self-driving cars
won’t save us just yet


Roadway deaths are
on the rise again, AND


There are two things that drive traffic fatalities — risk and exposure. Risk is the chance you’ll be killed per mile of driving, and exposure is how much you drive.


 We now know which factors helped to create that dramatic


drop in highway deaths from 2008 to 2011, and we also know the degree to which each factor contributed.



When you take a lot of the riskiest drivers off the road, you have fewer crashes and fewer fatalities from those crashes.



of vehicles are equipped with automatic braking.





For More Information

Robert Wunderlich


Roadway crash deaths, after a brief but sharp drop several years ago, are on the rise again. But equally troubling is the likelihood that this trend may continue.


There are two things that drive traffic fatalities — risk and exposure. Risk is the chance you’ll be killed per mile of driving, and exposure is how much you drive. The measurement for exposure, vehicle miles traveled, increases steadily along with a growing population and a healthy economy. As that measure increases, absent any dramatic steps to reduce risk, we can also expect crashes and deaths to increase.


Before we jump on the self-driving bandwagon, let’s just remember that automated vehicles — while they do offer the promise of safer daily travel — may not constitute a universal cure for the highway fatality epidemic, and certainly not anytime soon.


That’s in large part because self-driving cars, to make a profound contribution to safety, need to constitute a dominant presence on the roads. But as a general rule, lots of people like to hang onto their cars for a long time, beyond a decade on average. The number of automated vehicles on the road will grow slowly in relation to conventional autos, so we’re still many years away from the day when self-driving vehicles represent the majority of cars and trucks on our roadways. We’re that many years out from the time when drivers really don’t do much driving at all, and consequently make far fewer crash-causing mistakes. And even then, we still won't be at the automated point.



If that’s how we characterize the vehicle operators of the future — drivers who don’t really do much driving — then it’s fair to say that a large percentage of the drivers of tomorrow have yet to learn to ride a bicycle, and another big share of them haven’t even been born. For those of us who have been driving for a while, we should plan to vote in at least three or four more presidential elections before we can expect our new self-driving cars to transport us safely to the polls.


Until that day comes, there’s still a lot we can do because we know more about what drives the overall level of fatalities based on recent research by the Texas A&M Transportation Institute and the University of Michigan Transportation Research Institute. We now know which factors helped to create that dramatic 21 percent drop in highway deaths from 2008 to 2011, and we also know the degree to which each factor contributed. Taking a close look at those factors can offer guidance for how we approach safety in the next 20 years, and it can also remind us of the limitations we face.


Economic Factors


Economic Factors

We can attribute about 85 percent of the credit for the 2006–2011 decrease to economic forces. This should come as no surprise to those of us who remember the historic recession at that time. The U.S. unemployment rate rose as high as almost 10 percent in those years, but the jobless rate was double that for 16- to 24-year-olds. Consequently, it appears that this high-risk group was driving less for work reasons, and because they had less money to spend, they were likely also driving less for discretionary or leisure reasons.


The models we created in our research demonstrate that for every 5 percent increase in the unemployment rate for young adults, we can expect a similar percentage decrease in roadway deaths.


So, what’s the challenge going forward? As effective as the economic meltdown was at reducing crash fatalities, we shouldn’t consider a recession to be a safety strategy. It did, however, reinforce an important point: When you take a lot of the riskiest drivers off the road, you have fewer crashes and fewer fatalities from those crashes. The question then becomes, how do you do that in the absence of economic calamity? It is clear that we need to target this group to help improve their driving skills and reduce their risk-taking.


Vehicle Safety Improvements


Vehicle Safety Improvements

The generally improved safety of the cars we drive accounts for about 12 percent of the reason why traffic deaths declined so much from 2006 to 2011. Those improvements consist of things like more crashworthy designs, anti-lock brakes, electronic stability control and more and better airbags — standard features in an increasingly large share of cars now on roads that were built in 1991 or later.


So what’s the challenge going forward? The auto industry has been making meaningful and often dramatic improvements in vehicle safety since that work started in the 1960s. That begs the question of whether we’ve reached a point of diminishing returns. Have we run out of potential design benefits, and will we have to wait for those self-driving cars before we see another big safety step? There are very effective technologies such as automatic braking available now, but only 2 percent of vehicles are equipped with it today. It would take rapid adoption of this and many more advanced safety features to make a difference in the near term.


Regulatory Policies


Regulatory Policies

DUI laws and other roadway regulations account for the remaining 2 to 3 percent of the credit for a decline in crash deaths. Predictably, stricter laws contribute to a higher degree. But many of the DUI, safety belt and helmet laws have either already been adopted or haven’t changed much recently.


So what’s the challenge going forward? It’s fair to ask whether people are nearing the limits of how much they’re willing to be regulated. Even if we pass more laws, the presence of those regulations doesn’t ensure compliance. The daytime safety belt use rate in Texas, for example, is currently better than 90 percent; even so, half the Texans who die in crashes aren’t wearing safety belts. Two states, California and Washington, have driven seat belt use higher through year-round and extensive enforcement.





Infrastructure Spending

Infrastructure investment can make a difference too, but it’s generally a more modest difference in the bigger picture because those improvements only decrease crash risk for the relatively small number of people who travel where the improvements are made. That’s not to say they’re not important, as we learned in Texas with the establishment of a safety bond initiative in 2003. That policy dictated that 20 percent of the funding from voter-approved highway bonds would be spent specifically on safety-related construction. Our analysis showed that adding shoulders on narrow roads alone is preventing about 300 crash injuries and saving more than 40 lives every year since the improvements were made. So, it’s not just about how much you spend; even more important is where you spend it. Research has given us much better knowledge about what works, and to what degree, than we have ever had before.


That brings us to a central point: spending (or rather, investing). We’re talking about a very complex problem, but in this case, the solution is relatively simple. Increased investment in highway safety engineering along with better driver behavior education and enforcement can help slow the growth in roadway deaths, and maybe even reverse it. Our research, in fact, suggests that a $2.7 billion investment in roadway safety in Texas over 5 years would save as many as 4,600 lives and produce a return of about $32.5 billion. That’s $12 in benefits for every $1 spent.


Absent any sweeping policy changes (like raising the legal driving age and/or making current laws even stronger), more investment in projects and programs is absolutely essential to implement potential solutions to the problem.


Automated travel does hold considerable promise for safer roadways; no argument there. The potential is even greater with connected travel (in which cars are able to transmit safety signals between each other and the roadway). But we can’t pin our hopes on a solution we won’t see in widespread use anytime soon. While we’re waiting to enjoy the transportation dream we have for tomorrow, we need to do a lot more with the safety strategies we already have today.