The Bailout, the Election, and Strategic Politics
1. Most of the U.S. political leadership, including both condidates, supports the bailout.
2. A sizable majority of the electorate is at least suspicious of the bailout. One congressman reported that his mail and email was running 50/50—50% "no" and 50% "hell no."
3. Obama appears to be opening up a substantial lead over McCain in the polls.
This suggests an obvious opportunity for McCain: Come out against the bailout, with some plausible sounding explanation of why he was initially for it. It's risky, but if he is going to lose anyway it might be worth taking the risk.
McCain has not done it, and surely knows more about how to win elections than I do, which raises some interesting possibilities.
Suppose McCain, and those around him, are interested not only in winning this election but in longer term consequences for McCain and the Republican party. It may make sense for themto support the bailout, not because it is a good idea but because future events are more likely to look like evidence for the bailout than like evidence against it. Consider the possibilities:
1. No bailout, the economy does terribly. Opponents of the bailout look terrible—whether or not the bailout would have worked.
2. No bailout, lots of investors lose a lot of money, the current recession continues for a while. Opponents of the bailout don't look terrible but neither do supporters, who will argue that things would have gone much better with the bailout.
3. Bailout, the economy does terribly. Supporters of the bailout will argue that this is evidence of how much worse things would have been without a bailout—and who can prove them wrong?
4. Bailout, the current recession continues for a while. Supporters of the bailout will claim to have saved us from a depression.
Looking at these alternatives, it seems plausible that if the bailout has no effect either way on the condition of the economy, a politician is better off supporting it. Only if the bailout makes worse outcomes substantially more likely or if some bad outcome directly associated with the bailout occurs, such as spending 700 billion and then having lots of firms fail anyway six months later, does opposition look like an attractive long run gamble.
I have a feeling that this argument could be generalized. One ought to be able to predict at least some patterns of political outcomes by asking what policies are more likely to be followed by outcomes that look like clear evidence for or against them. The point is related to the familiar observation that one would expect politicians to design programs whose benefits are easy to see and whose costs are hard to see.