The $667 Million Mirage: Wendover’s Case Against Airport Security Doesn't Survive Its Own Sources
Wendover Productions makes a compelling case that airport security is expensive, intrusive and frequently incompetent. But its most alarming conclusions depend on statistics that do not mean what the video tells viewers they mean. The $667 million price tag, the supposed 500 road deaths and the spectacular TSA failure rates are all real pieces of a much more complicated story. I find the presentation guilty of turning estimates into facts and possibilities into conclusions.
That matters. An argument about public safety should be held to a higher standard than an entertaining explanation of why airport queues are frustrating.
The $667 million figure is not an actual measurement
The video declares that the TSA spends an average of $667 million for every life saved. That sounds like an extraordinary finding from an economic study. It is actually the output of a highly assumption-dependent calculation.
The source is a 2020 paper, An Economic Study of the US Post-9/11 Aviation Security, by James Ford, Ardeshir Faghri, Dian Yuan and Saumabha Gayen. Its model assumes approximately $5 billion in annual checkpoint expenditure and 300 potential fatalities per year without enhanced security. That figure of 300 is not an observed historical death rate: it comes from assuming a 9/11-scale event would occur approximately once every ten years.
SCIRP+1
The researchers then assume that other pre-boarding security measures reduce risk by 50%, while assigning checkpoint screening an efficiency of just 5%. That last assumption comes from the 2015 undercover tests in which screeners failed 67 of 70 attempts. The resulting $667 million is therefore a modeled scenario, not a measured historical cost per life actually saved.
There is an even more damaging omission. The same paper explicitly warns that the 95% failure rate does not establish that TSA checkpoints are 95% ineffective. It considers 50% checkpoint effectiveness a more reasonable assumption, producing a cost estimate of approximately $66 million per life saved instead. Its abstract ultimately concludes that TSA's spending on aviation security is justified. Wendover presents the most dramatic scenario while leaving out the qualification and the paper's overall conclusion.
SCIRP+1
That is not a minor editorial oversight. It changes the meaning of the statistic by an order of magnitude.
The TSA failures were real. The interpretation is the problem.
The 2015 results deserve serious criticism. DHS Inspector General investigators reportedly passed prohibited items through checkpoints in 67 of 70 attempts. Inspector General John Roth subsequently described the performance as universally disappointing, including failures involving technology, procedures and human judgment.
ABC News+1
I have no interest in making excuses for that performance. But a test designed to uncover vulnerabilities is not a statistical measurement of the probability that an actual terrorist will defeat the entire aviation security system.
The distinction matters because the 67-of-70 result was converted into a 5% effectiveness assumption in the economic calculation. It measured success against a particular set of covert attempts at checkpoints. It did not establish that 95% of real-world threats would pass through, or that the entire system prevented only 5% of potential attacks.
The Government Accountability Office's subsequent findings make the situation more interesting, not less troubling. Its 2019 investigation observed 26 covert tests and found that TSA had improved parts of its testing process, but still had problems with test quality and vulnerability management. Nine vulnerabilities submitted for mitigation since 2015 had not been formally resolved by September 2018. In some cases, assigning an office to begin mitigation took up to seven months.
U.S. GAO
The proper indictment is therefore not that TSA has a proven 5% effectiveness rate. It is that serious vulnerabilities were identified, and the agency struggled to resolve them. That is a much stronger criticism because it is supported by evidence rather than a convenient mathematical shortcut.
The 500 road deaths are an extrapolation, not a body count
Wendover says that more than 500 people died in 2002 because airport security delays pushed travelers into cars. The underlying research is considerably more specific.
A 2007 study by Cornell economists Garrick Blalock, Vrinda Kadiyali and Daniel Simon found that mandatory baggage screening reduced originating passenger volumes by approximately 6%, rising to 9% at the 50 busiest airports. Crucially, federalizing passenger screening had little effect on passenger demand. The video turns a finding about baggage screening into a broader indictment of post-9/11 airport security. (University of Chicago Press)
The researchers estimated approximately 129 additional road fatalities in the fourth quarter of 2002. They calculated this using 3.2 million trips not taken, 2.7 passenger enplanements per trip, and an estimated 15 additional road deaths per million lost enplanements. The frequently repeated figure of approximately 500 deaths annually is an extrapolation from that quarterly estimate, not 500 individually established deaths in calendar year 2002. The researchers themselves called the fatality calculation a very rough approximation. (Original study; WIRED's examination of the 500 figure)
There is another important distinction. A separate 2005 Cornell study estimated approximately 344 additional driving deaths per month during the final three months of 2001, with a cumulative estimate of 2,170 deaths attributable to the public's response to September 11. But the researchers identified two contributing factors: fear of flying and the inconvenience of tighter airport security. They did not establish that screening delays alone caused those deaths. (Cornell working paper)
These are meaningful findings. Security inconvenience can have deadly consequences. But Wendover collapses separate studies, periods and causes into a single emotionally powerful claim. The evidence supports a serious policy trade-off, not the confident assertion that TSA procedures killed more than 500 people in 2002.
Israel's security record is not proof of an impenetrable system
The video spends considerable time presenting Ben Gurion Airport as the gold standard. It describes interviews, passenger profiling, risk classifications, plainclothes officers and armed air marshals. It then uses the absence of successful hijackings or bombings on departing aircraft as evidence that the system works.
There is a fundamental problem with that reasoning.
Wendover correctly acknowledges that there is no control group of developed countries operating without airport security. Yet it applies a different standard to Israel: a record without successful attacks becomes evidence that its security system is nearly impenetrable.
A zero-attack record cannot establish which part of a security system prevented an attack, whether attackers were deterred before attempting one, or whether the same results would occur under different threat conditions and passenger volumes.
The human cost of Israel's approach is also better documented than the video suggests. A 2012 peer-reviewed study by Badi Hasisi, Yoram Margalioth and Liav Orgad surveyed 918 Ben Gurion passengers, including 306 Palestinian citizens of Israel, specifically examining the individual and social costs of airport screening for Israeli Arabs. (American Law and Economics Review)
A further study published in Law & Society Review examined 1,970 passengers. It found that fair treatment by security personnel could reduce the perceived hostility of additional screening, but two screening measures retained independently significant negative effects. (Law & Society Review, 2016)
Neither study establishes that Israeli profiling is ineffective. But neither establishes that it prevents attacks. They document the burden imposed on passengers, not the security benefit that supposedly justifies it.
The video's assertion that Arab travelers can hardly avoid receiving the highest risk classification also deserves more than an unsupported declaration. Where are the classification statistics? What proportion of flagged passengers actually present a threat? How many innocent passengers are subjected to intensive questioning?
Those questions matter because a system can be operationally successful while imposing substantial costs on innocent people. Calling it nearly impenetrable does not answer them.
The missing counterfactual makes the conclusion premature
There is a deeper problem with the entire argument: security successes are much harder to count than security failures.
The 2020 economic study itself acknowledges that aviation security performance is classified, that there is no public way to determine how many threats TSA stops, and that reported failures are much more visible. That means the available evidence is inherently incomplete. (Ford et al., 2020)
The video recognizes that deterrence may explain the absence of attacks, but it never establishes how much protection comes from screening, intelligence, airport surveillance, hardened cockpit doors or other security layers.
A 2018 study by Mark Stewart and John Mueller modeled 18 layers of aviation security and estimated a 93% reduction in the risk of a terrorist-organization attack involving a passenger-borne bomb. This is a theoretical risk assessment, not proof of actual performance, but it illustrates why measuring one checkpoint's failures cannot establish the effectiveness of the entire system. (Journal of Transportation Security)
I am not suggesting that the TSA deserves credit for attacks that never happened. I am saying the opposite: nobody can responsibly calculate the lives saved or lost without confronting the uncertainty on both sides.
Wendover gives the uncertainty considerable weight when discussing American security, but less weight when presenting Israel's record. That is an editorial inconsistency, not a scientific conclusion.
The verdict: a legitimate warning presented as a settled calculation
Wendover identifies genuine problems. TSA checkpoint failures were serious, corrective action was inadequate, screening imposed measurable costs, and the Israeli model raises difficult questions about discrimination and passenger treatment.
But the video does not establish that airport security costs more lives than it saves. Its most dramatic figures depend on assumptions, extrapolations and incomplete comparisons. It presents the $667 million scenario without the source paper's qualification, turns an estimated quarterly road-fatality figure into a much larger annual claim, and treats Israel's absence of successful attacks as stronger evidence than the absence of attacks in the United States.
I would publish the criticism of TSA's failures. I would publish the economic case for faster, more intelligent screening. I would investigate the human cost of Israeli profiling.
I would not publish the video's implied verdict as established fact.
Airport security deserves scrutiny. So does anyone claiming to know how many lives it saves or destroys. Wendover has identified an important question, but it has not done the mathematics necessary to answer it.
References
1. How to Design Impenetrable Airport Security — Wendover Productions, YouTube. Original video. The work examined in this critique.
2. An Economic Study of the US Post-9/11 Aviation Security — James Ford, Ardeshir Faghri, Dian Yuan and Saumabha Gayen, Open Journal of Business and Management, August 7, 2020. Read study. Establishes the assumptions behind the $667 million calculation and explicitly discusses the alternative $66 million scenario.
SCIRP+1
3. EXCLUSIVE: Undercover DHS Tests Find Security Failures at US Airports — ABC News, June 1, 2015. Read report. Reports the 67 failures in 70 covert tests; the underlying Inspector General findings were not publicly released in full.
ABC News
4. Aviation Security: TSA Improved Covert Testing but Needs to Conduct More Risk-Informed Tests and Address Vulnerabilities — U.S. Government Accountability Office, April 4, 2019. Read GAO report. Documents persistent vulnerabilities, weaknesses in testing and failures to formally resolve identified problems.
U.S. GAO
5. The Impact of Post-9/11 Airport Security Measures on the Demand for Air Travel — Garrick Blalock, Vrinda Kadiyali and Daniel H. Simon, Journal of Law and Economics, November 2007. Journal publication · Full paper. Separates the effects of baggage screening from federalization and documents the assumptions behind the estimated road fatalities.
RCNi Company Limited+1
6. Driving Fatalities After 9/11: A Hidden Cost of Terrorism — Garrick Blalock, Vrinda Kadiyali and Daniel H. Simon, Cornell University, December 5, 2005. Read working paper. Estimates indirect road deaths following 9/11 and identifies both fear and airport-security inconvenience as contributing factors.
blalock.dyson.cornell.edu
7. Is Airport Security Killing 500 People a Year? — David Dobbs, WIRED, April 5, 2012. Read article. Traces the widely repeated 500-deaths figure back to the original quarterly estimate and its annual extrapolation.
WIRED
8. Ethnic Profiling in Airport Screening: Lessons from Israel, 1968–2010 — Badi Hasisi, Yoram Margalioth and Liav Orgad, American Law and Economics Review, September 17, 2012. Read study. Provides passenger-survey evidence concerning the social costs of Israeli airport profiling.
Oxford Academic
9. Is It the What or the How? The Roles of High-Policing Tactics and Procedural Justice in Predicting Perceptions of Hostile Treatment — Tal Jonathan-Zamir, Badi Hasisi and Yoram Margalioth, Law & Society Review, September 2016. Read study. Examines the relationship between intensive screening, passenger treatment and perceived hostility using 1,970 passengers.
Cambridge Core
10. Risk and Economic Assessment of U.S. Aviation Security for Passenger-Borne Bomb Attacks — Mark G. Stewart and John Mueller, Journal of Transportation Security, December 2018. Read study. Models the combined risk reduction of 18 aviation security layers, illustrating the limitations of evaluating checkpoint screening in isolation.
IDEAS/RePEc