America Hasn’t Seen a Spike in Traffic Deaths This Bad in 50 Years

This graphic from NHTSA tells you how many people were killed in motor vehicle collisions last year, but says very little about the systemic causes of America's abysmal traffic safety record.
This graphic from NHTSA tells you how many people were killed in motor vehicle collisions last year, but says very little about the systemic causes of America's abysmal traffic safety record.

In 2016, 37,461 people were killed in motor vehicle crashes, according to official statistics recently released by U.S. DOT — a 5 percent increase over the previous year.

Coming on top of the 9 percent increase in 2015, that adds up to the worst two-year swing in traffic deaths in more than 50 years. Not since the early 1960s has the country seen such a spike. Safety is even getting worse according to federal officials’ preferred metric — deaths per mile driven rose 2.6 percent.

People walking or biking account for a rising share of total traffic deaths. Last year drivers killed nearly 6,000 pedestrians — an increase of 9 percent. The number of people killed while cycling rose slightly to 580 — still the highest toll since 1991.

Even before the current increase in the traffic fatality rate, America was falling far behind its international peers on street safety. But despite the preventable loss of tens of thousands of lives, the federal agencies that put out this update did not make any appeal for policy changes to turn this trend around.

The National Highway Traffic Safety Administration released the above graphic on Twitter. It provides a rough breakdown of the primary factors causing fatal crashes (but not for pedestrian and cyclist fatalities), but these statistics are of limited value. As Boston University professor Ital Verdi has written, focusing entirely on driver error glosses over systemic causes like dangerous street design and car-centric transportation systems.

Earlier this year, the National Transportation Safety Board made a breakthrough on this front with a major new report calling on state and local governments to reduce the prevalence of lethal speeding. That kind of message is completely absent from U.S. DOT’s by-the-numbers data release last week.

There is no call to action accompanying this news about the staggering death toll on America’s streets. No reflection on the country’s conventional traffic safety policies and how they have failed. There’s barely even an acknowledgment that things are getting worse.

If anything good can come out of this awful news, it’s a heightened awareness that our streets and transportation networks need to change. Federal transportation officials aren’t getting that basic message out.

72 thoughts on America Hasn’t Seen a Spike in Traffic Deaths This Bad in 50 Years

  1. this breakdown is very odd.
    can’t there be multiple causes for any fatal crash?
    and, for a crash to be fatal, how can excessive speed _not_ be a factor?

  2. Easy, the motorist is unbuckled, which can turn even a relatively low speed crash fatal.

    I’m kind of wondering about the speed related category; is it just a catch all for those that can’t be fit into other categories? Like let’s say someone is driving the same speed as everyone else, but above the speed limit (this is ubiquitous), and doesn’t shoulder check when switching lanes, merges into another motorist, and causes a fatal accident. Would this be considered speed related, and more importantly, should it be?

  3. What would really aid understanding is to combine deaths and permanently life-altering injuries. My guess is that there are at least as many of those.

  4. Is it really fair to call it an “accident” if the motorist doesn’t shoulder check when switching lanes? Some type of collision just sounds like the predictable result of that behavior.

  5. Call it what you want. You can call being for or against abortion pro life or pro choice all you want, it doesn’t change what it is. I don’t agree with the whole ‘molding language to fit our agenda’ thing, but you do you.

    I can argue that in common usage, “car crash” is overly general, as it includes both accidents and intentional acts. As in, someone deliberately ramming another car is a “car crash,” and so is someone hitting a patch of ice and slamming into the guard rail. But only the second one is an ‘accident.’ However, if terminology is more important than substance to you, you can call it what you wish.

    Any thoughts on the “speed-related” issue?

  6. Taking a dive into the DOT numbers, there’s some surprising news there:

    Distraction related fatalities are down 2.2%.
    Drowsy Driving fatalities are down 3.5%.
    Unrestrained fatalities are up 4.6%.

    So people are looking at their phones less and driving drowsy less, which is wonderful news, but for some reason reason some people are no longer buckling up. Why?

  7. hitting a patch of ice and slamming into a guardrail is a intentionally negligent act.

    The driver was driving recklessly for the freezing conditions.

    Hence a Crash

  8. US DOT values a life at $9,600,000. The total cost to our society from traffic related deaths. $359,625,600,000 In just one year!

    To put it in perspective if we were to divide that value by total 2016 gasoline and diesel sales it would be $1.77 per gallon! If we factored in all the externalities to driving very few Americans would be able to afford a car.

  9. Very easy for a crash to not be caused by excessive speed. Consider that maybe you’re driving on the highway at or under the speed limit and you hit a large animal which suddenly crossed the road. There was no excessive speed as you weren’t exceeding the speed limit. And yet you could easily have fatalities.

    I think we’re too obsessed with speed. This is distracting us from the true cause of most crashes, which is incompetent driving.

  10. Instead of looking at bandaids, perhaps it’s time to just acknowledge that car travel is dangerous compared to other modes. This in turn should spur us to a national effort to get as many people out of cars as possible. Even if we didn’t change the rate of fatalities, if we dropped the VMT by 90%, we would be under 4,000 dead annually. However, as we get more cars off the road, the fatality rate should drop also as there will be fewer vehicles in close proximity to collide with each other.

  11. I’ll tell you what would focus attention. A requirement that auto and gasoline companies cover the health insurance costs attributable to this, over and above the minimal auto liability insurance people are required to carry. Today health insurance carriers end up covering most of the bill.

    My daughter had a congenital hip problem that required repair. The health insurance company challenged the bill, for one reason. To see if some auto insurance policy ought to be liable for some of it. I wonder of health insurers are starting to do this more and more?

    A few years earlier, my brother in law was riding a bike and slammed by a car, and ended up in the hospital for a few weeks. Even though it was 100 percent the drivers’ fault, her auto insurance company refused to pay anything other that the near-zero, depreciated value of his bike, bike helmet, and glasses. After all, they claimed, he had vacation time and sick leave to use, and would be covered by his employer’s health insurance.

    Just imagine 100 percent of the health care costs associated with motor vehicles showing up in the price of buying or using motor vehicles, rather than in the cost of health insurance (including taxes for public health care programs such as Medicare and Medicaid).

    (And the health benefits of bicycle transportation showing up in health insurance-based discounts and subsidies as well).

    It could be massive — particularly if it were opened ended. I’ll bet my sister in law is still spending money on her knee as a result of a collision with a high drug addict who went over the centerline. Covered by health insurance.

  12. While ‘Unrestrained’ may be listed as the cause of ‘death’ it doesn’t explain the cause of the crash and at 28% that’s a significant percentage that could be divided up among various crash causes.

  13. Yes, you will notice the numbers don’t add up to 100%. One crash can involve a pedestrian, involve speeding, and be alcohol related. Unfortunately I know of one fatal crash in my town last year that involved all three.

  14. If I had to pick a reason why fewer people are wearing seat belts, I think I would choose the obesity epidemic. Seat belts can be horribly uncomfortable if you’re very fat. In fact, they might not even fit. End result is fewer people bucking up.

  15. Fair points here – especially about roadkill.
    I guess my point is this: it’s hard to imagine a crash in which death or injury still could not have been avoided if the driver had been driving more slowly.

    Here’s my longer rant:
    A road can be designed for a speed which is unsafe.
    It’s also plausible for a driver to exceed a safe speed without exceeding the speed limit. Take into account weather conditions, traffic, etc.

    And many speed limits are more about politics than safety.

    For example, in Pennsylvania, traffic engineers and streets departments do not have the authority to set any speed limit under 25 mph. On state-maintained streets, the default speed limit is 35mph, even through densely populated residential neighborhoods. Any change requires at least local political action, if not state political action.

  16. I’m guessing that they allow accidents to be listed under multiple causes. (The numbers add up to 126%.)

    What I’m wondering is how do they determine the cause of a death. Most cars are traveling above the posted speed limited, yet only 26% are listed as speed related. So there must be some criteria other than simply speeding at the time of the crash.

  17. You’re right, i didn’t even check.
    For the lazy, it’s about 125%

    My point, though, as I replied to Joe R, is that *any* fatal crash could have been avoided if the driver(s) had been going slower. In the extreme: you can fall asleep at the wheel without anything bad happening, provided the vehicle isn’t moving!

  18. I don’t want to come across as somebody who doesn’t think that we need to scale back how car-reliant we are as a culture, but I’m afraid the statistics are quite clear about walking being significantly more dangerous than driving.

    I mean, the thing that makes walking dangerous is that there _are_ people driving and that you’ll need to share the same surfaces at certain points, but you’re 1.5x more likely to die in a car accident as a pedestrian than you are as a person in a car.

  19. Different states treat comparative negligence differently, but in most cases when a drunk driver makes an illegal turn in front of a speeding car, the collision is going to be blamed on the drunk.

  20. They are measuring what is related, not the cause. It’s the number of collision fatalities that involved a pedestrian, not the number caused by pedestrians.

  21. Add to these numbers people who’s have died from diseases caused by pollution.

    And yet no one seems to take this threat to our safety seriously. Everyone is just going to shrug their shoulders and consider this part of the cost of driving. A truly staggering cost.

  22. You should also include the health care costs related to air pollution. Cancer treatments are hideously expensive, and many (most) cancers can be attributed to some type of environmental pollution. While we’re at it, let’s also add in the obesity caused by car dependence.

  23. I don’t see the relevance of your statement, less cars means less auto-pedestrian accidents.

    Whether or not it’s the driver who is injured, it’s the car part that is dangerous, I don’t think there’s much point in “well, actually”-ing this discussion.

  24. The relevance is that the statement made by Joe is simply not factually true. Car travel is statistically _not_ dangerous compared to the alternative options, so the rest of the statement made has no foundation.

    Again — I’m not arguing that we still shouldn’t do what we can to diminish drivership: I’m just saying the proposition made is not a case that will convince anybody but those of us who *already* agree that the outcome is desirable.

  25. Joe’s statement is incorrect if you obtusely read it as “you are safer if you walk than if you drive” instead of “cars are the greatest source of danger to pedestrians, cyclists, and other cars” or “we would all be safer if we all walked/biked instead of drove”.

    Of course you’re going to experience increased survivability when you get hit by a 2-ton metal cage if you’re in another 2-ton metal cage.

  26. It’s factually incorrect for you obtusely believe that I’m somehow your enemy here. I’m just saying that it’s not convincing to anybody but us when you you make a statement like “cars are dangerous, we should all walk or bike” when anybody who isn’t on our team is going to make a quick glance at simple and easily-available statistics to arrive at the the obvious and correct conclusion that they’ll in fact *increase* their own personal danger by making other transportation choices.

    If you want to make an actual change in peoples’ real-life behavior, you’re going to have to come up with a better argument. That’s all I’m saying.

  27. Adam, try not to get upset, I never said you were my enemy. Pull yourself together.

    Why did you misquote me when my actual quote was right there? I never said everyone should walk or bike, I just said we’d be safer if we did.

    Your goal here doesn’t seem to be to advance the discussion or strengthen the cause at all, just to nitpick and misinterpret and generally waste everyone’s time.

  28. Let’s just be clear here: what I’ve said is “this isn’t an argument that’s going to convince anybody”. Keeping in mind that the person that needs to be convinced about the dangers of our car-oriented society isn’t me, I can’t help but wonder what the point you’re trying to make is, here. Take a deep breath and read this statement a few times; maybe it’ll get through to you if I spell things out more explicitly: telling people to put their own lives at risk is a very ineffective way of convincing people to get out of their cars, and if we want to change this aspect of our culture, we’re going to have to make a better argument than that.

  29. For some reason you think we are arguing over a consumer-targeted messaging campaign designed to get people out of cars. (A thing no one has mentioned)

    When in reality this is going to be handled by street and transportation design, the viability of walking/driving/public transportation, cost viability of the different options, consumer needs, etc.

    The people designing streets or making laws regarding transportation don’t care about the safety of a single hypothetical person who switches from driving to walking/biking, but rather the overall safety metrics for the group as a whole.

  30. The point is relevant if you look at what might replace car travel. You’re NOT going to replace most car travel by either walking or biking. If you get cars out of the equation with mostly separate infrastructure it does indeed turn out walking or cycling are both statistically much safer (i.e. >90% of cyclist deaths were a result of collisions with motor vehicles). However, those modes largely aren’t what I was referring to. Walking is good for very short distances up to maybe a mile. Cycling is good up to perhaps 10 miles. However, many car trips are either longer than that, or there are other reasons preventing cycling or walking.

    So what would the replacements be? For short or medium distance trips you would have buses, streetcars, light rail, subways, and commuter rail. For longer distances you would have high-speed trains. Very long trips are typically done by air now, not car, so not much mode share to be gained there. Anyway, all of these modes are at least an order of magnitude safer than car travel, in some cases, like high-speed rail, several orders of magnitude safer.

    Getting those who now drive but can’t or won’t bike or walk into these modes will drastically reduce the number of cars on the road. This is turn WILL make walking and cycling much safer than they currently are now. If you think about it, how many people die walking solo? Maybe the small number who trip and hit their head the wrong way. The number is tiny compared to the numbers killed by motor vehicles.

    I’d love to hazard a guess at numbers once we do all this but that’s hard. I’d say it’s highly likely we’ll go from ~40K transportation related deaths (most of those are due to cars) to under 1K. The real issue here isn’t just that cars themselves are dangerous to those who use them. It’s also the fact they make some other modes much more dangerous than they would otherwise be.

  31. Strictly speaking, if everyone drove 5 or 10 mph all the time almost nobody would ever be killed by motor vehicles. The reality is we’ve chosen a balance between safety and speed. In some cases, particularly residential streets, that balance has often shifted too far in the direction of speed.

  32. True – The safety and efficiency gains will not happen until all cars are autonomous.

    The switch to autonomous cars will be similar to the switch between analog to digital television. Possibly led by California legislation, a date will be set that all cars need to be autonomous. To ease the transition to retrofit older cars, vouchers will be distributed (remember the TV conversion box vouchers from 2009). Older cars may get a “Classic” designation. A limited use permit or a requirement to have a transponder that communicates to autonomous cars that it is human operated. Drivers’ Licenses will truly become a “Privilege” with more difficult testing.

    What may delay the full roll-out will be the infrastructure. To realize the full potential of autonomous vehicles, traffic/intersection management systems will need to be installed. Cars will have to communicate with a traffic control system. Right now it takes an hour to drive 10 miles during rush hour in the Los Angeles Westside. Even if the car travels at a constant 20 mph, it will cut the time down to 30 minutes. Vehicles could be grouped as such that there are one to two minute gaps for pedestrian crossing cycles (scramble crosswalks).

    Pedestrian and cyclist fatalities will be similar to fatalities along railways. Vehicles will be programmed to operate 100% under the rules and laws of the road. But the vehicles will rely on all road users to follow the rules and laws. There may be increases in fatalities for jaywalkers and wrong-way cyclists, but near-zero fatalities for VC cyclists, cyclists using designated infrastructure, and pedestrians during walk cycles at intersections. Similar to when a motorist or pedestrian goes around a railway crossing gate and dies; will it be victim blaming or Darwinian natural selection?

  33. Wouldn’t it make more sense to just give pedestrians and cyclists the right-of-way all the time? Basically what you’re saying is if they don’t follow the rules, vehicles will actually be programmed to run them down. Besides not being possible (it violates the first law of robotics), can you imagine the public outrage? The fact is pedestrians and cyclists don’t reliably follow traffic controls now, and they’re not going to magically start once we have AVs. It also makes cities less livable when you have to wait to cross streets. AVs could be a real boon for cities if they make cycling and walking both safer and faster. They could be a disaster if we basically turn every street into a highway where people can only cross at certain places or times. I hope we go the former route.

    The way I think it should work is vehicles negotiate with each other at intersections (hence no need for traffic lights and probably no need to stop). If pedestrians or cyclists are crossing, the vehicles yield to them. If it turns out a particular intersection has such heavy pedestrian traffic that vehicles can’t get through, then they get rerouted to less busy intersections. Obviously you also reroute any vehicles on the way to the intersection which haven’t yet reached it.

    I suspect on limited access highways it will work more like how you say. Since pedestrians aren’t allowed on the highways, the vehicles will likely be traveling at speeds where yielding is often impossible. Doubtless the vehicles would still try to avoid hitting any trespassers, but they might not always succeed.

  34. A good point in time to realize that the breakdown in data beyond the binary ‘dead’ is to be taken with a grain of salt. There is no way in God’s name that driver distraction is now playing a lesser role in deaths. No way. Just as when you dig down into individual police reports about crashes and see just how wrong they are, it is more likely that the police investigators are simply raising their ‘distracted’ threshold to match the baseline distraction in drivers – which has skyrocketed.

  35. “Car travel is statistically _not_ dangerous compared to the alternative options,”

    Really? It might be safer if you look only at death rate for the people traveling, but if you look at total deaths caused (including to third parties) per mile traveled, how does it look?

  36. No, it’s not. I hit a patch of black ice years ago. It was days after conditions had cleared, but melting snow during the day had refrozen where it ran across the road at an intersection. Impossible to know it is there when you can’t see it and the rest of the road conditions were dry and clear.

    That isn’t negligent.

  37. Yes, it’s estimated that there are 200,000 premature deaths a year in the U.S. due to air pollution. The statistic doesn’t elicit a big emotional response though, because the deaths are not sudden. But they are very real. So is the massive healthcare cost and lost productivity caused by all that air pollution.

  38. of course it’s negligent – temps were below 32 during the night.

    You choose to drive
    You choose to drive too fast for conditions
    You didn’t train how to operate your machine safely in case you did hit black ice

    It’s a crash not a accident

  39. I’m an interesting test subject, if I don’t mind saying so. I’ve been car-less for over 13 years and I have just recently joined the driving public. What I’ve notice, that’s quite glaring, is drivers aren’t paying attention when they’re in the turn lane for turning left. Many, many, MANY, times drivers have to honk the horn to get drivers attention to go. The inattentive drivers make it through the turning process. Who ends up paying the price of the driver’s inattentiveness is the drivers further back in the turning lane. It used to be that quite a few car got through the turning process. Now, not so much.

    Hmmmmmmmm wonder what the drivers at the front of the line are doing that has them distracted? Hmmmmmmmmmm………..

  40. Regarding the “driver error” category, driver’s education basically disappeared from schools around 1995 due to funding cuts, and most Departments of Motor Vehicles in most states will give an intoxicated monkey a driver’s license, and *none* of them will revoke a driver’s license from someone with a record of reckless driving.

    Imagine if we managed commercial pilot’s licenses this way.

    “Driver error” is not an excuse. What it means is “we don’t have any standards for drivers licenses”.

    I have seen worse and worse and worse driving on the roads in recent years, and it’s quite clear there’s no effort being made to regulate it at all.

  41. I agree wholeheartedly. In many other countries the licensing exams are quite rigorous. There are in depth written exams, skills tests on closed tracks, lengthy road tests in mixed city and highway traffic, and you must demonstrate extensive knowledge of the performance and maintenance needs of your registered vehicle. They also require you be periodically re-evaluated. In this country you get 12/20 on a road sign picture test, drive for 5 minutes on a side street, and you are certified for life. When will the state licensing agencies start taking more responsibility for the carnage all the people they license are causing? In most US states you can get a 5-10% discount on insurance premiums for taking a defensive driving course every 3 years, precisely because it makes you statistically less likely to be involved in a serious crash. Just as a starting point, why wouldn’t something like that simply be required for all drivers every 3 years? That would probably save thousands of lives each year.

  42. And that’s really how it should be. Back in the old days, cars were toys of the rich. It would be nice to go back to that. At least the rich back then who could afford cars took some pride in their driving skills. The idea of mass motoring belongs on the top of the dustbin of history’s awful ideas.

  43. I don’t put much stock in their distracted driving figures. I think it only includes instances where distraction was specifically listed in the police report. And nobody admits they were distracted. And it’s difficult to prove and likely vastly underestimated

Leave a Reply

Your email address will not be published. Required fields are marked *



America’s Traffic Death Toll Is a National Disgrace

More than 40,000 Americans were killed in traffic last year, according to new estimates from the National Safety Council, the worst toll in a decade. The U.S. transportation system claims far more lives each year than peer countries. If America achieved the same fatality rate as the UK, more than 30,000 lives would be saved each year.

It’s Official: 33,561 People Killed in Traffic on American Streets Last Year

The official 2012 death toll is out for our nation’s poorly-designed, auto-centric transportation system. According to the National Highway Traffic Safety Administration, traffic injuries on the nation’s roadways claimed the lives of 33,561 people. The headline of the agency’s press release, “NHTSA Data Confirms Traffic Fatalities Increased In 2012,” is quickly walked back by the subhed, […]

U.S. Traffic Fatalities Rising Fast — Especially Pedestrian and Cyclist Deaths

Traffic fatalities in America hit a seven-year high in 2015, with pedestrians and cyclists accounting for a disproportionate share of the alarming increase, according to preliminary data from the National Highway Traffic Safety Administration. Last year, 35,200 people were killed in traffic — a 7.7 percent increase over 2014 and the worst death toll since 2008. The number of people killed while […]

America Has a Terrible Traffic Safety Record Because We Drive Too Much

Even though the U.S. traffic fatality rate per mile driven has fallen by two-thirds in the last 50 years, America today still has the deadliest road system per capita in the developed world. Much of the improvement from safer driving and better emergency care has been wiped out by increases in total traffic. The American approach to traffic safety has emphasized seatbelt use, vehicle standards, […]