Monday, December 23, 2013

The TSA: Mistakes Were Made

As so many of us pass through airports at this time of year, contemplate the Transportation Safety Administration. The politics of creating the TSA were straightforward. The terrorist attacks of September 11, 2001, occurred. By November 2001, the Aviation and Transportation Security Act was passed into law, creating the TSA. This looks like a classic example of that old benighted syllogism of public policy: "Something must be done. Here is something. Therefore, it must be done." Chris Edwards reviews the history and makes the case for "Privatizing the Transportation Security Administration" (Cato Institute, Policy Analysis No. 742, November 19, 2013).

"TSA’s main activity is operating security screening at more than 450 commercial airports across the nation. The agency also runs the Federal Air Marshal Service (FAMS), analyzes intelligence data, and oversees the security of rail, transit, highways, and pipelines. TSA has 62,000 employees and an annual budget in 2013 of $7.9 billion." 

No other high-income country has put airport screening under central government control. Edwards writes: "More than 80 percent of Europe’s commercial airports use private screening companies, including those in Britain, France, Germany, and Spain. The other airports in Europe use their own in-house security, but no major country in Europe uses the national government’s aviation bureaucracy for screening. Europe’s airports moved to private contracting during the 1980s and 1990s after numerous hijackings and terrorist threats, and it has worked very well. Canada also uses private screening companies at its commercial airports, and some airports also use private firms for general airport security. After 9/11, the government created the Canadian Air Transport Security Authority, which oversees screening at the country’s 89 commercial airports.136 But the
screening itself is carried out by three expert private firms—G4S, Garda, and Securitas—which are each responsible for a group of particular Canadian airports."

There's an old but useful rule for deciding when a certain desired activity should be carried out by government itself or instead if the government should pay others to do it: "Government should steer, not row." Or to put it another way, do I feel more secure in a situation where undercover government inspectors are continually testing private airport security companies, who know they are likely to lose their contract if they fail to perform? Or in a situation where I'm depending on one branch of government to inspect the security efforts of another branch of government, with all the inter-bureaucratic incentives not to embarrass each other, and where no one in a position of authority is going to lose their job or their annual bonus if security fails to perform?

In addition to pure security issues, a contracting approach to airport security has other possible benefits. One is that a contracting firm has better incentives to provide friendly and rapid customer service than does a government agency that can't be fired. Another is that local control allows more rapid adjustment of schedules across the 450 airports for days and times when demand is high or low, compared with a rule in which all such adjustments need to be cleared with a faraway bureaucracy. Improved methods of providing security--either less time or different equipment--are more likely to emerge from private firms competing for work at airports across the county than from a federal bureaucracy.

Not unexpectedly, Edwards cites a litany of reports that suggest poor management at the TSA. Out of 240 federal agencies, the TSA ranked 232nd in employee satisfaction. A GAO report found large increases in employee misconduct at TSA. A former TSA chief has reported that the agency is "hopelessly bureaucratic." The TSA bought 207 "puffer" machines to detect explosives, but then decided they didn't  work well and mothballed them. Before TSA bought the full-body scanners, GAO requested that ti do a cost-benefit study, but none was done. In June 2011, a federal judge required that such a study be done, but it still hasn't been done. At least one outside study suggests that the machines would fail a cost-benefit analysis quite dramatically.

Maybe we needed more airport screeners in the aftermath of 9/11, although I'm not familiar with any evidence on the point. But did we need the number of airport security screeners to more than double from 16,000 before the TSA to more than 40,000 immediately after, and 53,000 today?

The 2001 legislation allowed airports to opt out, and five immediately did so--with San Francisco being the largest. Now there are 16 airports that have opted out, and others are applying to do so. There's no evidence that the airports which have opted out of the TSA are any less secure. And there aren't any airports moving the other direction, from privatized airport security to join up with the TSA. If the TSA can't be abolished outright, at least such opt-outs can be allowed and encouraged.