The TSA Has Caused More Deaths Than Terrorism — But Maybe it No Longer Does

This week I linked to a post by Bruce Schneier where he explained how post-9/11 increased airport security led to more deaths than terrorist attacks have cuased. That’s because people switched from flying (short haul routes) to driving with the increased hassle and time taken by security.

That predictably led to some consternation in the comments and in a barrage of emails. One of the smartest responses I got pointed out that overall traffic fatalities are on the decline.

Now, the statistic may get overused or there may be overclaims about what it says, but I’ve read John Mueller (Woody Hayes Chair of National Security Studies, and Professor of Political Science at Ohio State University) on this issue and the post appears to be based on sound research. And there’s not a claim that I’ve seen about traffic fatalities overall rising.

Mueller cites data which suggests that the more cumbersome security screening process that went into place post-9/11 led a subset of people to make the decision to drive rather than fly on short-haul routes.

And since driving is not as safe as flying, a certain number of them got into traffic accidents and died (who would otherwise not have died had they flown). They calculate the order of magnitude as a few hundred people per year.

And so the claim that more people died from traffic fatalities as a result of flying instead of driving, a decision they made because of more cumbersome security screening, would be correct and that the number of fatalities is greater than the number of people killed by terrorism is also likely correct.

I do think that the fair and reasonable counter argument is that what this data shows is not reflective of current TSA security practice. In the time after 9/11 security lines were much much longer than they generally are today.

A decade ago 40 minute security lines at Dulles at 5pm on a Thursday or Friday afternoon were commonplace. Those long lines, while they certainly happen they are no longer as commonplace. So it is no longer likely (though I do not have data on this) that the same trade-off is happening between taking short haul flights and driving, and so you can no longer attribute the same number of traffic fatalities to the screening process.

It can be simultaneously true that “more people died as a result of TSA screening than died in terrorist attacks” and also true that this phenomenon is not continuing (though as I say I do not know of research on this latter point and have not myself tried to piece together data so it is just a hypothesis).

I have also written by the way that I do not really blame the TSA as such for the current state of screening.

I actually think that there are people there who would like to focus their resources on security rather than security theater,

The political environment simply will not let them do that. They recognize that focusing on small threats takes resources away from focusing on real threats, but when they announce a change political opportunists jump on them and they back down. There’s ton of waste and bureaucratic mission creep to criticize them for.

I do not like the nude-o-scopes but millimeter wave scanners are better than backscatter, and the software they use to address privacy concerns is a real improvement as well. Not that these really foil terrorists, but they are less intrusive than what the TSA was doing before (a fairly low bar to exceed).

Their behavior detection programs are silly because a few weeks of training does not mean you have an Israeli model, you cannot just scale what’s done in Tel Aviv (one medium-sized airport) across the entire US in a meaningful way in a short period of time if it would be desirable to do so.

I do think the research behind the claim appears sound – the question is what you want to do with the claim that is at issue.

    You can join the 30,000+ people who see these deals and analysis every day — sign up to receive posts by email (just one e-mail per day) or subscribe to the RSS feed. It’s free. Don’t miss out!

About Gary Leff

Gary Leff is one of the foremost experts in the field of miles, points, and frequent business travel - a topic he has covered since 2002. Co-founder of frequent flyer community, emcee of the Freddie Awards, and named one of the "World's Top Travel Experts" by Conde' Nast Traveler (2010-Present) Gary has been a guest on most major news media, profiled in several top print publications, and published broadly on the topic of consumer loyalty. More About Gary »

More articles by Gary Leff »


  1. I disagree. I have heard this argument before, as this is my primary topic of coverage, and have seen absolutely no evidence that there are more traffic fatalities due to people avoiding flying due to security.


  2. John Mueller is a known crazy person, also known as a libertarian. Another nonsensical view from the right wing.

  3. Gary, to rove a hypothesis, you need to demonstrate cause and effect. If there is no statistical correlation to implementation of TSA procedures, then the hypothesis is unproven. Even a few hundred additional deaths would show up statistically and the evidence is no there.

  4. Glenn,

    I wouldn’t be so sure about your conclusion given how poor people are at handling statistics and how many mistakes even statisticians make. By the way, given that the affordability of driving cars has fallen — even more so for those with higher insurance rates (presumably in some part due to higher accident rates)– in the US in the last 12 years, I’d say that a major economic recession suppresses driving/road fatalities (more than would otherwise be the case) and thus mask at least some of the damage done by the TSA.

  5. If you amend the claim to be that there the TSA has caused more fatalities than terrorists have *on airplanes or in airports since TSA’s inception*, I think it’s more defensible (because the latter number is zero).

    However, there is still a problem with the statement (as much as I hate to defend the TSA, as a “crazy person Libertarian” who is certainly not “right wing”). I certainly won’t defend their practices (anything past what goes on at pre-check should be abolished), but as a pedantic person I have to challenge statements that are problematic.

    Using motor vehicle accidents under the assumption of causality is troublesome. It’s likely true (especially with my modification), but it’s approaching the “butterfly effect”. Because of this you can easily counter it in the opposite direction: “TSA has saved more lives than caused people to die”. For evidence you would only need to look at airline crashes and find those same people who refused to fly because of the TSA and count them as having been saved. Also, more directly, those people who were supposed to be on those flights who either missed the flight because the TSA delayed them or “would not be flying that day” because the TSA didn’t like the color of their hair.

    What really matters is that the TSA is a tremendous waste of money spent on security theater and should be abolished, full stop, on its non-merits. It’s the most effective outcome of the 9/11 terrorists, as it’s continued for so many years after the attacks, and has inconvenienced, frightened, assaulted, and sexually molested millions (including children).

  6. You should be embarrassed to be parroting this nonsense. I’ll stick to reading Lucky from now on.

  7. While I agree with Bruce’s overall position on these issues, his use of statistics is naive, as he is comparing apples and oranges from a statistical/risk point of view. Traffic deaths are in the realm of the Gaussian; you will never have a year where you have millions of traffic deaths- they will trend up or trend down, but never with that kind of wild variation. Not so for terrorist attacks, where a single large event (say a large bomb in a major city) could cause hundreds of thousands of deaths, and would completely reverse the trend Bruce is seeing. We cannot make these kinds of claims about the statistical properties of terrorist attack deaths for this reason.

    I love Bruce’s work in general, and certainly believe both that the TSA is ineffective, and that on philosophical grounds the overall trade of liberty for security is not one we should make, but he is overstepping his expertise here.

  8. Very interesting article. “Those long lines, while they certainly happen they are no longer as commonplace.” Unfortunately, the reality is that most people base their time estimate for going through security not on the average time, but on their estimate of the maximum time. That estimate is based on their own experiences and that of their friends. It takes a long, long time for those personal estimates of a maximum time to go down, no matter how quickly the average goes down. However, this is rational behavior as it is not an average time in security that makes you miss a flight, it is an outlier long security line.

  9. The comparison is also flawed because it doesn’t capture all affected fatalities. Just off the top of my head, I can personally attest to the fact that long security lines make me smoke less and therefore live longer. I’m sure there are other similar variables, on both sides of the ledger.

  10. I am just curious as frankly I haven’t really made up my mind on this whole subject, but to the longer-term readers of this blog, has Gary ever put out his specific ideas on what his plan would be? I would just like to see this because I would find it interesting. I was surprised to see him dismissing an analog of the Israeli model because usually that’s what the anti-TSA folks point to as a gold standard.

    Personally, I am hesitant to bash an idea when I don’t have a better one myself. That’s why I usually don’t say much about the TSA, or even find myself coming across as an apologist.

  11. My biggest gripe on this study is that it apperently doesn’t take into affect the number of people who after 9/11 didn’t (and still don’t) fly places because of a fear of terrorism due to that day. I personally know of a handful of people who drive on every vacation bc of this. I know not of a single person who chooses to drive bc of sercurity lines.

    Add those traffic deaths to the terrorism deaths (which by the author’s logic we should), and the numbers would flip greatly

  12. The problem is that in the US we can’t stereotype. We all now very well where these terrorists are coming from and what they look like, but we choose to interrogate 5-year-old americans and old ladies and old people flying to a hospice center instead dealing with young terrorist looking dudes from the middle east. I don’t agree with many things that go on on Israel, but they have one good thing, at least, they do stereotype and that saves the good dudes the harassment, and it saves everyone time. you don’t stereotype because you’r a racist you do do it because you want to save lives, and not be politically correct!!!

  13. @CW I do think that Israel does a generally good job, my criticism is not of the Israeli model for Israel, it is that the model does not scale to the US. Ben Gurion airport in Tel Aviv handles traffic on the order of San Jose airport. You would have to multiply that out across the entire US system. So far the TSA has figured out how to give a few weeks of training and it’s become just more or less a license to profile.

    Reinforced cockpit doors is a good system.

    The biggest deterrent now is that terrorists have a hard time getting into the cockpit, procedures have changed not to let them in even if they are threatening the lives of passengers (and terrorist organizations know this), and passengers will no long sit idly when someone attempts to take over an aircraft.

    Basically the changes in procedure, in hardware, and in passenger mindset make US airliners not a very good target for the kind of terrorist activity that took place on 9/11.

    There are still plots that the current system doesn’t do a very good job of weeding out. That’s why you have stories of ‘Christmas Day’ bomers, shoe bombers, liquid bombers etc. Mostly the plots just don’t work very well.

    I also think we cannot get perfect security, and it would not make sense to do what’s necessary to get it.

    But there are incremental reforms of the TSA that would be better than the status quo, that the TSA is generally in favor of even (like not focusing on non-threat items they’re forced to search the checkpoint for) but that the TSA is not permitted to change its rules. The TSA is not currently being allowed to focus on real threats at the checkpoint, there’s a tradeoff searching for liquids and small knives and it distracts from actual dangers.

  14. Gary – thank you very much for your thought-out and detailed response. Your raise some good points, ones that I do not disagree with at all. I’ve still yet to make up my mind on the whole, big-picture entity that is the TSA and all the policy and bureaucratic baggage that comes along with it, but the points you mention are really hard to disagree with.

  15. “post-9/11 increased airport security led to more deaths than terrorist attacks have caused”, while probably true, is not a particularly relevant statement. I a relevant statement would be “post-9/11 increased airport security led to more deaths than were prevented by the TSA procedures”. But no one can make this statement, since it is impossible to know how many deaths were prevented by the mere presence of the TSA in airports. That presence certainly discourages terrorists from trying to use box cutters to take over a plane, for example.

  16. “a few hundred per year?” Does that really make sense intuitively given that total annual driving deaths are roughly into the 30,000 range and 10,000 or so are from DUI etc.?

    Quite honestly that number is not credible – it would suggest that massive numbers of people are driving instead of flying, not to mention freeway miles (daytime) are generally safer than city driving, mature drivers (businness travelers) are better drivers than young kids etc.

    While I generally concur with your other comments about TSA, I’d recommend you stay away from bogus estimates of car deaths attributable to aviation security measures. More likely more people are driving due to airline consolidation and fare hikes.

Comments are closed.