DETROIT, Mich. - Driver pay has a strong influence on road safety, according to Michael Belzer from Wayne State University in Detroit. The higher the pay, the lower the risk of a crash.That's an over-...
DETROIT, Mich. – Driver pay has a strong influence on road safety, according to Michael Belzer from Wayne State University in Detroit. The higher the pay, the lower the risk of a crash.
That’s an over-simplification of a complex report Belzer and two co-authors, Daniel Rodriguez from the University of North Carolina and Stanley Sedo from the University of Michigan, submitted to the U.S. Federal Motor Carrier Safety Administration (FMCSA) in September.
But, whether or not it’s an over-simplification, it captures the thrust of their findings.
“Truck driver pay is an extremely strong predictor of driver safety,” according to the three professors.
The conclusions are based on three separate studies in the report.
The most compelling is an examination of a decision made by J.B. Hunt in October 1996.
This large, truckload carrier decided to increase driver’s mileage pay by over one-third, which is one heck of a pay raise.
According to Belzer, “Hunt’s goal was to increase both safety and productivity of drivers by paying a wage somewhat higher than that necessary to attract drivers at the margin.
The new wage allowed them both to identify and hire drivers with the characteristics they desired.
It also allowed them to terminate drivers who did not meet either safety or productivity standards.”
And what were the results of J.B. Hunt’s decision?
According to Belzer, in the 13 months prior to the decision being made (September, 1995-to-September, 1996) J.B. Hunt’s drivers were involved in 1,535 crashes. “Crash,” in this case, is any incident causing $200 or more in damage, an injury or a fatality.
In the 13 months after implementation (February 1997), crashes fell to 783-a drop of 50 per cent. This reduction in crashes allowed J.B. Hunt to recoup at least 80 per cent of the cost of the pay raise.
Belzer’s analysis of the J.B. Hunt case is based on 11,540 individual drivers, all of whom worked for Hunt for at least one month during the 26-month period.
The average pay for the first 13-month period was 26.2 cents per mile; in the second 13-month period, it was 33 cents per mile. Other things, like a safety bonus were also added after February 1997.
As Belzer’s analysis shows, every penny increase in the average rate of pay for the two periods (30.27 cents) reduces a driver’s probability of having a crash by 11 per cent.
This percentage reduction is only valid at the mean pay rate of 30.27 cents per mile-the higher the wage, the smaller the impact on safety of a one penny increase.
For example, at some very high mileage rate, an extra penny has no impact on safety.
Turning this around, the impact of a raise on safety seems to be larger the lower the initial rate of pay.
To speculate, it seems that the lowest paid drivers have the greatest incentive to take short cuts like incomplete trip inspections, under-reporting time on logs, speeding, working long hours, and so on.
Giving these drivers a raise has the biggest impact on reducing these incentives.
Other findings from the J.B. Hunt study include the fact that a driver’s risk of having a crash decreases with age, to 41 years old, then starts climbing again.
Young and old drivers have something in common-their risk of crashing.
Also, drivers who are not married have a lower risk of an accident than married drivers.
This comes as a surprise as conventional wisdom suggests married drivers are more stable and, hence, safer, than unmarried ones.
Another surprise is that driving in winter months reduces the probability of a crash by 13 per cent.
These findings may be unique to J.B. Hunt, so they aren’t necessarily transferable to the industry at large (or to Canadian winter driving).
The second study collected information from 102 truckload carriers in the states and then, through elaborate procedures, tested to see if there were valid statistical relationships between the number of crashes a company experienced and the way it paid its drivers.
The crash data is from the U.S. DoT’s “MCMIS” crash data file, which records only those accidents serious enough to result in a vehicle being towed away, an injury or a death.
The average carrier in the sample of 102 had 64 DoT-recordable crashes in 1998.
These are large firms, averaging 127.5 million miles a year, so the 64 crashes aren’t as many as they appear to be (roughly one accident every two million miles).
Pay rates were obtained from a private sector company (Signpost) that surveys trucking companies quarterly.
Belzer and his team used pay rates in the fourth quarter of 1998 and the crash file for the whole year.
The term “pay” in this study is actually a long series of variables on such things as the base rate (average 28.6 cents per mile), unpaid time (3.62 hours per trip), the rate of increase in the mileage pay (0.7 cents per mile per year), safety bonus (49 per cent of the drivers), health insurance ($166.84 of driver-contributed payments per year), and life insurance (the average driver had insurance worth over $15,000).
The results of the second study suggest that for every 10 per cent more in average compensation (mileage rate, unpaid time, anticipated annual raise, safety bonus, health insurance, and life insurance), a carrier will experience 9.2 per cent fewer crashes.
And the biggest part of this comes from the base mileage rate – a 10 per cent increase in it, drops the number of crashes by 5.2 per cent.
The third study is based on a 1997/1998 driver survey conducted by the University of Michigan at 19 truck stops in the U.S. Midwest.
Information on pay and working conditions was collected from over 1,000 drivers.
However, in Belzer’s work, only 247 surveys are used as all but the mileage-paid driver employees were excluded.
The results from this third study are not as robust as the first two. As the authors note, there is “noise” in the data.
Notwithstanding the noise, the results again confirm the relationship between driver pay and safety.
For all 247 drivers, 13.8 per cent reported a crash in the past year.
The average pay was 29.5 cents per mile.
For every 10 per cent increase above this, the probability of a crash fell by 21 per cent.
Two other findings from this third study are interesting. Drivers in large companies (more than 100 drivers) had a lower probability of a crash than drivers from small companies.
And, for every 10 per cent increase in paid days off (holidays, vacation) the probability of a crash fell by seven per cent.
The average driver in the sample had 14.7 paid days off a year.
Belzer and his co-authors are cautious about suggesting what ought to be done with their findings.
The report to the FMCSA contains no recommendations for government policy.
But, when questioned about this, Belzer suggests there are a number of things that could be done:
Governments could re-regulate the industry thereby making it harder for new firms to enter the business with low-paid drivers and cut-throat rates.
Enforcement agencies could use stiffer sanctions for infractions of safety regulations.
The insurance industry could improve its ability to assess risk thereby increasing costs for carriers who employ low-paid (unsafe) drivers.
A minimum wage for drivers could be set.
More favorable legislation could be passed making it easier for drivers to unionize (higher union rates result in safer drivers).
Belzer says all of these are possibilities.
He won’t say which, if any, is his preferred option.