Ten years after 9/11 it’s a good time for our government to review its approach to the problem of terrorism — especially with respect to airline security. We should retain the policies that work well and discard policies that have proven to be wasteful or counterproductive. One program that falls into the latter category is the Federal Air Marshal Service.

The ironic truth is that the attacks on commercial aviation on 9/11 provided both the impetus for a rapid increase in air marshal numbers and the increasing irrelevance of air marshals in deterring terrorist attacks.

Two major changes took place after 9/11 that forever altered the security calculus against terrorists. First, airlines installed hardened cockpit doors that make it very difficult to gain control of the aircraft. Second, the flying public realized that the rules of the game had changed. Terrorists were now using jets as cruise missiles instead of holding them on the tarmac and negotiating for the release of political prisoners. The passengers’ heroic stand on Flight 93 marked the end of the passive hostage era.

9/11 also marked the beginning of a rapid expansion in the ranks of air marshals, from a few dozen in 2001 to thousands today (the actual number is classified, as it should be). While some experts proposed that we embrace an Israeli-style policy of putting air marshals on every single flight, we haven’t. Doing so would have proven prohibitively expensive, costing over $9 billion annually — more than we currently allocate for the whole of the TSA.

We already know that about 90 percent of the airline flights proceed without any air marshal aboard. The $930 million we spend on air marshals annually affords about 10 percent coverage. The program produces an average of four arrests a year.

The air marshals’ deterrent effect has largely withered away because of a change in al Qaida tactics. The would-be shoe- and underwear-bombers were merely trying to blow up aircraft, not take control of the cockpit. Both were tackled by the passengers and crew of their target flights, not shot or apprehended by air marshals. The flying public has subdued unruly, intoxicated or suspicious passengers on flights bound for Newark, Tel Aviv, Los Angeles and numerous other destinations.

The airlines have taken note of the change in terror tactics as well. They recently asked the air marshals not to sit in first class; the cockpit is not the target anymore, so giving air marshals prime real estate is not worth the hit to the bottom line.

Academics are also questioning the cost-effectiveness of the marshal program. A study by Professors John Mueller of Ohio State and Mark Stewart of the University of Newcastle found that the estimated cost of a life saved by the presence of air marshals was $180 million, far more than the $1 million to $10 million that the Office of Management and Budget recommends. Hardened cockpit doors proved more cost-effective, with an estimated $800,000 spent per life saved.

The TSA’s program to allow pilots to keep firearms in defense of the cockpit also operates at a fraction of the cost of employing air marshals. It’s entirely voluntary, and the pilots pay the cost of attending training courses.

The taxpayers and the flying public would both be better off if Congress abolished the air marshal service and left it to the airlines to provide security personnel. Doing so would cut nearly a billion dollars from the federal budget.

Moreover, privately employed security personnel would have a greater deterrent effect than federal air marshals. We only know how cost-ineffective the air marshals are because their budget is a matter of public record; if the cost of employing air marshals were hidden within an airline’s overall operating expenses, the airlines could more effectively claim broad deployment of air marshals.

One of the reasons there is so much anger toward Washington these days is the seeming unwillingness of the president and Congress to cancel agencies and programs that are wasteful and unnecessary. All of the federal government’s programs should be reviewed with an eye toward cost-effectiveness. The Federal Air Marshal Service fails the test, and should be shuttered in favor of the other layers of security we have in place.

David Rittgers is a legal policy analyst at the Cato Institute.