Recent declines in US marriage are reflected both in delayed marriage and increases in permanent singlehood, punctuated by intermittent spells of nonmarital cohabitation. One argument is that the traditional economic foundations of marriage have been eroded by a deteriorating job market, a consequence of automation, deskilling, deunionisation, and global competition for cheap labour.
Indeed, sociologist William Julius Wilson’s (1987) “marriageable male” hypothesis provides a useful theoretical and empirical benchmark, claiming that declines in marriage are driven at least in part by reductions in employment prospects and earnings among men, especially less-skilled racial and ethnic minorities at the bottom of the education distribution.
High rates of incarceration and substantial out-marriage to White women, especially among Black men, have also left many minority women without marital partners. The fact that women’s educational levels now exceed men’s further implies that young women—by necessity—are less financially dependent on husbands than in the past and that educational hypogamy has become more commonplace. Young women seemingly face shortages of demographically similar men to marry.
This article provides new estimates of spousal mismatches in the marriage market. Specifically, we compare the demand-side sociodemographic characteristics that women typically seek in male partners with the availability or supply of these characteristics in the marriage market. We use methods for imputing missing data (in effect, creating “synthetic husbands”) to infer the likely sociodemographic profiles of the husbands of unmarried women if they married. We make the assumption that these women would marry men comparable with the husbands of demographically similar women who are currently married.
We accomplish our goals using national and subnational data from the most recently released cumulative 5-year files (2008–2012 and 2013–2017) of the annual American Community Survey. By identifying the counterfactual case (i.e., the likely demographic profile of husbands if unmarried women became married), we provide a direct assessment of whether women currently face demographic constraints in the marriage market. Our study—for the first time—identifies both surpluses and deficits of so-called synthetic husbands in the marriage market.
This didactic exercise shows that unmarried women face overall shortages of economically attractive partners with either a bachelor’s degree or incomes of more than $40,000 a year. Most previous work suggests that women are more likely to remain unmarried than to “settle” by marrying partners who are mismatched on age, education, or race.
A recent study by Qian (2017), however, indicated that patterns of assortative mating have shifted, switching from a tendency in 1980 for women to “marry up” in socioeconomic status to a pattern today of “marrying down.” This reversal suggests, at a minimum, that growth in the pool of marriageable men has not kept pace with the rapid rise in women’s socioeconomic status. Our study reinforces the commonplace view that women today face new marriage trade-offs at a time when finding a suitable marital match has become more difficult.
Our overall goal is largely descriptive: to appropriately characterise US marriage market conditions for currently unmarried women with different sociodemographic profiles. We have two specific objectives.
First, we use data imputation methods to infer what the sociodemographic characteristics of each woman’s spouse would be if they married a man with similar characteristics to the husbands of comparable women.
Second, we compare the distribution of characteristics of synthetic husbands with the distribution of all unmarried men in our sample. The goal is to identify the shares of women without a suitable marriage match and the specific female subpopulations that face the greatest risk of a “tight” marriage market—one with a demographic shortage of men to marry.
Our discussion of marriage market imbalances focuses primarily on (a) low-educated or poor women who are sometimes the target of recent marriage promotion programmes and (b) highly educated women who have ostensibly “priced” themselves out of the marriage market and now face shortages of economically attractive men to marry. Or, stated differently, men may have become less competitive in the marriage market, falling behind on those economic and demographic traits that made them attractive to women as marriage partners.
Matching spousal characteristics
Table 1 provides summary statistics for each of the key variables used in our matching exercise reported separately for each of the four groups. Married men and women are on average older than their unmarried counterparts, and they have higher education levels. Unmarried women are slightly more likely to be employed but earn slightly less than their married counterparts.
Table 1: Summary statistics
These observed similarities and differences are largely consistent with the conventional wisdom that married men are more “economically attractive” or “marriageable” than unmarried men and that most single women (by definition) must rely on their own employment and earnings to support themselves and their families.
For example, the average total personal income of married men is $70,000 compared with $35,000 for unmarried men (measured in 2017 dollars). Thirty-seven per cent of married men are college graduates compared with only 25 per cent of unmarried men. Although the difference is small in absolute terms, the relative difference in employment status is large. About twice as many unmarried as married women are unemployed (7.05 per cent vs. 3.79 per cent). The largest relative difference between married versus unmarried women is the percentage Black (6 per cent vs. 18 per cent), a result that highlights the persistent marriage gap between Blacks and Whites.
Imputing synthetic spouses
The key empirical goal is to determine the characteristics of the spouse to whom the unmarried women in our sample would likely be married, assuming they exhibit the same mate selection patterns as currently married women.
We estimate the characteristics of synthetic spouses using two alternative approaches. Our first approach is to use a standard hot-deck imputation in which we randomly draw a spouse out of the set of possible matches and repeat this process for all unmarried women in our samples. The second imputation approach takes the average of each characteristic across the set of possible matches for each unmarried woman. We then use these averages to estimate the characteristics of each synthetic spouse.
Results: Baseline estimates of marital mismatch
The synthetic spouses (Table 2) had an average income that was about 55 per cent higher ($53,000 vs. $35,000), were 26 per cent more likely to be employed (87 per cent vs. 70 per cent), and were 18 per cent more likely to have a college degree (29 per cent vs. 25 per cent) than the actual unmarried men who were available in the United States. These estimates suggested large differences in the demand and supply of unmarried men with certain characteristics.
Table 2: Comparison of Synthetic Spouses and Unmarried Men
In Figures 1 and 2, we overlaid the distribution of age, income, education, and race of the synthetic spouses and the unmarried men observed in these data.
Figure 1: Comparison of distribution of synthetic spouses and actual unmarried men using hot-deck imputation
Figure 2: Comparison of distribution of synthetic spouses and actual unmarried men using mean imputation
The results in both Figures 1 and 2 clearly highlighted large income- and education-based mismatches in the marriage market. Specifically, there was an excess supply of men with incomes less than $20,000 (with a shortage of men with incomes greater than $40,000) as well as a marriage market mismatch in education—too many men had only a high school degree and too few had a college or graduate degree. However, there was some evidence in previous studies that fathers who marry their child’s mother may, as a result, experience increases in income.
To the extent that this happens, our estimates of the shortage of higher earning men may be slightly inflated, but nevertheless still cannot fully explain the magnitude of the male shortage.
In contrast to these estimates, the racial distributions were well matched, except for the possible oversupply of unmarried Black men, a pattern clearly consistent with Wilson’s (1987) “marriageable male” hypotheses. Because less-educated racial and ethnic minorities have disproportionately high rates of incarceration, the evidence here of well-matched racial distribution seemingly indicated that any effects of the mass incarceration of Blacks on the overall marriage market were modest, a result consistent with the results reported by Lopoo and Western (2005).
In Table 3, our results showed how women’s sociodemographic characteristics jointly determined whether they experienced a demographic shortfall of unmarried men—those with a demographically suitable bundle of characteristics.
Table 3: Characteristics that predict whether an unmarried woman is likely to have a potential match
The overall results in Table 3 indicated that younger women and less-educated women were more likely to find demographically suitable potential marital partners available to them. Conversely, older and highly educated women were most likely to face shortages of marital partners. This finding was consistent with other related empirical evidence that sex-ratio imbalances increase with women’s age and that the gender reversal in educational attainment has upended traditional patterns of educational hypergamy among American women.
Race also placed constraints on marital opportunities. For example, within states, Black women were 15.01 percentage points less likely to have a suitable match. Asian women were 3.50 percentage points less likely to have a match. The difficulty in finding a match was larger within PUMAs than within states, especially among Asians (b = −27.23).
Discussion and conclusion
Our analyses provide clear evidence of an excess supply of men with low income and education and, conversely, shortages of economically attractive unmarried men (with at least a bachelor’s degree and higher levels of income) for women to marry.
One implication is that promoting good jobs may ultimately be the best marriage promotion policy rather than marriage education courses that teach new relationship skills. Of course, other policy efforts aimed at securing women’s economic independence (i.e., equal pay legislation) are also important in the case of single mothers who often face constraints on marital search behavior and have limited prospects for “marrying up”.
Our estimates of marriage market disequilibria are instructive, especially at a time when marriage is sometimes viewed as an economic panacea. In the case of unmarried minority women, for example, shortages of highly educated unmarried men also impose serious constraints on the marital search process. Black women, for example, are about 17 percentage points less likely than White women to have a match in their local marriage market area (PUMA).
Our findings also make the case that highly educated White women face shortages of marriageable men. For highly educated women, the marriage market implications of new gender imbalances in educational achievement seem increasingly clear. They will either increasingly remain unmarried or, alternatively, conventional patterns of marital educational hypergamy (i.e., women marrying up in education) may give way to educational hypogamy as women adapt to deficits in the pool of highly educated men.
This study is not without some limitations. For example, we acknowledge that there are unmeasurable selection factors that may differentiate married women from unmarried women. Our results should therefore be interpreted to indicate what the marriage market should look like if all women were to have a plausible match, regardless of their inclination toward marriage.
A large share of adolescents and young adults today expect to marry, and this is little changed from previous generations. This makes clear that most women—Black or White, rich or poor, highly educated or uneducated—have “high hopes” for marriage, yet growing shares of women today either delay marriage or remain unmarried altogether.
Our study uncovers the demographic reality of large deficits in the supply of men who are suited or well matched for today’s unmarried women. If nothing else, our empirical results indicate that the US marriage market is currently in disequilibria. The supply of unmarried men is out of demographic balance with the demand for marriageable men among America’s currently unmarried women.
Whether this is new or different from past generations is unclear, as is the question of whether marriage market mismatch is fully or partly responsible for the ongoing “retreat from marriage.” What is clear is that the characteristics of potential spouses—male and female—have become more diverse over time with rising educational levels among women, increasing racial diversity, and new patterns of income and educational inequality.
Daniel T. Lichter is Professor of Policy Analysis and Management and Sociology at Cornell University, US.
Joseph P. Price is an Associate Professor of Economics at Brigham Young University, US.
Jeffrey M. Swigert is Lecturer of Economics and Finance at Southern Utah University, US.
This is an edited extract from the authors’ paper, Mismatches in the Marriage Market, which was first published on Wiley Online Library, and is being reproduced here with permission.
Why news media is in crisis & How you can fix it
India needs free, fair, non-hyphenated and questioning journalism even more as it faces multiple crises.
But the news media is in a crisis of its own. There have been brutal layoffs and pay-cuts. The best of journalism is shrinking, yielding to crude prime-time spectacle.
ThePrint has the finest young reporters, columnists and editors working for it. Sustaining journalism of this quality needs smart and thinking people like you to pay for it. Whether you live in India or overseas, you can do it here.