Alex Tabarrok, at Marginal Revolution, has an interesting blog on asteroid deflection as a public good. He notes that neither markets nor governments are supplying it, even though the risk from dying from an asteroid strike are approximately the same as dying in a plane crash - which countries and markets spend lots of money to avoid - and a large asteroid strike could kill billions, or everyone.
Alex notes that the US spends millions on the detection of near earth objects; indeed, it is the only country with a substantial research program to identify possible sources of asteroid strikes before they occur. Detection is, of course, a necessary prerequisite to deflection. Thus, his initial assertion that no government is supplying asteroid deflection is a bit misleading. Moreover, the fact that the US is investing substantial sums in asteroid detection and on advanced propulsion systems that might be used to deflect asteroids in the future, raises an important question that Alex does not answer: why is the US moving at all in the direction of unilateral provision of a global public good?
An excellent answer to that question is provided in Scott Barrett's superb book, Why Cooperate? The Incentive to Supply Global Public Goods (Oxford 2007). In it, Scott distinguishes five types of global public goods: "single best effort"; "weakest link"; "aggregate effort"; "mutual restraint"; and "coordination." This is not the place for a complete explication of those categories, but, for example, climate change mitigation is an "aggregate effort" public good because its ultimate success depends on the total efforts of all countries. By contrast, large asteroid deflection is a "single best effort" public good because it may be in the interest of one very wealthy country, such as the US, to supply it, even if no other country cooperates. In other words, asteroid deflection may not require international collective action at all.