ץ Government Failure: AIDS Research as an Example. It is gratifying that the idea of governmen failure has reached Eugene Volokh and Clayton Cramer . The idea that government is not a benevolent individual who acts for the public good, but an intricate tangle of people who act to further their own principles and interests using limited information is one of the most important ones of the past fifty years. It is crucial both positively---to explain why governments do inefficient things--- and normatively--- because it implies that even if the market is inefficient, government intervention might be even worse. I wonder if this is in Hayek anywhere? Probably not, or I'd have heard of it--- I associate the idea with Gordon Tullock, George Stigler, and Milton Friedman. And perhaps H.L. Mencken, who did write more than 50 years ago.

Messrs Volokh and Cramer don't get it quite right, though. The issue they discuss is, "Does the U.S. spend a lot on AIDS research because of the political strength of homosexuals?" First, of course, it is not clear that too much is spent on AIDS research. Certainly the amount spent per case is extremely high compared to other ailments. But this might be justified because of the following considerations:

AIDS affects younger people than other major ailments, and so curing it saves more and better years of life.
AIDS is an infectious disease, unlike cancer and heart disease, so it might spread rapidly at some point.
The number of homosexuals and drug users in the population might be on the increase, so the ailment will be more important in the future than it is now.
AIDS is a relatively new disease, and so we have not gone so far down diminishing returns in research on it. (the most important item on this list, and one hard for a non-scientist to judge)
AIDS research is really research on immunology, and that is a subject more worthy of research as basic science than, e.g., heart valve replacement.
Yearly research on an ailment has diminishing returns regardless of the number of cases, so it is senseless to spend 1000 times as much on disease X as on disease Y just because X is 1000 times more common. The relationship should be nonlinear.

Of course, there are considerations going the other way too--e.g., the fact that AIDS is completely preventable. We already know how to eliminate AIDS by simple and cheap changes in behavior. But let us put both the pros and cons aside.

Eugene Volokh's original thought was: Since AIDS mainly hurts members of disliked minority groups-- homosexuals and heroin users-- if the government spends so much on AIDS research, it must be for the public good, rather than in response to political pressure.

This is an important and wrong idea. Along the same lines, we could assume that whenever the government helps a minority rather than the majority, its actions must be for the public good. Thus, we could deduce that affirmative action, tax breaks for particular corporations, restrictions on sugar imports, and so forth are all for the public good. We would only need to watch out for the tyranny of the majority--- things such as high taxes on the rich or segregation laws. But the idea is wrong, and minorities often achieve their goals politically.

Let's assume, for illustration, that indeed too much is spent on AIDS research, and on a straight up-and-down vote, the public would reduce spending by 90%. Why might we nonetheless have high spending?

One reason a minority position might win is because it is part of a majority coalition.

A second reason a minority position might win is because it goes "under the radar screen" of the majority. The name "rational ignorance" is often given to this idea. Majority Coalitions. Suppose homosexuals are part of a majority coalition--the Republican Party. Homosexuals gets AIDS research, steelworkers get protection from imports, old people get healthcare spending, and evangelicals get foreign policy efforts to reduce religious persecution and abortion. None of these things is efficient, and each of the four groups opposes the policy of the other three, but overall each one is happy, and the losers are voters outside the coalition.

The Majority Coalition explanation is the one Clayton Cramer seems to focus on. It does explain why many minority political positions are taken by parties seeking to win a majority. Abortion provides the best examples: pro-abortion people have secured support even for late-term abortions from the Democrats, and anti-abortion have secured hostility even to first-week abortions from the Republicans, even though the average voter no doubt dislikes both of those positions. But I don't think this is the reason that would apply to AIDS funding, because that policy has a much lower profile.

Rational Ignorance. The average voter does not know the optimal level of research on AIDS versus cancer. Indeed, almost no voters do-- and I don't know it myself. The people who are most interested are scientists, university administrators, and homosexuals. Scientists and administrators are split on which topic of research they prefer to see funded, depending on their interest, but all of them want to see research spending's total level increased. Homosexuals want more AIDS funding. The natural result is increased spending on AIDS research, maybe or maybe not with reduced funding for other topics of research. The general public would not approve if they knew the details, but they don't. And if they try asking scientists, they will get biased answers. (Of course, I share the scientist's bias: my first response, here as elsewhere, is: "Spend more on knowledge!" My recent post on abortion body parts is another illustration.)

The idea of rational ignorance is the one I wish had penetrated more into the intelligentsia (of whom I take Volokh and Cramer to be unusually well-informed representatives). Political power does not require having even 2% of the electorate on your side; it depends more on wanting influence on obscure and complicated issues that nobody on the other side cares much about.

Finally, in the case of AIDS research (as opposed, to, say, a tax break tailored to help just one corporation), any explanation has to further explain why an individual homosexual would incur personal costs to help homosexuals generally. That was the subject of my New Year's Day post on rational irrationality.

[ http://php.indiana.edu/~erasmuse/w/04.01.03a.htm . erasmusen@yahoo.com. ]

To return to Eric Rasmusen's weblog, click http://php.indiana.edu/~erasmuse/w/0.rasmusen.htm.