The Housing Theory of Marriage

Marriage, as IFS readers know, is in decline. That decline has not been equitable, with marriage dwindling far more among poorer Americans than the rich. In other words, marriage has become far more correlated with wealth—what was once a universal institution is now the reserve of the rich.

Is that relationship only correlative? Or does wealth protect against marriage-lessness? A new paper, published in the American Economic Journal, floats a novel theory: the value of a home works as a kind of collateral, making marriage more appealing than it otherwise would be.

The paper’s authors, Jeanne LaFortune and Corinne Low, start with the observation that marriage has grown less common among the poor than the rich. Specifically, they note, Americans who rent their homes are substantially less likely to have ever married than Americans who own. As late as 1940, they find, roughly 95% of both renters and owners ages 30 to 50 had married. Today, the figure is around 85% for owners but has plummeted to about 65% for renters.

This observation could reflect underlying differences between renters and owners. But, LaFortune and Low argue, it may also reflect the changing nature of marriage, and particularly the nature of the marital contract.

People tend to think about marriage as a matter of love and affection, but it’s also an economic institution. One way this is true is that in a marriage, both partners can split the home and labor-force work equally, but they can also specialize, yielding the sort of returns the division of labor often yields for market firms. If one partner (often mom) stays home and takes care of the kids, while the other (often dad) focuses on career, both can benefit more than if they divide labor evenly.

That arrangement, though, involves risk to the partner who foregoes working to benefit the shared project (what LaFortune and Low call a public good) of the family. If mom stays home for 15 years, only for dad to leave one day and take his earning potential with him, Mom is left in the lurch. She has neither savings nor, more importantly, accrued experience that she can use to get her own job.

Historically, this dilemma was resolved by strong legal constraints on marriage. But those have eroded. The advent of no-fault divorce makes marriage riskier for the specialized-to-home partner. Child support payments and legal parental rights for non-married fathers, meanwhile, have made non-marital fertility comparatively less risky. Why get married when dad could cut out on you at any moment, especially if he will have to pay for junior regardless?

This is where home ownership comes in. A home is a large, illiquid, hard-to-hide asset that partners divide the value of in the case of divorce. (Indeed, in many cases the custodial parent—usually mom—gets the house.) This, LaFortune and Low argue, serves as a kind of collateral, a guarantee to the home-specializing partner against marital “default” in the form of divorce. Mom can safely stay at home knowing that if dad up and leaves, she retains a huge portion of his earnings in the form of the home.

LaFortune and Low test this “collateralized marriage” model in part by looking at the relationship between housing prices and spouses’ specialization. Their method for doing this is complicated, using a number of public surveys to estimate various figures. Across the board, LaFortune and Low look at how housing prices at age 25 (the modal age of first marriage in their data) affect home ownership, and how that in turn affects household members’ workforce behavior. 

They find the expected effect: a 1 percentage point increase in housing prices at the time of marriage corresponds to a 3-point decrease in homeownership, and a 1-point increase in the probability of women working outside the home. Under certain aggressive assumptions, they estimate that a household going from not owning to owning increases male work by about 25 hours and reduces female work by about 20 hours—consistent with the division of labor theory.

These effects hold up across a variety of measures. For example, LaFortune and Low find that lower home prices at marriage reduce women’s wages, and increase time spent at home/reduce time spent at work for women in time use surveys. They even find that lower home prices affect fertility: a 1 percentage point increase in home prices reduces the number of children by at least 0.07. LaFortune and Low also look at historic data and argue that the advent of no-fault divorce increases the association between asset holding and marriage.

If LaFortune and Low’s model holds up—and they give enough evidence to suggest that it’s at least part of the puzzle—then it has several implications, both descriptive and policy.

On the descriptive side, it is more support for the view that changing labor market conditions, and particularly men’s declining earning potential relative to women, drives the unequal decline in marriage. This “marriageable men” hypothesis is not new. But LaFortune and Low show how marriageability can be determined not just by earnings, but by wealth, and how concerns about earnings can be allayed by joint investment.

More interesting, though, is the policy implication. LaFortune and Low argue that the historical marriage contract—no unilateral divorce, no paternity rights outside of marriage, etc.—helped keep historic marriage rates high. That arrangement is probably not coming back. But if home prices affect tendency to marry, then home prices are a plausible mechanism for increasing marriage, and thereby increasing fertility and potentially reducing inequality.

That high home prices produce many social harms is not news. But if politically popular policies focused on reducing home prices, like zoning reform or the legalization of multi-family housing, can also increase marriage, that adds a powerful policy tool to the often too-small arsenal of those who want to make marriage less rare. 

Charles Fain Lehman is a fellow at the Manhattan Institute and a contributing editor of City Journal.