers must continue to use the RODS until they begin using an EOBR.

Option 5 - Long-Haul and Regional Drivers must Use an EOBR

This is the same as option 2, except both long-haul and regional drivers must use an EOBR. These
carriers must continue to use the RODS until they begin using an EOBR.

Options

As mentioned above, the linchpin of all these options is to allow drivers adequate consecutive off-duty
time to obtain sufficient rest. Science indicates that there is no precise length of time all drivers can
operate safely. That point varies with the amount and quality of sleep a driver has obtained. Drivers
who have had only three hours of sleep in the last 24 hours, for example, are far less likely to drive
safely, for as many hours, as their counterparts who have slept 10 of the previous 24 hours. So rather
than focus on a maximum driving limit, the FMCSA proposes to establish minimum opportunity-to-rest
times.

As a result of this proposed focus, some drivers may be able to drive more hours in a work-rest cycle
than are currently permitted. For example, while drivers are currently limited to 10 hours of driving
before taking an 8-hour break, under these proposals some drivers could drive as many as 12 hours
before stopping. There are several reasons the Agency believe that this would not increase driver
fatigue.

First, although drivers are currently limited to 10 hours of driving, they may drive during, or after, their
12” hour on-duty Individuals who begin their work shift with non-driving on-duty tasks, which include
such strenuous activities as loading or unloading trailers, may end up driving into the 15” hour on-duty
The expert panel noted that there should be no distinction between driving and other non-driving onduty
tasks, arguing that “cumulative hours on duty increases fatigue.” Research suggests that crash risk
increases with on-duty time, rather than time on a specific task.

Second, a driver may drive as many as 16 hours in a 24 hour period under the current regulations. As
noted previously, a driver may drive 10 hours, take 8 hours off duty, and then continue driving for
another 10 hours. This equals 16 out of 24 hours, and 20 out of 28. By contrast, the proposed
options would limit drivers to 12 out of 24 hours (or fewer, for many drivers).

Third, the options would require that drivers obtained between 9 and 12 consecutive hours off-duty,
depending on the driver type and option. Currently, drivers are only required to have 8 consecutive
hours off-duty, which often does not leave sufficient time for rest (Wylie et al).

Fourth, the weekly limitations proposed guarantee that drivers obtain sufficient cumulative rest, and that
some of that rest takes place in the midnight to 6 AM period, when sleep is most restorative. The
existing HOS regulations make no distinction between daytime and nighttime driving.
Finally, as noted above, these proposals also encourage drivers to begin work at approximately the
same time each day, and to obtain sufficient cumulative rest at the end of each work week. Research
indicates this combination will enhance driver’s overall levels of rest and attentiveness.

Thus, while the Agency understands that these proposals may result in some drivers driving for longer
periods than are now allowed, we believe the cumulative effect of these changes will be increased
driver alertness and safety.

While drivers who comply with the regulations should be more rested, potential violations must also be
addressed. We believe that this proposal will result in fewer violators, and those who continue to
exceed the maximum allowable hours will still be subject to FMCSA fines and other penalties. A
frequent comment made at the thirteen nationwide outreach sessions held in support of the zero-base
review of the FMCSRs was the need to make the regulations more understandable. These proposals
are simpler and more comprehensible than the existing standards The Agency is proposing to simplify
the regulations by eliminating the on-duty not driving category, making many drivers subject to a simple
12/12 schedule, and removing the log book requirement. We believe that people are more likely to
comply with rules which are easier to understand and that fit their physiological needs.

In addition, Options 4 and 5, which require that certain drivers use EOBRs, will make it more difficult
for drivers to violate the HOS regulations. EOBRs must produce a paper printout or other format
which can be read at the roadside. This output will allow enforcement officers to determine the number
of hours the vehicle has been in operation, and when the operation began Enforcement officers will
have a minimum of two accurate data points, which will make it easy to determine how long a driver
has been on-duty. It may still be possible for drivers to conceal the number of their driving hours, but
the use of an EOBR will make it significantly more difficult to do so. EOBRs will also allow motor
carriers to better monitor driver’s driving time. Accordingly, the FMCSA assumes that options 4 and 5
will lead to a greater reduction in fatigue-related crashes than the other three options.

Many drivers have argued that the current regulations do not comport with their schedules, forcing them
to violate the hours-of-service regulations. Specifically, the fact that drivers are often not on a 24-hour
cycle skews their natural sleep/wake patterns, which may not be aligned with their driving hours.
Drivers may be required to go off-duty when they are not tired, or they may have to continue driving
during a circadian trough. Some drivers choose not to go off duty and violate the HOS regulations
rather than discontinue driving if they do not feel tired. By encouraging drivers to operate on a 24-hour
schedule, drivers’ driving schedules and circadian rhythms should be more nearly congruent, lessening
the tendency for drivers to drive over hours and while fatigued.

While work truck and bus drivers are allowed the least driving time, they may have the greatest amount
of time on-duty. A work truck driver could be on duty for 65 hours in a week, and may operate a
CMV up to 15th hours after commencing work. The FMCSA is proposing these more lenient
provisions because of these drivers’ reduced risk of fatigue-related accidents. Research indicates that
these drivers are an order of magnitude less likely to be involved in a fatigue-related crash than are
other drivers. Chapter 3 describes the research in more detail.

Chapter 2
Baseline Vehicle and Driver Data

The regulatory options under consideration distinguish between drivers based on the characteristics of
their operations. Therefore, the FMCSA generated estimates of the number of vehicles and drivers in
the relevant categories. This chapter presents these estimates, as well as information on driver
characteristics.

While the Agency is reasonably confident in the estimate of the total number of drivers, there is more
uncertainty about the distribution of drivers by operational type. The options envision 5 classes of
drivers: long haul drivers, regional drivers, split-shift drivers, work-truck drivers, and local non-work
drivers There is no publicly available data which classifies drivers into these categories, so the
FMCSA developed them from data collected for other purposes.

The FMCSA relied largely on data from the Truck Inventory and Use Survey (TIUS), a survey
conducted every five years by the Bureau of the Census of the U.S. Department of Commerce. TIUS
is described in detail in Appendix A. The 5 driver categories were culled down to 4, because no data
is available on split-shift drivers. We assumed that vehicles with an average trip distance of more than
500 miles are long haul, those with average trip distances of between 200 and 500 miles are regional,
and those less than 200 miles are either local or work trucks, depending on their primary use and
vehicle body type. These simplifying assumptions are reasonable but not unassailable. Therefore,
results pertaining to costs and benefits by type of operation are considered less reliable than overall
results.

1) Number of Drivers and Vehicles

The number of drivers was based on data from the FHWA’s 1996 Controlled Substances and Alcohol
Testing Survey, supplemented by information from the Motor Carrier Management Information System
(MCMIS) census file. The survey is a statistical sample of several thousand motor carriers selected
from the MCMIS file, and it included a question about the number of CDL-drivers employed. These
data and sampling weights from the survey were used to generate an estimate of CDL drivers. This
number was multiplied by the ratio of non-CDL to CDL drivers from MCMIS to estimate non-CDL
drivers, and the numbers were summed to generate a total. The FMCSA estimates that there are 6.4
million interstate and intrastate CMV drivers subject to the current hours of service regulations.

This regulatory analysis includes both interstate and intrastate drivers. While the current HOS
regulations only apply to interstate drivers, all States have adopted compatible regulations governing
intrastate drivers. 48 States and the District of Columbia receive grant money through the FMCSA’s
Motor Carrier Safety Assistance Program (MCSAP). One requirement of participation in MCSAP is
that the State adopt MCSAP compatible regulations. Accordingly, the FMCSA believes that most
States will adopt regulations substantially similar to what the Agency chooses. Therefore, the options
under consideration will effectively regulate both interstate and intrastate drivers, and both have been
included in this analysis.

The FMCSA categorized the 6.4 million drivers by operational type, compatible with the definitions in
the HOS options. Vehicle data from the TIUS was used. The 1992 TIUS, the most recent survey
completed, obtained detailed physical and operational data on more than one hundred and twenty five
thousand trucks. After eliminating trucks under 10,000 pound gross vehicle weight rating (GVWR)
from the sample, vehicles were stratified based on the percent of miles accounted for by various trip
lengths. Missing data were distributed based on the distribution for similar non-missing vehicles.
Single-unit trucks with predominately short trips were stratified as work trucks or non-work trucks,
based on the trucks’ repotted major use. The FMCSA assumed that drivers were distributed
proportionately to trucks, which resulted in the following driver counts.

Table 1
Number of Drivers by Operation Type

Appendix A explains in detail the process used to generate the number and distribution of drivers.

We also used TIUS to develop estimates of the vehicle miles traveled (VMT) by trip distance and
vehicle type. Because the Trucks Involved in Fatal Accidents (TIFA) file uses the same trip distance
definitions as TIUS, these estimates allow us to analyze differences in risk for fatal crashes, including
both fatigue- and non-fatigue-related crashes. TIUS respondents report the percent of miles in each
trip category, as well as the total number of miles driven. These figures were multiplied for each
vehicle, then summed to yield a total. A given vehicle could have some miles in each of the trip
distances listed below (although this was not often the case). Estimates are presented in Table 2.

Table 2
100 Million Vehicle Miles Traveled by Power Unit Type and Trip Distance

Straight trucks and tractors are used differently. Almost 60 percent of straight trucks have a most
frequent trip distance of less than 50 miles, compared to just 12 percent of tractors. At the other
extreme, tractors accumulate over one third of their miles in trips of greater than 500 miles, compared
to one in 50 miles for straight trucks.

Partly because of the difference of travel between straight trucks and tractors, the distribution of trucks
differs considerably from the distribution of travel. Regional and long haul trucks (those with a most
frequent trip distance of more than 200 miles) account for over 40% of miles but just 13% of vehicles.
The breakdown of trucks by most frequent trip distance is presented in the following table.

Table 3
Truck Registrations by Power Unit Type and Trip Distance

Additional information on the development of these numbers can be found in the UMTRI report in the
public docket.

2) Baseline Driver Survey Information

The FMCSA relied extensively on a driver survey conducted by the University of Michigan Trucking
Industry Program (UMTIP) to analyze the costs of the proposed options. The DOT did not sponsor
this survey, and therefore some of the definitions do not precisely match those in the NPRM. This
section describes the survey and presents summary information on the results. While most of the text is
based on the following tables, some of this discussion relies on information not included in this
evaluation. The UMTIP survey analysis has been placed in the docket.

The driver survey used a two-stage randomized design in five Midwestern states. UMTIP selected
truck stops at random from the population of buck stops in these states, stratified based on the number
of parking spaces for trucks as a proxy for truck traffic. Subjects were chosen at random (every nti
individual who walked through the door). The survey took approximately 45 minutes and drivers were
paid $20 to participate. The response rate was 60% (including conversions for those who had
insufficient time at the truck stop but were interviewed at home). An additional 6% response was
achieved using a five minute questionnaire. UMTIP also surveyed drivers who were refueling to
confirm sample validity, and achieved an 96% response rate.

Drivers were asked if they were local or long haul. Drivers who called themselves local were classified
accordingly. Drivers who called themselves long haul drivers were so classified if their average trip
length was greater than 500 miles; otherwise they were categorized as regional. These definitions are
used throughout this section. However, the reported traveled distances may be high, especially for
local drivers. It is extremely unlikely that local drivers actually drive more than 80,000 miles per year,
as reported in the survey. The survey most likely undersamples local drivers, as they are unlikely to
stop at truck stops. Those who do, regardless of how they categorize themselves, are not likely to be
representative of most local drivers. Therefore, we did not use these definitions in the analysis in
Chapter 5. We believe that both long haul and regional drivers in the survey are likely to fit the
NPRM’s definition of long haul drivers. In this evaluation, long haul driver characteristics (such as
wages and hours worked) are assumed to equal an average of those reported for long haul and regional
drivers in the survey. Likewise, we assumed that local drivers in the survey are most likely to match the
NPRM’s definition of other (non-long haul, non-work truck) drivers.

Table 4 shows the trends for mileage driven by individual drivers. The average driver covers 112,765
miles annually, and long-haul drivers drive 124,475 miles. Means exceed medians, demonstrating the
extent to which high mileage figures dominate. Indeed, the top 25 percent of all long haul drivers
reported exceeding 150,000 miles and ten percent exceeded 170,000 miles.

Table 4
Annual Miles

The report also shows that employee drivers drive 5.8 percent more miles than do owner-operators,
and that union drivers run nearly 7 percent fewer miles.

The driver survey shows that, on average, drivers worked 64.3 hours in the past seven days, including
driving and on-duty-not-driving. The survey also shows, somewhat surprisingly, that local drivers work
as many hours as do long-haul drivers. We suspect this is probably because of the lack of
representativeness of local drivers in the survey discussed above. In any case, the data presented in
Table 5 suggest a broad pattern of violation, with the top 10 percent of all drivers averaging 94 hours,
including a 97 hour average for local drivers. At the median, only the local drivers stay below the 60
hour limit and regional drivers hit that limit exactly.

Table 5
Hours Worked in Last 7 Days

Again, non-union and employee drivers work longer hours than either union members or owner
operators.

The following two tables report hours working and driving in the last 24 hours. They reveal a direct, if
modest, relationship between the number of hours worked per day and the length of trip taken by the
driver. Long-haul drivers work the greatest number of hours; at the mean, long haul drivers work 12
percent more hours during any single day than do regional drivers, and 15 percent more hours on any
single day than do local drivers. Not surprisingly, they also drive more of those hours and spend less of
their time in any given day performing non-driving labor. At the mean, long-haul drivers perform nondriving
work 23 percent of their time, while comparable figures for regional and local drivers are 25
percent and 37 percent.

Table 6 suggests that many drivers may routinely exceed the daily hours-of-service rules. At the 75th
percentile, long-haul drivers are working 15.5 hours. At the 90th percentile long haul drivers work I9
hours, leaving only 5 in the last 24 for non-work activity. This suggests that a significant percentage of
all long-haul drivers may be in daily violation of the HOS limits. As noted above, other studies have
also found high violation levels.

Table 6
Hours Worked in the Last 24

Table 7
Hours Driven and Worked in the Last 24

Table 8 shows that long-haul drivers put in the most work time per trip, and their trips are considerably
longer than those of their local or regional counterparts. However, long-haul drivers spend a
disproportionately large percentage of their time waiting during trips. The average driver waits more
than twice as many hours as he performs non-driving work, but the distribution is skewed. Local
drivers wait somewhat less than they work (possibly because they are usually paid by the hour) but long
haul drivers wait almost three times as long as they perform non-driving work.

A great deal of efficiency is lost when drivers spend their on-duty time waiting. While drivers are
arguably creating value when they are working loading or unloading a truck, they are not creating any
value when they are waiting for a dispatch or delivery This inefficient use of their time is the source of a
great deal of slack in the system. Many drivers are not paid directly for this time, or earn a very small
piece-work rate for activities. The Fair Labor Standards Act requires that employees be paid for all
time they work or are “engaged to wait”, which includes time driver’s spend waiting for a dispatch or
for a vehicle to be unloaded. Drivers’ weekly wages must be high enough that they receive the
minimum wage for all hours worked, including both driving and waiting time. Therefore, employee
drivers may be legally paid for waiting time, but their marginal pay is generally close to zero. Piece
work drivers (those paid by the mile or the load) would generally receive the same pay whether they
waited 10 hours for a dispatch or 1 hour. As long as the drivers cumulative pay results in an average
hourly rate equal to, or greater than, the minimum wage, the driver’s employer is in compliance with the
relevant provisions of the FLSA. However, because marginal hourly pay equals zero, drivers perceive
this time as unpaid, a perception that most economists would support. Accordingly, this discussion
refers to this time as unpaid, while making no judgement about compliance with the FLSA.

This unproductive time is relatively costless to the shipper, consignee, and the trucking company, but
represents a significant opportunity cost to the driver. The economic cost is reflected in high turnover
and low human capital investment, as well as a tendency for drivers to pack in working (mainly driving)
hours in addition to unpaid waiting time to make up for their lost earnings.

Table 8
Time Working and Waiting in the Last Trip

Drivers clearly spend a significant proportion of their on-duty time on non-driving tasks. When this time
is unpaid, it may contribute to excessive hours, as drivers may illegally log unpaid time as off duty.
Table 9 shows paid time as a percent of all non-driving time, and indicates that much non-driving work
time is unpaid. At the mean, 29 percent of ail non-driving time is paid, but at the median the percentage
of paid time is zero. This means that more than half of all drivers earn nothing for this labor. Most
union drivers are paid for their time, as 72.7 percent of the union driver’s time is paid at the median,
while for non-union drivers, at the median their ratio of paid time to total non-driving time is zero
percent. At the 75” percentile 70.6 percent of the non-union driver’s total non-driving time is paid,
while the corresponding figure for union drivers is 100 percent. Owner-operators have an even worse
problem than do ordinary non-union drivers, as at the 901b percentile only 66.7 percent of their total
non-driving time is paid. (These data are not included in this table, but can be found in the complete
report in the docket).

Table 9
Paid Time as a Percent of Total Non-Driving Time

Earnings

The costs associated with the alternate proposals will largely be reflected in changes in drivers pay.
Accordingly, a brief discussion of the level and method of drivers pay is warranted. The UMTIP report
in the docket contains a more comprehensive discussion of driver pay.

While mean driver earnings are relatively high for somewhat skilled but generally not highly educated
workers, drivers work an excessive number of hours to achieve their earnings levels. While the average
driver earns more than $36,500 annually, he also works an average of about 3,300 hours per year to
do it. This is more than half again as many hours as the full time standard year in the United States, and
considerably greater than 50 percent more hours than the average employee actually works.

Table 10 shows that long haul drivers make less than regional drivers. This is consistent with previous
research that showed that the lowest paid drivers worked for long-haul TL carriers. That research
showed that rates for drivers working for regional carriers averaged 3 1 .O$ per mile while those working
for national carriers averaged 25.1# per mile. The highest pay rate went to national LTL carriers at
40.16 per mile, while the mileage rate for national TL drivers was 22.76, 43% less than national LTL
carriers (and they are generally not paid for their non-driving labor). (Belzer).

Table 10
Annual Wage

Owner-operators earn somewhat less than do company drivers, suggesting profits may be quite low
(owner-operators often commingle these concepts). While median earnings are the same, mean
earnings of owner-operators are about 5 percent lower than those of company drivers. Owner
operators drive fewer miles and work fewer hours, as was noted above, but their returns on capital
investment appear relatively low.

As previous research has shown, the most stiking difference in driver wages comes from the influence
of unions. Collective bargaining clearly provides union drivers with great advantages in comparison
with their non-union counterparts. Collective bargaining appears to provide a nearly 26 percent
earnings advantage over the non-union employees. Since non-union employees also work 8.3 percent
more hours, the real advantage may be closer to 34.3 percent, not including the value of benefits (which
also is considerably higher for unionized employees). Data on owner-operators and union employees
may be found in the UMTRI report in the public docket.

Most over-the-road drivers are paid on a contingent basis, that is, by the mile or by a percentage of the
load revenue. The latter method is most common among owner-operators, who usually act as
subcontractors for motor carriers. It also is common among non-union drivers, but is relatively
uncommon for union drivers.

Night Driving

The UMTIP survey asks drivers how many hours they worked between the hours of 11 PM and 7
AM, commonly known as the “graveyard shift.” Drivers reported a relatively small amount of night
driving, 29.0 percent at the mean. Local and long haul drivers put in a somewhat smaller proportion of
their time in night driving, and regional drivers a higher proportion.

Option 3 proposes limiting night driving (defined as between 12:00 AM and 6:00 AM) to 18 hours per
driver per week. The average driver works 64.3 hours a week (Table 5), of which 26 percent (Table
7) is non-driving. Multiplying 64.3 by .74 suggests that the average driver drives 47.6 hours per week.
Based on the “last full trip” information, the average driver in the sample spends 29 percent of his or her
driving time between the hours of 11:00 p.m. and 7:00 a.m. (the hours asked about in the survey). I f
we assume a uniform distribution of reported night driving over this period, then we can estimate that
the average driver would have 75 percent of his overall night driving hours in the policy-relevant six
hours (between midnight and 6 AM). We calculate that the average driver would be driving 21.75
percent of his time during the proscribed hours 0.75 x 0.29). Using the sample mean driving total of
47.6 hours, this suggests the average driver currently drives 10.4 hours during the midnight to 6:00 AM
period each week, well within the prospective limits. UMTIP believes this figure is conservative, since
they ended up adjusting their interview schedule because fewer drivers were available to interview in
the early morning.

These calculations are based on the mean; depending on the characteristics of those drivers exceeding
the mean either in hours worked or percent of night driving, these characteristics might be different at
the extremes. For example, in the LTL sector most of the regional carriers’ drivers operate through the
night, five days a week, but do so within the 60 hour weekly limit. In the national LTL and in the
package delivery sector, tractor trailer combinations run throughout the night as well as during the day.
The following tables show the percent of hours driven between midnight and 6 AM for different operational types.

Table 11
Estimated Night Percent

Table 12
Estimated Night Percent

Chapter 3
Baseline Safety Analysis

The objective of this proposal is to reduce the number of fatigue-related truck and motorcoach crashes.
The overall benefit will therefore depend on the current number of these crashes. This section discusses
the number of fatigue-related truck accidents, in three parts. First, we discuss existing estimates of the
number of fatigue-related truck crashes, and attempt to explain why these estimates differ. Second, we
use the Trucks Involved in Fatal Accidents database to generate estimates of fatigue-related truck
crashes by operation type, hours driving, and time of day. Finally, we present adjusted estimate of the
number of fatigue-related truck crashes by operation type.

1) Existing Crash Estimates

There are significant differences in published estimates of the number and proportion of fatigue-related
truck crashes. Much of the difference results from the differing analytical approaches used, particularly
differences in the set of crashes analyzed. Generally speaking, these studies can be divided into two
classes: those relying on large scale accident data tiles, and those based on more intensive analysis of a
smaller number of crashes.

The FHWA and the National Highway Traffic Safety Administration (NHTSA) have conducted several
fatigue studies using large scale data bases, primarily the Fatality Analysis Reporting System (FARS)
and the General Estimates System (GES). These databases, which are managed by NHTSA, are
based largely (but not exclusively) on police accident reports (PARS). Most police accident forms
contain a field for driver contributing factor, and among the choices are driver fatigue, drowsiness, or
asleep at the wheel. In most analyses, crashes in which one of these fields is checked are classified as
fatigue-related.

Crash analysts frequently criticize use of PARS for fatigue analysis, as they assert that PARs understate
the true extent of fatigue. There are a number of difficulties police face in determining whether fatigue
contributed to an accident. First, the responding officer’s primary concern is assisting accident victims
and restoring the flow of traffic. Investigating the causes of the accident is often a second (or lower)
level concern. Second, few police officers are trained in accident reconstruction, and they therefore do
not have the training to conduct a detailed investigation of the physical and mechanical evidence.
Therefore, many police officers must rely on eyewitness and other oral evidence.
This results in an additional problem. By the time an officer interviews surviving crash-involved drivers,
any signs of fatigue are likely to have worn off. The stress of the crash produces an adrenaline surge,
eliminating any traces of fatigue and in fact enhancing the drivers sense of alertness and awareness and
acuity, at least for the short term.

Some analysts have argued that in two-vehicle truck-involved fatal crashes, the investigating officer may
have to rely inordinately on the testimony of the truck driver, since the truck driver is five times more
likely to survive a fatal crash than the driver of the other vehicle. The truck driver has an obvious
incentive to minimize the role of fatigue in his actions, which could result in underreporting of fatigue
involvement. However, a report on two-vehicle truck-involved crashes does not find significant
evidence of a “survivors bias”. Dan Blower of the University of Michigan Transportation Research
Institute (UMTRI) examined almost 5,500 two-vehicle truck and passenger vehicle fatal crashes in
1994 and 1995. In 4,55 1 of these crashes, the passenger vehicle driver died while the truck driver
survived. In these case, 82% of passenger vehicle drivers were coded for contributing to the crash, as
opposed to only 24% of truck drivers. Truck drivers were the only fatality in 90 cases, and truck
drivers were coded in 58% of these cases, while passenger vehicle drivers were only cited in 47% of
these crashes. This suggests some sort of “survivors bias”. However, Blower notes that this explanation
is too simple, and he examined the larger number of fatal crashes where both drivers survive. The car
driver is assigned a factor in 74% of these cases, and the truck driver in 34.5%. The author notes that
“If driver survival explained the overall preponderance of driver factors for passenger vehicle drivers,
one would expect factors to be about equal where both survived”, which is not the case (Blower).

Smaller scale studies have a different set of problems, the most significant of which is generalizability.
While limiting the number of accidents studied can allow for more in-depth analysis of each specific
event, the results of these studies can not automatically be applied to all crashes. Thus, the results of
the National Transportation Safety Board’s (NTSB) 1995 study of single-vehicle large truck roadway
departure crashes where the truck driver survived may not be applicable to crashes which do not tit this
description.

Also, it is unclear what should be counted as a fatigue-related crash. Clearly all crashes where fatigue
is cited should be included, but there are other crashes where fatigue may play a less direct role.
Crashes involving inattention, distraction, or other driver failures may be related to fatigue, as a sizeable
literature demonstrates that fatigued individuals are prone to a variety of mental and physical errors.
Pilcher’s meta-analysis of 19 studies reveals that “sleep deprivation has a significant effect on human
functioning”, with cognitive performance subsiding more than physical performance. This supports
Brown, who argues that “the main effect of fatigue is a progressive withdrawal of attention from road
and traffic demands”. This suggests that in some cases, mental errors cited on a PAR may be the result
of fatigue. Dinges supports this logic, stating that “A loss of 10 percent in the detection of salient visual
stimuli (e.g., ‘slow speed’ signage) and a 10 percent increase in reaction time (e.g., stopping to avoid a
rear end collision) both of which can be demonstrated in even moderately sleepy persons (Dinges
1992) may contribute to many traffic and work accidents that are otherwise attributed to operator
inattention”.

Not only does fatigue demonstrably diminish individuals performance, but the type of errors made by
fatigued drivers are a major causal factor in crashes. For example, the classic Indiana Tri-Level Study
of the Causes of Traffic Accidents (Treat et al.), perhaps the most in-depth study ever performed in the
US on crash causation, found that “recognition failure” was involved in 56% of the crash cases
analyzed. While driver drowsiness/fatigue was found to be a certain or probable factor in only 2% of
the cases, 23% involved faulty visual surveillance, 15% involved inattention, and 13% involved
distraction. More recent studies have also found high levels of inattention and distraction. In a study of
nearly 700 Crashworthiness Data System (CDS) and GES crashes, Najm et al determined that
recognition errors were the primary causes of 45% of the cases studied, compared to 3.7% primarily
due to driver drowsiness (Najm et al). General Motor scientists reviewed over 1,000 PARs from
Michigan, and reported that 17% were attributable to “daydreaming” and 18% to improper lookout,
with just 1% due to “dozing” (Deering). While these studies were not limited to CMV crashes, they
demonstrate the prevalence of mental errors in crashes.

A recent study by the US Coast Guard also suggests that direct measurement of fatigue may understate
its true extent. Coast Guard researchers developed a “fatigue index”, based on the number of fatigue
symptoms reported, and the number of hours worked and slept in the 24 hours prior to the incident.
Using this formula upped the percentage of critical vessel cases categorized as fatigue related from 1.2
percent to 16 percent. For critical casualty cases, the fatigue index resulted in an adjustment from 1.3
to 33 percent. These reports indicate the need to be expansive in defining fatigue related incidents, and
the likelihood that measurements of fatigue based solely on accident reports are likely to underestimate
the extent of fatigue (US Coast Guard).

The FHWA recently completed an analysis of large truck crashes related primarily to driver fatigue, a
copy of which has been placed in the docket. This analysis reviewed existing studies of fatigue related
crashes, by a variety of characteristics, including vehicle body type and crash severity. Table 13,
reprinted from that study, shows that based on PARs, 1.98% of all fatal large-truck involved crashes
were clearly indicated to be fatigue-related. Fatal to truck occupant only (FTO) crashes were much
more likely to be fatigue-related, and crashes fatal to non-truck occupants were significantly less likely
to have fatigue cited as a related factor. For all severity levels, tractor-trailer combinations had a higher
rate of fatigue-related crashes than single-unit trucks.

Table 13
Large Truck Crashes and Percentages Associated with Truck Driver Fatigue, 1992-1997 Average Annual Crashes and Percentages