The Growing Racial and Ethnic Divide in U.S. Marriage Patterns

Summary

The United States shows striking racial and ethnic differences in marriage patterns. Compared to both white and Hispanic women, black women marry later in life, are less likely to marry at all, and have higher rates of marital instability.

Kelly Raley, Megan Sweeney, and Danielle Wondra begin by reviewing common explanations for these differences, which first gained momentum in the 1960s (though patterns of marital instability diverged earlier than patterns of marriage formation). Structural factors—for example, declining employment prospects and rising incarceration rates for unskilled black men—clearly play a role, the authors write, but such factors don’t fully explain the divergence in marriage patterns. In particular, they don’t tell us why we see racial and ethnic differences in marriage across all levels of education, and not just among the unskilled.

Raley, Sweeney and, Wondra argue that the racial gap in marriage that emerged in the 1960s, and has grown since, is due partly to broad changes in ideas about family arrangements that have made marriage optional. As the imperative to marry has fallen, alongside other changes in the economy that have increased women’s economic contributions to the household, socioeconomic standing has become increasingly important for marriage. Race continues to be associated with economic disadvantage, and thus as economic factors have become more relevant to marriage and marital stability, the racial gap in marriage has grown.

Today’s racial and ethnic differences in children’s family experiences are striking. In 2014, 70 percent of non-Hispanic white children (ages 0–18) and roughly 59 percent of Hispanic children were living with both of their biological parents. The same was true for only a little more than one-third of black children.¹ Although many children raised in single-parent households thrive and prosper, at the population level, single-parent families are associated with poorer outcomes for children, such as low educational attainment and teen childbearing.² Some social scientists argue that single-parent families may harm children’s development directly, by reducing fathers’ and mothers’ ability to invest in their children. Others suggest that common factors, such as economic distress, contribute both to family instability and to developmental problems in children.³ That is, in this view, family structure itself is not the source of children’s disadvantages. Regardless, even if many single-parent families function well and produce healthy children, population-level differences in family stability are associated with distress for both parents and children.

To explain racial and ethnic variation in children’s families, we must better understand the differences in marriage patterns across groups. We begin by describing racial and ethnic differences in marriage formation and stability, then review common explanations for these differences. We also discuss how these gaps have evolved over time and how they relate to social class. To date, many explanations have focused on the poor and working class, even though racial and ethnic differences in family formation exist across the class spectrum. We argue that the racial gap in marriage that emerged in the 1960s, and has grown since, is due partly to broad changes in ideas about family arrangements that have made marriage optional (but still desirable). As the imperative to marry has fallen, alongside other changes in the economy that have increased women’s economic contributions to the household, socioeconomic standing has become increasingly important for marriage. Race continues to be associated with economic disadvantage, and thus as economic factors have become more relevant to marriage and marital stability, the racial gap in marriage has grown.

Although we primarily focus on black-white differences in marriage, we also consider contemporary family patterns for other racial and ethnic groups (Hispanics, Asians, and Native Americans). New waves of migration have added to the diversity of the United States, and blacks are no longer the largest minority group. Moreover, considering the family patterns of other minority groups, whether disadvantaged or comparatively well-off, can give us insight into the sources of black-white differences. Our ability to analyze historical marriage trends among Hispanics, however, is limited due to changing measurement strategies in federal data, shifts over time in the characteristics of migrant populations, and the fact that the marriage patterns of migrants differ from those of U.S.-born Hispanics.

Black-White Differences in Marriage and Marital Stability

Young adults in the United States are waiting longer to marry than at any other time in the past century. Women’s median age at first marriage currently stands at 27, compared to a median marriage age of 24 as recently as 1990 and a low of just over 20 in 1955.⁴ Although social scientists debate whether today’s young people will eventually marry in the same numbers as earlier generations, marriage remains commonplace. In 2013, more than eight women in ten in their early 40s were or had ever been married.⁵

Contemporary Differences

At the same time, racial and ethnic differences in marriage are striking. The median age at first marriage is roughly four years higher for black than for white women: 30 versus 26 years, respectively, in 2010.⁶ At all ages, black Americans display lower marriage rates than do other racial and ethnic groups (see table 1, panel A). Consequently, a far lower proportion of black women have married at least once by age 40. Our tabulations of data from the U.S. Census Bureau’s American Community Survey for 2008–12 show that nearly nine out of 10 white and Asian/Pacific Islander women had ever been married by their early 40s, as had more than eight in 10 Hispanic women and more than three-quarters of American Indian/Native Alaskan women. Yet fewer than two-thirds of black women reported having married at least once by the same age.

Table 1.

Women’s Age-Specific Rates of First Marriage and Divorce by Race, Ethnicity, and Nativity

	Panel A. Marriage
Age	White	Black	Asian/Pacific Islander	American Indian/Native Alaskan	Hispanic, Total	Hispanic, U.S. born	Hispanic, foreign born
15–19	8.7	5.0	8.5	20.3	16.7	13.1	32.6
20–24	58.9	23.0	41.4	53.5	59.1	50.4	81.3
25–29	115.6	43.0	133.7	76.6	81.0	75.9	89.2
30–34	130.6	47.6	152.5	74.9	87.4	83.0	92.1
35–39	123.0	44.6	129.1	70.5	80.4	72.7	86.8
40–44	111.6	39.4	100.5	51.8	77.9	72.6	82.2
	Panel B. Divorce
Age	White	Black	Asian/Pacific Islander	American Indian/Native Alaskan	Hispanic, Total	Hispanic, U.S. born	Hispanic, foreign born
20–24	48.44	40.13	12.23	63.61	26.79	36.74	16.13
25–29	38.80	44.29	13.23	52.02	26.71	40.43	15.31
30–34	31.60	44.43	15.95	40.15	25.03	37.09	16.83
35–39	29.66	41.20	12.98	41.58	23.70	36.31	16.43
40–44	26.33	38.86	13.07	48.60	21.47	30.15	16.78

Source: Authors’ computations from the 2008–12 American Community Survey, Integrated Public Use Microdata Series.

Note: Rates are calculated as the number of marriages per 1,000 unmarried women and number of divorces per 1,000 married women.

In addition to later age at first marriage and lower proportions ever marrying, black women also have relatively high rates of marital instability (see table 1, panel B). At nearly every age, divorce rates are higher for black than for white women, and they are generally lowest among Asian and foreign-born Hispanic women.⁷ Recent demographic projections suggest that these racial and ethnic gaps in marriage and marital dissolution will continue growing.⁸

Thus far we’ve relied primarily on data from the U.S. Census and other similar sources (for example, the American Community Survey). These sources offer historical continuity and large sample sizes, but they generally offer only limited information about women’s marital histories and background characteristics. Moreover, they almost certainly underestimate the size of racial gaps in marital instability, as black women tend to transition more slowly than white women do from separation to legal divorce.⁹ For our final look at contemporary marriage patterns, we now turn to a smaller data set, the National Survey of Family Growth, to get a better sense of how women’s accumulated life experiences of marriage vary across race, ethnicity, and nativity. This data set contains retrospective histories on the formation and dissolution of cohabiting and marital relationships for a nationally representative sample of women aged 15–44. Table 2 displays these results.

Table 2.

Women’s Marital Life Profiles at Ages 40–44: Percentage with Life Histories of No Marriage, Stable Marriage, or Unstable Marriage

	All Women					Percentage of Ever-Married Women Experiencing Unstable Marriage	Percentage of Unstably Married Women Who Have …
			Unstable Marriage
Race, Ethnicity, and Nativity	No Marriage	Stable Marriage	Total	Married Only Once	Married 2+ Times		Married Only Once	Married 2+Times
White, non-Hispanic	7	54	38	16	23	41	41	59
Black, non- Hispanic	34	29	35	21	15	53	58	42
Hispanic, total	14	48	39	18	21	45	46	54
Hispanic, foreign born	11	48	41	19	21	46	48	52
Hispanic, U.S. born	21	46	34	15	19	42	43	57

Source: Author’s calculations from 2006–10 National Survey of Family Growth.

Consistent with other sources, we again see lower levels of marriage among black women than among white or Hispanic women. Among those who do marry, black women experience more marital instability than do white or Hispanic women. About 60 percent of white women who have ever married are still married in their early 40s, compared to 55 percent of Hispanic women but only 45 percent of black women. After accounting for women who have never married at all, then, roughly half of white and Hispanic women in their early 40s are stably married, compared to less than a third of black women the same age. The nature of instability also varies by race: Among women who’ve experienced any marriage that ended (in table 2, our “unstable marriage” group), black women are more likely to have been married only once (58 percent, versus 42 percent who have been married two or more times), whereas white women are more likely to have married multiple times (59 percent, versus 41 percent who married only once.)

Historical Trends

Although social scientists sometimes attribute racial differences in family patterns to long-run historical influences such as the legacy of slavery, marriage was common among black families in the early 20th century.¹⁰ Thus the racial divergence we see now in marriage formation is relatively recent. From 1890 through 1940, black women tended to marry earlier than white women did, and in the mid-20th century first marriage timing was similar for black and white women.¹¹ In 1950, black women aged 40–44 were actually more likely to have ever married than were white women of the same age (figure 1). Racial differences in marriage remained modest as recently as 1970, when 94.8 percent of white women and 92.2 percent of black women had ever been married.¹²

Figure 1. Percentage of U.S. Women Aged 40–44 Years Who Had Ever Married, by Year, Race, and Ethnicity.

Source: 1930–2000 U.S. Decennial Census and 2012 American Community Survey, Integrated Public Use Microdata Series.

The likelihood of ever marrying by midlife (which we define as age 40–44) conveys important information about the nature of group differences in marriage, yet these figures reflect age-specific marriage rates that prevailed at earlier points in time. If we understand the historical timing of the racial divergence in marriage rates with greater precision, we may shed light on what caused the change and variability in family patterns. Sociologists Robert Mare and Christopher Winship report that during the 1960s, marriage rates began to decline much more rapidly for black women than for white women across all age groups.¹³ Thus looking at age-specific marriage rates suggests that the racial divergence in marriage patterns gained momentum about 10 years earlier than figure 1 suggests, after about 1960.

Although before the 1960s age at first marriage and the proportion of women ever married were similar among whites and blacks, blacks had higher rates of marital dissolution during this period. If we examine the percentage of ever-married white and black women who were currently married and living with their husbands at midlife, the historical story about trends in the racial marriage gap changes somewhat. Figure 2 displays these results. We now see large racial differences in the likelihood of being married even as early as 1930, when only 69 percent of ever-married black women in their early 40s were married and living with a spouse, compared with roughly 88 percent of white women the same age. Some of this difference reflects higher rates of mortality among black men, but some is due to higher rates of separation. In the early 1900s, very small percentages of women, whether black or white, were officially divorced. Somewhat more were married but not living with their spouses, though the percentage was small by today’s standards. Still, the proportion was twice as high for black women as for whites.¹⁴ Between 1940 and 1980, both white and black women experienced large increases in divorce, but the increase occurred sooner and more steeply for black women.¹⁵ By 2012, roughly 73 percent of white women in their early 40s who had ever married were still married and living with their spouses, compared with just over half (52.7 percent) of black women the same age.¹⁶

Figure 2. Percentage of U.S. Women Who Are Currently Married, Spouse Present, by Year, Race, Ethnicity: Women Aged 40–44 Who Had Ever Married.

Source: 1930–2000 U.S. Decennial Census and 2012 American Community Survey, Integrated Public Use Microdata Series.

In short, we can learn much from taking a longer-run view of the black-white marriage gap. We see that the racial gap in marriage formation was minimal through about 1960, both in terms of marriage ages and rates, but that the higher rate of marital instability among black than among white women has deeper historical roots. Divorce rates increased earlier and more steeply among black than among white women. After about 1970, we see marital instability continue to diverge between black and white women, but we also begin to see a new racial gap in the likelihood of ever marrying, driven by a decline in marriage formation among blacks. As we’ll see below, when we explore variation by social class, a similar pattern has appeared more recently among less-educated whites.

Explaining the Black-White Marriage Gap

Social scientists can’t fully account for the racial and ethnic differences in marriage, even though these differences have been intensely debated for decades. Given the large differences between them, marriage patterns of white and black women have been of particular interest. Empirical research best supports explanations for the black-white marriage gap that involve labor market disparities and other structural disadvantages that black people face, especially black men. These explanations are rooted in classic demographic arguments about the affordability of marriage and about imbalances in the numbers of men and women available for marriage.¹⁷

In their highly influential 1987 book The Truly Disadvantaged, sociologists William Julius Wilson and Kathryn Neckerman hypothesized that black women’s low marriage rates in the 1970s and 1980s were due to a deficit of marriageable men.¹⁸ An enormous decline in unskilled manufacturing jobs during the 1970s and 1980s hit black men particularly hard.¹⁹ The black-white unemployment gap grew rapidly, and by 1985 unemployment rates for black men aged 25–54 were two times higher than for white men in the same age range. Among men aged 16–24 the racial disparity was even greater, with the unemployment rate for black men three times that of white men.²⁰ Black men were also much more likely to die or be incarcerated, and this (combined with low rates of interracial marriage) depressed the number of men available for black women to marry. Unemployment rates for black men continue to be much higher than for white men, and black men’s rates of incarceration have increased dramatically since 1980, suggesting that these factors are still relevant today. Indeed, in the early 2000s, more than one-third of young black men who hadn’t attended college were incarcerated, and nearly twice as many black men under age 40 had a prison record than a bachelor’s degree. Overall, black men are seven times more likely than white men to be incarcerated.²¹

Yet men’s demographic availability, unemployment, and low earnings don’t completely explain black-white differences in marriage.²² Moreover, black marriage rates fell at the same time that racial discrimination was declining and black men’s wages were growing. Between 1960 and 1980, employed blacks saw real increases in wages relative to whites, partly due to increases in their educational attainment and partly because returns to education also increased.²³ During this time, the proportion of blacks who were in the middle class (defined as between 200 and 499 percent of the federal poverty line) increased substantially.²⁴

Not all black men were reaping the benefits of increasing opportunity that came via civil rights legislation. As we’ve seen, black unemployment rates were growing, and the racial disparities are even greater if we account for the high rates of incarceration among less-educated black men.²⁵ Still, the proportion of blacks who are poor is lower today than in 1960, and blacks’ median household income, after adjusting for inflation, is higher.²⁶ Black marriage rates began to fall even while the black middle class was growing, and they continued falling after 1980 even as black men’s unemployment rates and real wages improved (although not relative to white men’s). We’ll return to this problematic mismatch between historical trends in marriage and labor force patterns toward the end of this article.

Other explanations for the black-white marriage gap focus on additional constraints on the availability of partners for black women. For example, women tend to marry partners who have accumulated at least as much schooling as they have.²⁷ Among both blacks and whites in the United States today, young women tend to be more educated than young men.²⁸ This constrains the pools of desirable partners for marriage. But the education gap between men and women is larger for blacks, making this constraint particularly important for black women. Moreover, rates of intermarriage among blacks differ substantially by gender.²⁹ Black men are more than twice as likely as black women to marry someone of a different race.³⁰ This, too, constrains the pool of potential partners for black women.³¹

Finally, some explanations emphasize racial differences in the ratio of men’s to women’s wages, as opposed to men’s earnings alone. A specialization model of marriage suggests that the gains to marriage are greatest when men’s wages are high relative to women’s, so that men can specialize by working in the labor market while women work in the home.³² The ratio of men’s to women’s wages is much smaller among blacks than whites. Thus the specialization model suggests that marriage rates should be lower for blacks. Although family scholars are quick to point out that black marriages have historically been less characterized by specialization, considerable evidence suggests that the expectation that men will provide for their families economically is strong across groups.³³ Yet the ratio of men’s to women’s wages can’t explain lower marriage rates among blacks. Declines in black women’s marriage rates between 1968 and 1996 don’t track changes over time in women’s wages relative to men’s. Marriage rates fell, while the female-to-male wage ratio remained similar across time.³⁴ Moreover, other analyses show that both women’s and men’s earnings are positively associated with marriage and that the positive association between women’s earnings and marriage has been increasing over time, suggesting that the argument that gender specialization supports marriage may be outdated.³⁵

Although differences in men’s (and women’s) employment, earnings, incarceration, and education contribute to the racial gap in marriage, they give an incomplete account. We’ve argued elsewhere that taking a broader view of marriage and how it relates to other social institutions may uncover additional sources of black-white differences in marriage.³⁶ The United States has become increasingly stratified by class, in terms of earnings, wealth, and occupational and residential segregation. Consequently, the sources of racial inequality likely vary by social class.³⁷

Social Class and the Racial Gap in Marriage

If rising unemployment and incarceration among black men fully explained the racial gap in marriage, we would expect racial differences in marriage among people with the same level of education to be small; we would also expect such differences to be concentrated among economically disadvantaged blacks. After all, black men without any college education were affected most by both trends.³⁸ Yet, although the racial marriage gap is largest among those who didn’t go to college, we see a gap at all levels of the educational distribution. For example, among college-graduate women in 2012, 71 percent of blacks had ever married, compared to 88 percent of whites (see table 3). Moreover, while we see differences by education in the proportion of black women in their early 40s who have ever married, there are no clear educational differences among white women. We see a similar pattern in the proportion of men who have ever married, although data from 2012 show some evidence that white men with a high school degree or less are moving away from marriage.

Table 3.

Percentage of Women and Men Ages 40–44 Who Had Ever Married, by Year, Race, and Education

	Women				Men
	1980	1990	2000	2012	1980	1990	2000	2012
	White, Non-Hispanic
Total	95.8	93.4	90.9	87.9	93.9	91.4	86.3	81.6
<=12 years	96.7	95.1	92.4	87.1	94.0	91.4	85.6	77.6
13–15 years	96.0	94.5	91.6	88.9	94.6	92.4	86.6	82.6
16+ years	91.1	89.4	87.8	87.9	93.0	90.5	87.2	85.5
	Black, Non-Hispanic
Total	88.7	83.2	72.8	62.4	88.5	82.6	73.7	65.3
<=12 years	88.4	81.8	70.0	55.8	87.7	79.8	69.5	57.6
13–15 years	91.5	84.9	75.7	64.6	91.3	86.2	79.4	73.1
16+ years	86.9	85.0	77.1	70.9	90.4	86.4	82.9	76.5
	Hispanic, Total
Total	93.3	90.6	88.0	82.7	92.4	89.9	85.4	77.3
<=12 years	93.9	90.4	88.2	81.0	92.4	89.2	85.1	76.0
13–15 years	91.8	92.4	87.9	85.5	92.9	92.3	86.7	79.9
16+ years	87.1	87.8	87.2	85.8	92.2	89.2	85.5	80.8
	Hispanic, Foreign Born
Total	93.1	90.8	89.4	84.7	92.8	90.7	87.9	79.6
<=12 years	93.8	90.2	89.7	83.4	93.0	90.3	87.5	78.7
13–15 years	89.2	94.1	88.7	89.0	91.8	92.5	89.6	82.7
16+ years	90.7	90.6	88.0	88.0	92.0	90.8	88.8	83.0
	Hispanic, U.S. Born
Total	93.4	90.4	86.2	79.6	92.2	89.0	81.8	73.5
<=12 years	93.9	90.6	85.8	75.1	91.9	87.7	80.8	69.7
13–15 years	93.9	91.6	87.3	83.0	93.6	92.1	84.4	77.6
16+ years	82.8	85.6	86.5	84.0	92.4	88.0	82.1	79.0

Source: 1980–2000 U.S. Decennial Census and 2012 American Community Survey, Integrated Public Use Microdata Series.

But, as we’ve argued, looking at the proportion of people who are married by midlife doesn’t capture the most recent changes in marriage patterns among younger women. To overcome this problem, we calculated age-specific marriage rates using data from the 2008–12 American Community Survey (see figures 3a and 3b). Here we see signs that white women with a high school degree or less are beginning to retreat from marriage. Starting in their early 20s, white women with a bachelor’s degree have higher marriage rates than white women with lower levels of education. In fact, marriage rates for college-educated white women in their late 20s and early 30s are higher than those for white women with less education at any age. Their higher marriage rates persist through the peak marrying ages, until their mid-40s. This is a dramatic change from white women’s marriage patterns in the late 1970s, when peak age-specific marriage rates for less-educated women were considerably higher than those ever observed among college-educated women.³⁹ In the near future, the proportion who have ever married at age 40 may fall among white women with less than a college degree, both absolutely and relative to their better-educated counterparts.⁴⁰

Figure 3a. Age-Specific First Marriage Rates, by Education: White Women.

Source: 2008–12 American Community Survey, Integrated Public Use Microdata Series.

Note: Rates are calculated as the number of marriages per 1,000 unmarried women.

Figure 3b. Age-Specific First Marriage Rates, by Education: Black Women.

Source: 2008–12 American Community Survey, Integrated Public Use Microdata Series.

Note: Rates are calculated as the number of marriages per 1,000 unmarried women.

We find further evidence that white women’s marriage patterns diverge by education when we consider marital stability, as table 4 shows. In 2012, the likelihood that ever-married white women were currently married in their early 40s was much lower among the least educated than among the most educated (65.5 percent versus 82.7 percent, respectively). This reflects growing socioeconomic differences in divorce risk, which have also been documented elsewhere.⁴¹ This difference by education in the endurance of marriage among white women is relatively recent, but it has deeper historical roots among black women. Back in 1980, there was no clear relationship between educational level and the likelihood that ever-married white women would be currently married at midlife (see table 4). The story is quite different for black women. Though table 4 again shows that stable marriage is lower overall among ever-married black women than among ever-married white women, within each educational group, marital instability increased earlier and more dramatically among black women with a high school degree or less. Even in 1980, ever-married black women with low levels of education were less likely than the relatively more educated to be married at midlife.

Table 4.

Percentage of Women and Men Ages 40–55 Who Are Currently Married (Spouse Present) among Those Ever Married, by Year, Race, and Education

	Women				Men
	1980	1990	2000	2012	1980	1990	2000	2012
	White, Non-Hispanic
Total	83.9	78.3	77.4	73.5	88.4	82.6	79.2	76.8
<=12 years	84.1	78.3	74.5	65.5	88.1	79.7	73.9	68.2
13–15 years	82.5	76.1	76.0	69.9	88.0	80.9	79.6	76.2
16+ years	84.5	81.1	83.4	82.7	89.4	86.9	87.8	86.4
	Black, Non-Hispanic
Total	55.6	51.5	52.6	52.7	72.9	64.2	61.4	60.5
<=12 years	54.5	49.3	49.5	45.6	71.5	60.9	55.9	53.6
13–15 years	56.6	50.5	53.1	52.3	75.0	65.3	65.8	61.4
16+ years	65.7	60.9	60.9	62.8	80.9	73.4	74.9	74.5
	Hispanic, Total
Total	75.8	68.8	71.2	68.9	83.0	75.8	72.8	73.1
<=12 years	75.4	69.1	71.1	68.6	82.2	74.6	71.3	71.6
13–15 years	77.3	68.1	68.1	64.6	83.4	77.1	74.1	73.8
16+ years	78.3	68.1	76.1	75.6	88.5	79.3	80.1	79.8
	Hispanic, Foreign Born
Total	79.2	72.5	74.7	71.8	83.0	75.1	75.0	75.6
<=12 years	78.7	72.7	75.0	72.3	81.2	73.7	74.1	75.1
13–15 years	83.4	71.3	70.7	66.5	88.5	77.1	77.7	75.5
16+ years	79.6	72.4	77.3	75.5	88.6	81.1	79.7	79.2
	Hispanic, U.S. Born
Total	73.1	65.4	66.8	64.1	83.0	76.6	69.2	68.7
<=12 years	73.0	65.1	64.8	58.3	82.9	75.9	66.0	62.3
13–15 years	72.5	66.4	66.3	63.2	80.4	77.2	71.2	72.3
16+ years	76.6	64.4	75.2	75.7	88.4	77.9	80.5	80.3

Source: 1980–2000 U.S. Decennial Census and 2012 American Community Survey, Integrated Public Use Microdata Series.

To summarize, increases in divorce preceded declines in marriage, beginning first among the most disadvantaged blacks. Whites and blacks of all classes have experienced delays in marriage, but declines in the proportion who have ever married at age 40–44 also appeared first for blacks with low levels of education. By 1980, we began to see an educational divergence in family patterns for whites. First, the college-educated saw declines in divorce, while those without college maintained high levels of divorce. More recently, whites with the lowest levels of education are beginning to experience delays in marriage relative to college-educated women, and an increasing proportion are likely to never marry.

Explanations for the Black-White Marriage Gap by Education

Black-white differences in marriage appear at all levels of education, suggesting that something more than class status is at play. At the same time, we’ve seen that class status has become increasingly associated with marriage patterns. Among black women, and more recently among white women, lower levels of education have become associated with higher levels of divorce and declines in marriage. This increasing connection between education and the formation of stable families suggests that the structural forces that generate racial differences in marriage and marital stability might vary across different educational groups.⁴²

As we’ve said, classic arguments that link lower marriage rates among black women to a shortage of marriageable men tend to focus on differences in men’s employment prospects and incarceration. Because unemployment and incarceration are highest among black men who are disadvantaged to begin with, we would expect these factors to suppress marriage rates most strongly among poor and working-class black women.

A shortage of marriageable men may be part of the explanation for low marriage rates among better-educated black women, but it’s harder to see how the ratio of men to women can explain low marriage rates among better-educated black men. Some scholars argue that the scarcity of better-off black men relative to black women, which is compounded by black men’s relatively lower levels of education and higher rates of interracial marriage, may increase black men’s bargaining power and make marriage less attractive to them as an option in early adulthood.⁴³ This argument assumes, however, that men would rather have informal relationships with women than marry, despite having access to a larger pool of women eligible for marriage. Because nearly all studies linking the gender ratio to marriage have focused on what predicts marriage among women, we don’t have good evidence on this point. A true test of this argument would analyze men’s marriage.

Another possibility is that both middle-class black men and middle-class black women have more trouble finding spouses because their social worlds consist mostly of people who are not likely to connect them to potential mates. Marriages between black people and people of other races continue to be rare.⁴⁴ More broadly, our social networks tend to be homophilous; that is, they include only people of our own race.⁴⁵ Even friendships that cross racial boundaries tend to be less close and involve fewer shared activities.⁴⁶ Although the social networks that form around work may provide some access to potential mates, this is likely to be less true for blacks who work in mostly white environments.⁴⁷ For example, research shows that black adolescent girls who go to schools where the student body is mostly white are less likely than white girls to be involved in romantic relationships.⁴⁸

Finally, many studies have documented important racial differences in the economic returns to schooling. As young adults, black men have more trouble transitioning into stable full-time employment than white men do, and this racial difference is particularly pronounced among men with lower levels of education. In early adulthood, even college-educated black men earn less than white men, however.⁴⁹ These differences in career entry alone help explain why black men are slower to marry than white men. But a difficult transition to stable employment is an even greater barrier to marriage for black men than it is for white men.

Blacks’ greater sensitivity to labor force transitions might be explained at least partly by the fact that black families accumulate less wealth than white families do. For example, home ownership is less likely to lead to wealth among blacks than among whites, because of high levels of residential segregation and a general reluctance among whites to live near blacks.⁵⁰ Thus young black couples are less likely to have a nest egg to fall back on if they lose their jobs. They are also less likely to be able to rely on their parents for support during rough times. Research shows that differences in wealth can account for some of the racial gap in marriage, especially among men.⁵¹

In sum, differences in employment, earnings, and wealth might account for a sizeable portion of the contemporary racial gap in marriage. Additionally, persistent patterns of racial stratification, such as high rates of residential segregation (which affects the accumulation of wealth, as well as school quality and young men’s risk of incarceration), combine with economic disadvantage to depress black marriage rates today. Yet we still don’t know why black marriage began to fall in the middle of the 20th century and why it continued to do so through good economic times and bad.

Another puzzle is that Hispanic marriage patterns more closely resemble those of whites than those of blacks, despite the fact that Hispanic and black Americans face similar levels of economic disadvantage.⁵² A common explanation is that a large proportion of the Hispanic population in the United States consists of first or second generation immigrants who come from collectivist countries where the imperative to marry remains strong.⁵³ Yet studies that have tried to link race- or ethnic-specific attitudes and beliefs to variation in marriage patterns have generally not found clear supporting evidence. Compared to whites, black women and (especially) men are less likely to say they want to marry, but so are Hispanic women.⁵⁴ Moreover, differences in attitudes about marriage can’t explain lower rates of marriage among blacks.⁵⁵ Even if the attitudes that immigrants bring from other countries buoy Hispanic marriage rates, over time and across generations Hispanic women in the United States experience lower levels of marriage and higher rates of unmarried childbearing. In the third generation and beyond, Hispanic women’s family patterns increasingly resemble those of black Americans. Exposure to economic disadvantage in the United States, then, combined with the widespread individualistic ethos here, eventually trumps whatever pro-marriage disposition Hispanics might have had.⁵⁶

The Growing Importance of Economic Status for Marriage

To understand the dramatic declines in marriage among blacks, we must consider broad changes in the labor force as well as changing ideas about gender and family relationships. These changes made employment and earnings, especially those of women, more important for forming stable families. Changing ideas about family affected both whites and blacks, but they affected black families earlier and more strongly because blacks were and continue to be more economically vulnerable. Since 1980, as economic restructuring has eroded opportunities for less-educated whites, they too are seeing dramatic changes in family life.

Over the past century, families in the United States and most of Europe have undergone sweeping changes across all social and demographic groups. The age at marriage rose, nonmarital cohabitation became common, and divorce rates skyrocketed. Some demographers refer to these broad changes in family life as the Second Demographic Transition. (The original Demographic Transition was the shift from high birth and death rates to low birth and death rates experienced first by Western Europe and eventually by all countries). Because these changes have occurred in both good economic times and bad, and have affected all socioeconomic groups, many believe that changing ideas about the family have helped drive them.⁵⁷

For example, during the 1960s and 1970s divorce and premarital sex both became more widely accepted.⁵⁸ Changes in attitudes toward divorce appear to have followed rises in divorce, suggesting something other than growing acceptance was responsible for the rise in divorce that started around the beginning of the 20th century.⁵⁹ Nonetheless, rising divorce rates combined with growing acceptance of premarital sex might have encouraged people to delay marriage and cohabit outside of marriage.⁶⁰ Altogether, this reinforced the notion that decisions to marry or divorce are a private concern, not something subject to social sanction.

Shifts in the labor force likely also contributed to the Second Demographic Transition’s changes in family life. The service-based economy’s growth since 1950 has enhanced the incentives to get an education for both men and women, but especially for women.⁶¹ Because marriage in early adulthood would interfere with college and starting a career, men and women have been delaying marriage for the past 50 years.⁶² Nonetheless, until recently, most women have continued to marry eventually.

Since 1980, marriage and divorce patterns have become increasingly stratified by class. For example, in the late 1970s, the percentage of marriages that dissolved within 10 years was not that different among women with a college degree (29 percent) than among women with just a high school diploma (35 percent), a difference of only 6 percentage points. For marriages beginning in the early 1990s, this gap had grown to over 20 percentage points.⁶³ As we’ve noted, differences in marriage are also beginning to emerge by social class. Historically, college-educated women were less likely to marry.⁶⁴ But beginning with people born in 1955–64, college-educated women became more likely than other women to ever marry.⁶⁵ Recent projections suggest that the educational gap in marriage will continue to widen over time.⁶⁶ Other evidence has shown that higher-earning women are also increasingly more likely to marry.⁶⁷

Young adults who don’t earn a college degree face diminishing prospects in today’s information economy. Wage disparities by education have grown substantially since 1980, mostly due to the growing demand for college-educated workers.⁶⁸ Compared to their more highly educated counterparts, people without a college degree are less likely to achieve the economic security they feel they need for marriage, and those who do marry are more likely to divorce.

In sum, in the early part of the 20th century, urbanization and other shifts in the economy occurred alongside gradual but modest increases in divorce, especially among blacks. In the years immediately following World War II, unanticipated economic prosperity boosted marriage rates, but only temporarily. Broader cultural trends that emphasized individual choice and gender equality contributed to a growing divorce rate. Divorce among blacks had begun to rise earlier, and the postwar marriage boom didn’t last as long for blacks as it did for whites. By the 1960s, the proportion of blacks who ever married had started to decline. Divorce among whites began rising later, but divorce rates for both whites and blacks accelerated substantially in the 1970s. Starting in 1980, as the gap between the wages of more- and less-educated people started to widen, the educational gradient in divorce began to grow as well. Today, divorce rates are substantially higher for the less-educated than for those with a college degree. Most recently, it looks as if the proportion of less-educated white women who ever marry has begun to fall. Although college-educated women delay marriage, most will eventually get and stay married. This divide between more- and less-educated white women helps us understand black-white differences, because it makes clear that over time, marriage has become increasingly linked to employment and earnings, especially for women. Even though blacks’ economic opportunities have improved in some respects, they still aren’t nearly equivalent to those of whites.⁶⁹ Thus black-white differences in marriage have grown so much since 1960 because economic factors have become increasingly important for marriage formation and stability, and blacks continue to face economic disadvantage.

Inequality and the Continuing Significance of Race

A number of points emerge from our discussion. First, racial differences in U.S. marriage patterns remain large. On average, black women are less likely to marry and to remain married than are white women. Second, although racial gaps in marriage persist across the educational distribution, they tend to be largest among people with the least education. Moreover, for both black and white women, marriage appears to have begun to fall first among those with no more than a high school degree. Third, for both black and white women, marital instability rose before marriage formation fell. Finally, for both groups, educational gradients in marital instability emerged before educational gradients in marriage formation. These patterns have implications for change and variability in families that transcend racial differences in marriage.

No existing explanation alone can fully account for racial gaps in marriage patterns. But we are likely setting the bar too high if we expect any single theory to account for change and variability in processes as complex as marriage formation and dissolution. A broader lesson from studying racial differences in marriage is that if we seek to explain changing family patterns, we need to examine social class. Although no single explanation can account for all the racial gaps we see in marriage, individual theories offer useful (albeit partial) explanations for marriage gaps in specific socioeconomic strata. Most of the recent research on the racial marriage gap focuses on relatively disadvantaged populations and on women. Yet we could learn much about racial variability in marriage, and about family change more broadly, if we looked at marriage patterns among relatively well-off populations and among men.

There may be meaningful linkages between broad trends in marriage formation and marital stability and the differences we see by race. When the imperative to marry was high, as it was through the mid-20th century in the United States, the vast majority of women married despite high levels of poverty. But as an individualistic ethos took hold, the dominant model of marriage shifted from institutional marriage based on gendered roles and economic cooperation to relatively fragile marriages based on companionship, and divorce rates began to climb.⁷⁰ Rising divorce rates, in turn, have further increased the ideal of individual self-sufficiency, encouraging delays in marriage and high levels of marital instability, as demographer Larry Bumpass argued in his 1990 Presidential Address to the Population Association of America.⁷¹ As women and couples became increasingly aware of marriage’s fragility, investments in some marital relationships may have declined, lowering the likelihood that they would last. The growth in divorce may also have led some women and couples to be less willing to marry in the first place. Bumpass argued that no changes have altered family life more than the growth in marital instability.

Finally, people with less education appear to be leading the trends with respect to marriage and marital stability, regardless of race. Again, there may be lessons here for thinking about family change more broadly. Generally, as marital stability and, eventually, marriage formation became more strongly linked to the transition into stable employment for both men and women, blacks’ economic disadvantage became a greater impediment to marriage. The legacy of legal discrimination, as well as continued racial bias in friendship networks, residential preferences, and mate preferences, all contribute to racial inequalities within education groups. Yet whites are not immune to structural forces. Growing inequality has contributed to high rates of divorce among less-educated whites for decades, and, more recently, has started to erode their marriage opportunities as well.