Costs and benefits of homeland security

The economist's Gulliver blog summarizes a recent paper on the costs of homeland security measures since 9/11. It is hard to imagine that more than a modest fraction of them (e.g. hardening the cockpit doors on airplanes) come within shouting distance of passing a cost-benefit test. But, as with the FDA and drug approvals, politicians face an asymmetric loss functions and so exploit the limitations that the principal (the voters) have over them in their role as our agents to minimize their own probability of job loss by adopting measures that may reduce security, but not enough to justify their costs.