By on April 20, 2011

Edmunds’ Jeremy Anwyl asks:

The chart [above] shows the rate of fatalities per 100 million miles driven. We have assembled the data, going back almost 100 years. Look at the chart closely. Can you see a drop in fatalities that corresponds with when seatbelts were first introduced? Or when legislation mandating their use was passed? Or what about when air bags become prevalent? What about a jump in fatalities that ties to the current “epidemic” in texting while driving?

I can’t. The data does show that fatalities dropped markedly during the Great Depression and WW II. Aside from that, the rate has been declining  steadily for years. Decades, even. This is good news, but makes it hard to credit seatbelts, technology or the other factors that reflexively are given credit. I am not suggesting that we should all stop wearing seatbelts. I am suggesting that when thinking about transportation safety, there is more going on than we typically consider.

Get the latest TTAC e-Newsletter!

54 Comments on “Chart Of The Day: Why Have Road Fatalities Declined?...”

  • avatar

    more proof i think we can safely raise speed limits

  • avatar

    I would assume it had to do more with the modern interstate system; which if I recall correctly began after WW2. This allowed more people to drive many more miles with a lot less interaction or area for mistakes than smaller roads and city driving.

    • 0 avatar
      Robert Schwartz

      The Interstate Highway System was authorized by the Federal-Aid Highway Act of 1956 – popularly known as the National Interstate and Defense Highways Act of 1956 – on June 29.–
      The steepest part of the decline on the chart occurred before 1956.

    • 0 avatar

      I think road and interstate design is a big factor. Better roadways, but also better tires, safety glass, better auto handling and breaking. Cultural use of the vehicles also plays into it I think. There are many factors at play.

      • 0 avatar

        You’re not going to have sudden declines because new developments–seat belts, air bags, etc.–penetrate the fleet only slowly, and because it took many years to build the interstate system. On the other hand, people are always going to drive a bit harder when they perceive that they have more safety.
        Regarding the difference in rate of decrease in fatalities in the first half vs the second half of the last century, it’s important to remember the law of diminishing returns. I think it’s damned impressive that we’ve reduced the fatality rate to one fifth what it was 50 years ago.

  • avatar

    In Finland, there were a record of 1200 fatalities in 1973, when seat belts (and using them) became compulsory, as well as using daytime running lights and speed limits became mandatory. It’s been a steady drop since then, with only about 200-300, fatalities per year nowadays, and three times the traffic than in the 1970s.
    I’m pretty sure all that legislation had something to do with the decline of traffic deaths.

  • avatar

    Since it is per million miles driven then there must be other variables that are more important.  For instance, percentage of miles driven on the highway vs. city.  Since highway travel on a per mile basis must be much safer than city travel, it is likely that a significant amount of the early reduction has more to do with what type of driving was done and distances traveled by car.

  • avatar

    Shoddy analysis.

    Including the first few decades of the century makes more recent gains appear smaller than they actually are. Look closely at the graph and you’ll see that fatalities today are about a quarter what they were 40 years ago, before most of the safety regulations.

    Also remember that safety features only gradually spread through the entire population, as they are fitted to new cars. So figure 6-8 years before any new mandated feature is on nearly all cars. 

    Even seat belt use, while it was mandated at one point, gradually increased over the years. I still remember when many people refused to wear seat belts. Nearly everyone wears them today. This change took decades.

    Ditto drinking while driving. My sense is that this is much less common than it used to be.

    Texting while driving? How much of the population does this? And would this percentage be likely to increase if nothing were done to prevent it?

    • 0 avatar

      Such a graph needs a logarithmic scale to be useful. But statistics and lies are cousins for a reason.

    • 0 avatar

      Michael – you beat me to it – this chart shouldn’t start until post WWII as it would cover the introduction of safety technologies + improved roads i.e. interstates

    • 0 avatar

      The change in attitude regarding both seat belts and drunk driving has been huge during my lifetime. When I started driving in 1978, no one wore safety belts. Now, virtually everyone I know does. Even my parents, who started driving in the 1950s, and didn’t wear seat belts when I was growing up, won’t start the car unless everyone is wearing his or her safety belts.

      When I started college in 1980, drunk driving was still winked at by many people. Today it is taken very seriously. This goes hand-in-hand with different attitudes regarding alcohol abuse. We can see this in the two versions of the movie Arthur. In the original, which came out in late 1981, the title character’s heavy drinking and inebriated state throughout the entire movie was played for laughs. In this year’s remake, Arthur is in rehab, because most of us would not find it amusing that someone couldn’t get through the day without drinking heavily.

    • 0 avatar

      Well put. I think it’s also important to note that many of the safety features existed several years before becoming mandated; airbags, for instance, started making their way in to cars in the US in the mid-70’s, became more common by the mid-80’s but weren’t required to be in cars until the end of the 80’s, and weren’t required to be in light trucks until the mid-90’s. At the time that airbags were required, light trucks were more than 40% of the auto market.

    • 0 avatar

      “Also remember that safety features only gradually spread through the entire population, as they are fitted to new cars. So figure 6-8 years before any new mandated feature is on nearly all cars.”
      I scanned through the comments to see if anyone had posted this. If there are – say – 200 million cars on the road and a mandated safety feature is installed on every one of 10 million new cars sold in a particular year, the contribution of of that feature to overall safety will be initially quite small, but will grow over time.
      Increases in medical care have made severe injuries survivable, which of course will then not show up on a chart of fatalities. This is the same reason why fatalities in wars over time have to be closely examined.

      • 0 avatar

        Bingo! I was waiting to see if anyone else would mention this. For many it is a case of not being able to see the wood for the trees.
        Better medical care is by far the biggest factor in this for sure. That is NOT to say that all the other technical advances don’t also help, surely they must.

      • 0 avatar

        But I recall reading that, over the last 10-15 years, the number of accidents has declined, too. So something must be helping people to avoid having the accident in the first place.

    • 0 avatar

      Pretty much exactly my analysis Mike.

      I will add that the seat belt interlock mearly resulted in a lot of folks connecting the the belt and siting on it.  Mandated and enforced are two different things also.

      As kids we would crawl all over the car, can’t imagine that now.  Loved station wagons, more room to  roam.

    • 0 avatar

      Many excellent points made about gradual improvements in highways, vehicle quality and safety equipment, and the speed and effectiveness of medical care, but one statistical factor is also overlooked: the time spent in cars. As driving speed increases (without addl risk thanks to better cars and roads), miles traveled increase without a matching increase in time in the vehicle, and you can’t have vehicle fatalities if you’re not in the vehicle. I’ve seen this point made in comparisons of safety vis-a-vis mode of travel. Yes, air travel is safer than bus travel as measured by injuries per passenger mile, but passengers spend so much more time in a bus than in a plane to travel the same distance that the opportunity for injuries goes up dramatically.

  • avatar

    1. Gradual reduction of road hazards.
    2. Gradual adoption of upgrades such as air bags, seat belts, crash structures, ABS, stability control, etc. etc.
    3. Gradual and substantial improvements in acute health care.
    As for texting etc., I see no evidence, including texting or not, that distracted driving is increasing or decreasing.
    The next wave of safety improvements will be proximity control systems and systems to detect inattentive driving.

  • avatar

    It’s mostly due to a steady increase in the quality and availability of emergency medical care.
    If you look at the homicide rate in Baltimore you’ll notice a sudden drop that doesn’t correspond with any new police action or economic situation, but rather with the opening of the Cowley Shock Trauma Center. Same story.

    • 0 avatar

      The US was the only country in the West that didn’t see a significant drop in fatalities in the last 10 years.
      Hurrah for nationalized health care

      • 0 avatar

        At the scene of the accident, first responders do not ask victims whether they have insurance, so the availability of nationalized health care, or lack thereof, has nothing to do with it.

  • avatar

    You probably won’t see a big drop when air bags or seat belts were introduced because, in real terms, those things were phased in over a number of years.  In other words, not every car on the road suddenly had seat belts or airbags (or collapsible steering columns, or crunch zones, or whatever).  New cars had them, but not existing cars.  Compare when these things first came out to about 10 years in the future for meaningful numbers.

  • avatar

    I’d like to see a graph of average speed traveled charted for each year.  I bet average speed traveled has increased fairly steadily over time.  Then we could draw the (false) conclusion that increasing the average speed of travel has improved safety.  Yeah!

  • avatar

    More seatbelt use, less tolerance for drinking and driving, safer cars.

  • avatar

    Better EMTs, flight for life helicopters and the prevalence of mobile phones help as well. Previously, a “fatality” was anyone who died within a year after the accident. Now, the death time-frame as been reduced to a month or two. The data criteria and analysis has changed so much over the past 60 years to make any insight meaningless.

  • avatar

    I find the plateau in the 60s interesting.  Deaths begin increasing in 1960, then decline in 66.  Front seatbelts became standard in 1963, given the much shorter lifespan cars enjoyed back then I expect  that by 1966 a large percentage of the cars on the road had seat belts.

    • 0 avatar

      Demographics. 1966 was close to the peak of the baby boom generation turning 16. The decline after that may be partly credited to the improved safety devices mandated in 1968. The dip in the early 1960’s corresponds to a fairly deep recession which may have curtailed trucking. There are so many possible causes outside of just laws and technology. A similar graph exists for air travel. The steepest drop was about 1961 as jets rapidly replaced the props in the fleet.

  • avatar

    Ed – the reason why you don’t see a sudden drop in road fatalities is that, while technology does reduce fatalities, it is introduced slowly into the national fleet. For example stability control: while most new cars now have it, the majority of older cars on the road don’t and this smoothes out the positive effect on fatalities and spreads it over 10-20 years.
    I don’t know about the texting effect but I would guess that it is only yet another incremental increase in the ocean of distracted driving and thus doesn’t cause a big bump. The other possible reason is that people generally only text in slower traffic and thus it may be causing more crashes but not necessarily more fatalities.

  • avatar

    But why the big decline from about 1933 to 1960? Well that isn’t due to any single automotive technology when I look at the data. Remember, this chart shows fatalities, not injuries, and not having an injury line leaves out a vital part of the picture. Adavancement in medical technology.

    The first blood bank? Established in 1937. Rh type cross matching? Perfected in 1940. Battlefield medicine innovations in wound care, burn treatment, and infection prevention? Giant advancements made from 1941 to 1945. The first national network of Blood Banks? 1950. A good understanding of how to break down the components of blood to deal with hemmoraging? From 1954 to 1960.

    Pencillin medical trials started in 1942. The mass production of pencillin to deal with post accident infections? Perfected in 1945. Look at the huge decline from ’42 to ’45 as penicillin became available.

    The lessons of World War II generated the idea of aid stations and battlefield hospitals, patient mortality on the battlefield during the Korean War plummeted compared to World War II with the use of Aid Stations, MASH units, and having ambulances operating as mobile hosptials instead of for pure transport. The techniques learned of patient transport and advanced treatment before getting to a hospital was adopted in the civilian space starting around 1954.

    Paramedic programs and using helicopters to transport critically injured patients started rolling out in the mid and late 1970’s, interestingly the public demand for these kind of services was heavily fueled by the TV show “Emergency” which went to great lengths to be authentic in the implementation of emergency medicine. Emergency Medicine as a specific discipline, started up in 1979!

    If you look at the dramatic decline in fatality rates from 1942 to 1960 in particular, I would argue that decline was not fueled by a specific introduction of an automotive safety technology, but due to the rapid advancement of emergency medicine fueled by the lessons learned in World War II, our understanding of blood typing, unlocking the secrets of penicillin, the Axis Powers horrors of medical experimentation (which I am NOT in any way condoning, it was horrible, and it was a travesty that the Japanese monsters basically got a free pass compared to their German counterparts because they cooperated with the US government on providing the data on what they learned) and the lessons of advanced medical treatment from the Korean War that had the biggest impact on fatality levels.

    I would suspect that if an injury line were shown, it would show a different trend, and could potentially show reductions hand-in-hand with the introduction of a couple of key automotive technologies, specifically seat belts and air bags. But, that is my pure speculation.

    I know Ray LaHood wants me to believe it is Red Aslphalt and Blood Runs Red On the Highway out there, but the fatality and accident rate data doesn’t prove it out. There is a mountain of evidence that points to the opposite, and that speed limits could be increased, and in some states (hello, California) the results would probably be even lower fatalities (increased speed within reason on a well designed interstate equals better concentration and focus)

    Actually, the data doesn’t even support my personal view that we need better driving training. It actually shows we have darn good drivers.

    • 0 avatar

      It’s comments like this that makes this site great. I’ll add that changing the ambulances from modified hearses and Suburbans (you call, we haul, that’s all) to custom bodied Econolines with room in the back for performing CPR and hauling as many medical supplies as possible, saved many lives.
      One extremely minor quibble: “Emergency!” authentic? Not so much. Back when it was first on, my father, who helped found our town’s rescue squad, would watch it just to make fun of it! “Johnny, you forgot the neck brace!” “Get that guy on a backboard BEFORE you move him!” “Block that Torino or it’ll fall on you!”

  • avatar

    The thing that chart doesn’t deal with is hours driven per million miles.  It would seem to me its not a question of miles, but time on the road per death that is a better metric for determining the fatality level.   It’s more a question of how much time there is for something to wrong rather than speed.  Since the average speed traveled probably went up as cars got better, safer, and faster, and more people drove more as cars got better, you could easily lose the really interesting data by looking at the wrong way.

  • avatar

    I was going to write a long treatise as to why the banner graph presents a simplistic, and perhaps not even the best or a representative trend-line for discussing a complex topic, but there have been some very excellent analyses esp. by Holdenssvse, so I’ll only add a few comments.

    I would posit that in addition to the trendline shown, there should be, as suggested by others, a line for injuries, and then these additional lines: 1. costs of injuries, 2. cost of maintaining people in permanent vegitative states, 3. and other costs for healing or treating everything else in between or up until the person dies (this should cover everything but people that are quickly or never healed), and, finally, 4. the cost of capital investments to create the capacities and competence to deploy all those battlefield lessons learned since WW2.

    This would go a long way towards either coroborating or refuting (as inadequate evidence) the trendline presented.

  • avatar

    Perhaps the drop during the late 1930s is related to hydraulic brakes becoming mainstream.

  • avatar

    I’d like to know whether there is a correlation between the safety (crash) ratings of cars and the ones that the fatalities occurred in?  In other words, can we actually prove that cars with better safety ratings reduce fatalities?

  • avatar

    The decline isn’t because of cars, but because we save lives better.
    Can you imagine having a serious auto accident in 1960 along a secondary road in New Mexico? Have you any idea how long it would have taken for you to receive medical care? Do you think someone had a phone in their car, or a CB radio, or a HAM radio so that someone was contacted when another driver found you? Do you think folks waited around for someone to get an ambulance?

    C’Mon people. Think! The number of miles driven compared to the number of deaths does not equal a significant figure shedding light on how much better cars are built. This chart is just damn stupid.

    All these people pushing for more auto regulations, claiming that their demands are making cars safer…most of this is utter nonsense spun by people who don’t think very hard. We have gotten to a point where there are too many people in the federal government demanding that all automobiles have some kind of electric window mechanisms that sense when something is caught in them. The costs incurred by all are supposed to make driving safer. This graph comes from a similar cracked way of thinking.

    Regulations save lives? Please! There comes a point where adding more regulations doesn’t add value.

    • 0 avatar

      Regulations save lives? Please! There comes a point where adding more regulations doesn’t add value.

      Yup, and those being regulated will tell you that every single step of the way, decade after decade.

      • 0 avatar

        If you said there is a diminishing rate of return on those regulations, you would have an argument with some merit.  However if you think that we would enjoy the safety of the modern automobile that we do today if improvements were only driven by market forces, well I have one hell of a beautiful bridge in Brooklyn with your name on it.

  • avatar
    John Horner

    Yeah, only a bit over 30 thousand people killed in traffic accidents in the US every year. Victory! Explain to me why we freaked out over the ~3,000 people killed on 9/11 then?

    • 0 avatar

      Imagine if we converted this to a dollars per life figure?  If you spent the kind of money fighting, say, poverty, poor nutrition or such that get spent on fighting terror or drugs?
      It’s funny how it’s often the same people shilling for de-regulation also scream for more spending on security, innit?

    • 0 avatar

      Explain to me why we freaked out over the ~3,000 people killed on 9/11 then?

      Please tell me that I didn’t just read this…

  • avatar

    When seatbelts, airbags, or any other safety feature is mandated, it’s not retrofitted to all cars the next day.  It probably takes 10 years to see them in over half the cars, and decades to be in the vast majority.

    The crashworthiness of cars, tires, handling, brakes, safety glass, padded dash, seatbelt availability and use, airbags, road design, drinking laws, ABS/stability control, FWD, medical and paramedic technology, life-flight helicopters, phone availability, cellular phone proliferation and other things have all contributed to this dramatic reduction.  Auto crashes are still the leading cause of death for the under-40 crowd though.

  • avatar

    No one has commented on the 33% decline from 24 in 1921 to 16 in 1926. What was the cause?

    There was a similar decline during the Great Depression, from 15 in 1930 to 10 in 1942. It is well known that fatalities decrease during difficult economic times. There was a spike in 1940 and ’41. Why? Was the Depression easing?

    The graph is more or less flat, at about 11, during the Second Wold War, even though driving was even more restricted than during the Depression.

    Although the graph too hard to read, I know from other sources that multiplying the rates of the 1970s by today’s travel would yield more than 100,000 fatalities a year.

  • avatar

    I see the increase from 1960 to 1966 as being symptomatic of the first Boomers turning 16 and getting their license and then thinning out the herd.  I bet we see an upward trend for the next 20 years as the Boomers turn 65, keep driving, and start having those age related accidents.  Stay away from those farmer’s markets, everyone!

  • avatar
    Dan R

    Perhaps it’s due to stricter enforcement of traffic laws?

  • avatar

    The thing that chart doesn’t deal with is hours driven per million miles.  It would seem to me its not a question of miles, but time on the road per death that is a better metric for determining the fatality level.   It’s more a question of how much time there is for something to wrong rather than speed.  Since the average speed traveled probably went up as cars got better, safer, and faster, and more people drove more as cars got better, you could easily lose the really interesting data by looking at the wrong way.

    But has the average driving speed gone up?  Yes, some highway travel may be faster, but what about the coast-to-coast traffic congestion?  Bumper to bumper, stop and go, hour after hour, day after day.  Doesn’t it seem worse than ever, every year?

  • avatar

    There is a spike starting at the bottom of the Great Depression – 1932.  

    Since the data starts with 1921 I wonder how much the condition of roads has to do with fatality rates in the early 20s?     At that tiime most roads were not paved.  Condition of the roads probably dovetails with what’s been said about the availability of EMS.  

Back to TopLeave a Reply

You must be logged in to post a comment.

Recent Comments

New Car Research

Get a Free Dealer Quote


  • Contributors

  • Vojta Dobes, Czech Republic
  • Matthew Guy, Canada
  • Kamil Kaluski, United States
  • Seth Parks, United States
  • Ronnie Schreiber, United States
  • Bozi Tatarevic, United States
  • Chris Tonn, United States
  • Moderators

  • Adam Tonge, United States
  • Kyree Williams, United States

Get No-holds-barred, take-no-prisoners Automotive News in your Facebook Feed!

Already Liked