Weighing in on the Debate: Why Do Economists Urge College but Not Marriage?

by Barbara Ray

A debate is bubbling up about why economists focus on college as an earnings boost and not marriage, sparked by Megan McArdle at the Daily Beast, Pascal Emmanuel-Gobry at Fortune, and economist Bryan Caplan . When economist talk about options for reducing poverty and bolstering the middle class, they often note the strikingly higher incomes of those with a college degree compared with those with just a high school degree. This “college premium” has been growing since the 1980s, although it has recently flattened out. Meanwhile, the wages of those with the least education are losing ground to inflation. So it seems logical to promote higher education as a route to more secure futures.

But as those of you following this blog know, the Knot Yet findings show that another way to boost earnings is through marriage. The data show that for men at least, getting married is associated with both greater personal and household income. Interestingly, for women, marriages is not associated with does either personal or household income increases. (click on the charts to enlarge).

Figure15

Figure7Figure8

 

So why, some are asking, do we focus so much on education and not more on marriage as well for men at least?

I asked two economists who specialize in U.S. poverty (University of Michigan’s Sheldon Danziger) and labor markets (Georgetown’s Harry Holzer). Here’s their answer:

Harry Holzer (via email):

“Here are a few thoughts:

1) While both sets of estimates potentially suffer from selection bias, the evidence on the causality of education is stronger – David Card and others have presented pretty strong evidence on education, while I am not aware of really strong IV [instrumental variable] estimates for marriage.

2) The policy implications are also clearer for education. They are not perfect, but we have some sense of what might work: effective remediation, financial aid, affirmative action, better information (to lower-income populations), better K-12 efforts, stronger in-college supports. For marriage, it is much less clear. The experiments funded by the Bush 43 administration on marriage generated very few successes.

3) Economists do tend to overstate the education effect; the average estimated effect, even if accurate, might not hold for the marginal noncollege student or worker. I think we have overstated the benefits of 4-year college relative to other remedies, such as CTE [career technical education] and apprenticeships and other forms of workforce development, as well as direct demand/job creation strategies. And, given the very low completion rates for disadvantaged folks whom we send to college (often with Pell grants), maybe we should diversify our education/training strategies a bit more.

4) I also agree with whoever wrote in the [Daily Beast] article that marriage is widely viewed as a very personal decision, especially since the 1960s, and folks fear anything that might demonize or delegitimize single parents.

Sheldon Danziger (via email):

As usual, Harry has the correct answers. The only point I would make is that we have no research evidence on what government can do to promote marriage. There are some recent randomized demonstrations by Mathematica & MDRC–of programs that were launched under Bush the Younger: “Building Strong Families & Supporting Healthy Relationships.” The effects are very weak.

And, as Harry knows, MDRC has solid evidence on career academies, a program for high school students to encourage high school graduation and the transition to jobs. One result of the randomized trial is that for men it led to a substantial increase in the share married at age 25. So a program that increases education raises employment, earnings, and marriage.

At best, a program that increases marriage means that some kids are less poor because they live with a low-income mother and a low-income father. This is better than living with only one parent, but it is not clear that it is the most effective way to reduce child poverty.