Below the fold is a (slightly corrected) comment  that I left on this EconLog post by Bryan Caplan . My comment is in response to a comment left, on the same post, by Thaomas – who, I believe, is the Thomas Hutcheson who sometimes comments here at Cafe Hayek. Thaomas and Thomas Hutcheson frequently complain that libertarian and libertarian-leaning scholars are too ideological – a stance that causes us to unscientifically refuse to use cost-benefit analyses as often as a good scientific, objective disposition requires. And the typical example of this unscientific disposition that Thaomas and Thomas Hutcheson point to is the ‘libertarian’ analysis of minimum wages.
Thaomas writes :
Good analysis but it typically stops where so many Libertarian analyses stop, when it finds the first downside. It’s like saying we cannot have a minimum wage because it will cause some unemployment.
The minimum wage is a very poor example of Thaomas’s point. Economists over the years – including many of a libertarian bent – have, when analyzing the effects of minimum-wage legislation, often gone well beyond pointing out that minimum wages destroy employment opportunities for some low-skilled workers. Other consequences of minimum wages that have long been noted in the literature – including in the writings of libertarian-leaning scholars such as Milton Friedman, Deirdre McCloskey, Thomas Sowell, and Walter Williams – include:
(1) the loss of current earnings by those made unemployed by minimum wages is not the only loss; another loss is the loss of opportunities to gain job experience and on-the-job training; this latter loss also reduces (but in ways very difficult to measure accurately) the affected workers’ future earnings;
(2) the distribution of the minimum-wage’s benefits (higher wages for some workers) and losses (unemployment or worse jobs for other workers) isn’t random; the winners in the U.S. disproportionately are better-schooled whites (likely each with his or her own private means of transportation and no parental responsibilities) while the losers disproportionately are blacks and other minorities.
Philosophically mature analysts of minimum-wage legislation also understand two other realities.
First, if the relevant population is considered to be everyone – rather than, as Thaomas would seemingly have it, only low-skilled workers and their households – economic theory is quite clear that, in the absence of monopsony power in the market for low-skilled workers, the size of the winners’ gains from a minimum wage will be less than the size of the total losses suffered by society at large.
Second, even if we assume that the relevant populations to consider are to be confined to low-skilled workers and their households, considerations such as points (1) and (2) above make any cost-benefit analysis of minimum wages very impractical compared with the intellectual stance of treating the determination of the market prices of hours of low-skilled labor no differently from how we treat the determination of the market prices of hours of high-skilled labor, of toothpicks, of automobiles, and of feta cheese. If we have no itch to withhold judgment on the merits of allowing the market to price these (and countless other goods and services) until we conduct cost-benefit analyses on the use of price controls in each of these other markets, what reason is there to believe that the market for low-skilled labor should be treated differently? Once the analyst of minimum wages disproves (or casts healthy suspicion on) the widespread presumption among today’s minimum-wage proponents that minimum wages have no negative employment effects, that analyst – whether libertarian or not – bears no further burden to prove that minimum wages fail some cost-benefit test.
Let me put this last point somewhat differently: if the income gains, from a minimum wage, captured by some people are to be counted as ‘benefits’ to be weighed against the losses of other people – and if these gains can in principle be so high relative to the losses that the minimum wage passes a cost-benefit test that then is used to support the minimum wage – then there is flung open a Pandora’s box of utilitarian horrors.
For example, someone (call him CB) might propose that the state prohibit the employment of all blacks under the age of 20. CB would correctly point out that, while his policy would obviously have some losers, it would also produce winners – namely, the wages of non-black, mostly young workers will increase as a result. And he’d be correct. I can also imagine that, in reality, the measured increase in the aggregate pay of this policy’s winners would be larger than the measured decrease in the aggregate pay of its losers (especially if we confine “losers” only to the black teenagers who lose jobs). Yet who would counsel that we should, therefore, withhold judgment on CB’s proposed policy until a cost-benefit analysis is conducted? Who would think it to be “libertarian” or “one-sided” or “unscientific” to prejudge as unacceptable a policy of prohibiting the employment of blacks under the age of 20?