Sunday, August 06, 2006

The Marriage Market for Men Without BAs

The NYTimes publishes an article about the increasing tendency of men, particularly the less educated, to reach middle-age without marrying as part of its continuing series on the “new gender divide.” While the article presents several interesting explanations for how the marriage market has changed over the past 25 years, it is oddly organized and seems unfocused. The headline (Facing Middle Age With No Degree and No Wife) and first several paragraphs suggest that the article is about the differential growth in the propensity of men without college education to remain unmarried into their forties; however, the article devotes most of its space to factors like fears of divorce, comfort being alone, or difficulty finding the right person that likely affect everyone, and the authors do not explain why it is likely that these trends are more prevalent among less educated men.

The figure below presents the key descriptive facts that motivate the article.

Three trends stand out in the figure:
1) Among men and women in all education categories, the propensity to remain unmarried increased sharply between 1980 and 2004 (college women got an early start on this, but everyone else caught up during this period)
2) Within all education categories, the propensity to remain unmarried into their 40s increased more for men than for women.
3) Among both men and women, the likelihood of not marrying before 40 increased more for the less educated.
The setup of the article suggests that the discussion will cover the intersection of 2 and 3 – the decreasing likelihood that men without college degrees will marry before 40. As such, I would think that the authors would want to focus on explaining why, over the past 25 years, less educated men are more likely to remain unmarried than highly educated men and similarly educated women.

[An aside for students – to clearly establish the core motivating fact empirically one would want individual data for a large and representative sample of men and women between 40 and 44 years old in at least two periods (e.g., 1980 and 2004). Then you would want to run a regression of whether or not the individual was married on their gender, education, the interaction of gender and education, and other controls (like race, region, urban, etc.) separately for each sample. The authors of the Times article expect the interaction between male and low education categories to have grown more negative between the 1980 sample and 2004 sample (technically, one could do this in one regression by adding an additional variable for the sample and interacting it with the other stuff, but too many interactions confuse me – so I’d probably do it this way).]

The first part of the article does this. The authors outline how demand for less educated males in the marriage market fell as more women went to college (and presumably developed the taste for more educated males that accompanies this) and as the economic prospects of low education males plummeted (making them less desirable spouses for even those within their education category). The authors further speculate that weakening economic prospects induced men without college degrees to drop out of the marriage market (reducing supply) because their careers are hectic and unstable (and they fear these traits would carry over into their marriages).

These explanations work. They provide reasons, specific to low educated males, that explain why both demand and supply for them in the marriage market have fallen in the past 25 years.

The last two pages of the four page article, however, drift into general explanations for the decline in marriage propensity in all populations, and, as such, generate confusion. The authors talk about growing fears of divorce, general lack of desire to marry, and the inability to find someone to marry; however, none of these explanations seem peculiar to low educated males. These seem like general explanations for all groups increasing unwillingness to marry, and, relative to other factors, seem like relatively minor shocks compared to things like birth control and improved labor market opportunities for women that substantially reduced the returns to marriage.

On a slightly related note, I would love to see the BLS add some sort of question about active search for a spouse (or spouse like entity) to the CPS so that we could compute an unmarried rate analogous to the unemployment rate. This would give people like me who enjoy studying marriage markets a lot more interesting data to work with about entry and exit into the market (and its correlates).

Comments: Post a Comment

Subscribe to Post Comments [Atom]

<< Home

This page is powered by Blogger. Isn't yours?

Subscribe to Posts [Atom]