Why do married men earn more?


Jason Collins


February 29, 2012

Even after controlling for observable traits such as IQ, married men earn more. Bryan Caplan suggests there are three economic explanations for this male marriage premium:

  1. Ability bias: Qualities that make a man attractive to a woman are also attractive to employers.
  2. Human capital: Marriage makes men more productive.
  3. Signalling: Marriage signals to employers that a man has desirable traits.

Caplan notes a study that argues that ability bias accounts for less than 10 per cent of the premium. However, Caplan suggests that ability bias accounts for 50 per cent of the premium, while signalling accounts for less than 10 per cent. That leaves human capital to fill the gap.

I would make a different split, as there are two other candidate explanations. Each derives from the evolutionary biology insight that females seek high-quality genes, which they cannot directly observe, and they desire resources for their offspring.

The first explanation is also a signalling explanation, but in this case relates to signalling by men to women. When women are seeking a mate and trying to assess male quality, a man’s income (often displayed in the form of conspicuous consumption) is one of the most obvious signals. Income and conspicuous consumption are not perfect signals of quality, and there is likely to be noise in the signal through sources such as luck and men lying, but it is a useful proxy. Given that income is not a perfect signal of quality, female use of income as a proxy for quality would result in a higher marriage premium than would be expected on the basis of quality alone.

In relation to resources, women care about income in itself as they can direct those resources to raising offspring. Even if the female knows that he is an idiot who received his wealth through inheritance, that inheritance still has some value.

As a result, I would suggest that human capital accounts for a smaller proportion of the marriage premium than Caplan suggests, with mate signalling and resource acquisition filling the gap. I would also put a slightly different bent on the ability bias explanation by noting that women seek in men many qualities that employers favour, not only because are these are desirable attributes in a marriage partner, but because employers favour them and they can increase income.

Having made these claims, I should address the study by Ginther and Zavodny that suggests the ability bias and resource acquisition explanations have little power in explaining the male marriage premium. They use the interesting approach of comparing normal marriages with “shotgun” marriages, where a child was born within seven months of the date of marriage. If the likelihood of shotgun marriage occurring is not related to income or ability due to their unforeseen nature, we would expect no marriage premium to exist in those cases. However, the authors show that shotgun marriages also have an income premium and the premium is not significantly different from that in normal marriages. As a result, the premium must not be due to underlying quality.

The authors’ finding rests on this clever shotgun wedding mechanism, but ultimately it is flawed. One obvious problem noted by the authors is that many shotgun weddings were planned before the conception or would have occurred regardless of the pregnancy. In those cases, the pregnancy is not the primary reason for the marriage or only affected its timing.

The authors also address another problem, which is the issue of whether the occurrence of a shotgun wedding is correlated with the man’s income. First, they note studies that show that a man’s income does not affect the chance of a couple legitimating a birth. They also ask whether men who marry following a premarital conception are more desirable than men who do not, and note that there is no substantive difference in quality or income.

This explanation misses an important element. The decisions to mate with a man in the first place, to take less caution with contraception, and to keep the child are all likely to be contingent on the quality of the male and their income. If women prefer to mate with high quality, high-income men before marriage and to keep their children, shotgun weddings will not provide an independent comparison.

To disentangle these effects, the first place I would look is income of men before and after marriage. My rough understanding of the literature is that they are highly correlated, which is not what you would expect in a pure signalling to the employer model.

Another interesting area we could look at is the quality of the wives. We need to remember that these are two-sided markets. If women are attracted by income, and men buy the best they can afford, we would see a strong correlation between husband income and wife quality. We would expect to see this correlation under the signalling to females, resource acquisition and ability bias explanations. Under the signalling to the employer theory, we would see no link between male income and wife quality, whereas the human capital explanation would be only dependent on productive qualities of the wife, such as conscientiousness. We would also find that the male marriage premium is smaller at the low-quality female end of the scale.

*Caplan has also written a post on the marriage penalty for women.