Racial differences in relationship and childbearing patterns

Last Updated on January 20, 2024

The post documents racial differences in relationship and childbearing patterns. This covers a broad range of topics including racial differences in marriage rates, cohabitation rates, marital stability, marital well-being, sexual behaviors, pregnancies, abortions, fertility rates, single mother household rates, half-siblings, and more. Because of the broad range of topics covered, it is not feasible to include a somewhat comprehensive summary here. Instead, I will place summaries at the beginning of each section describing the main important findings. In general, there is a typical rank-order of desirable outcomes by race/ethnicity. Asians tend to have the best outcomes, followed by whites, Hispanics, and then blacks with the worse outcomes.

This post is primarily focused on compiling descriptive statistics. Not much effort will go into explaining these racial disparities. That will have to wait for another post. I also included a description of all of the major datasets referenced at the bottom of the post. I tried to only include studies that used nationally representative datasets, although there are a few exceptions to this rule.

Union rates


This section describes racial differences in the rates of union formation (unions include both marriages and cohabiting relationships). This section focuses on racial differences in marriages, cohabitation, union transitions, and post-conception marriages.

Summary of section

  • Black people are much less likely to marry than other racial groups. Asian people are more likely to marry than other racial groups. These racial gaps have only emerged in recent decades.
  • Black people are less likely to form unions of any type (marital or cohabiting union) than white people. Hispanics form unions at rates between the rates for blacks and whites. Asians have low rates of cohabitation but high rates of marriage.
  • Black cohabiting unions are less likely to transition to marriage than white cohabiting unions. Hispanic cohabiting unions transition to marriage at rates between the rates for blacks and whites.
  • After a non-marital conception, pregnant black mothers are much less likely to marry before the child’s birth than pregnant white mothers. Pregnant Hispanic mothers typically married before the child’s birth at rates between the rates for blacks and whites.

Much of the information presented in this section has also been described by Smock and Schwartz (2020) [archived] in their review on demographic trends and research on families in the United States.

Marriage

There are significant racial disparities in rates of marriage, although these disparities have only emerged in recent decades. In their review of marriage trends by race/ethnicity, Smock and Schwartz (2020) [archived] noted increasingly divergent rates of marriage by race/ethnicity:

Population-level marriage trends obscure substantial differences by race and ethnicity and by education. One important research emphasis over the past few decades has been a growing Black–White gap in marriage in which African Americans have lower marriage rates than Whites (). In 1970, among 40- to 44-year-old women, 95% of White women had married as had 92% of African American women. In contrast, data from 2012 indicate that fewer than two thirds of 40- to 44-year-old African American women had ever married compared with 88% of White women (). Generally, race–ethnic differences in the percentage of women who have ever married have grown. For example, Hispanic, American Indian/Native Alaskan, and African American women all have increasingly different percentages of women marrying when compared with White and Asian/Pacific Islander women and compared with one another ().

Raley et al. (2015) [archived] reported on trends in racial disparities in marriage using U.S. decennial census data and the 2012 American Community Survey (ACS). The following chart shows the percentage of women aged 40−44 years who have ever been married by year:

In recent decades, blacks have much lower rates of marriage than whites, who have slightly lower marriage rates than Asians and slightly higher marriage rates than Hispanics. However, many of these patterns were not present in the first half of the 20th century, when blacks often had higher rates of marriage than whites.

When looking at the percentage of women who are currently married with their spouse present (rather than the percentage of women who have ever been married) by year, the picture is slightly different. Now, the racial disparities have been fairly consistent since 1930:

The above two charts imply that, even though black women had similar or higher rates of marriage than white women in the first half of the 20th century, many black women had high rates of divorce and/or separation which has decreased over time (as will be shown below). These findings are consistent with the data reported in my previous post examining marital rates by race/ethnicity using Census data going back to 1890 (one difference is that my previous post plots marital rates of women of all ages, whereas Raley et al. 2015 focus on older women).

Large racial disparities also persist when breaking out the data by age:

In 2008−12, at each age group past 20, whites are several times more likely than blacks to have had a first marriage. Asians have rates of marriage that are fairly similar to whites, though there is some gap depending on the age. Hispanics seem to have rates of marriage at the midpoint between blacks and whites. There are also racial disparities in rates of divorce, though I will focus on divorce in a later section.

Racial disparities in marriage rates are so large that there are nearly as many Asian married family groups as there are black married family groups (about 4,944,000 vs 5,236,000 married opposite-sex couple family groups as of 2023) in the United States (Census 2023 [archived]). This is despite the fact that the black population is over twice the size as the Asian population (Census 2023).

There are also racial differences in rates of remarriage after ending a previous marriage. McNamee and Raley (2011) [archived] used the 2006-2009 NSFG and 2004 SIPP to examine the time for women to remarry after separating from or divorcing their first husband. They find that time-to-marry is greatest for black women, followed by Hispanic women, and white women with the fastest time to remarry:

By 3.8 years after separation, 25% of white women have remarried. U.S. born and foreign born Latina women have slower entry into remarriage, taking 5.1 and 5.2 years until 25% have remarried respectively. Black women had the slowest pace of remarriage and never reach 25%, as only 19% had remarried after 6 years of separation (not shown).

There were similar racial disparities in time to repartner (i.e. form a cohabiting union) after ending a marriage.

Cohabitation

The previous data shows racial disparities in rates of marriage specifically. There are also racial disparities in rates of unions more generally, which includes both marriages and cohabiting unions, although such disparities are not as great as the marriage disparities. Raley (1996) [archived] reported these findings using data from the 1987−1988 National Survey of Families in Households (NSFH). The sample was limited to just over 3,000 women aged 19−34. The following chart on the left shows the percent of women by race and age who have never been in a union and never been married:

The graph was described as follows:

Figure 1 shows the results of life-table estimates of women’s cumulative survival to first marriage and first union. A striking feature of these graphs is that the Black-White difference in age at first union is much smaller than for first marriage. At age 25, one-third of the White women are never-married, but well over half of Black women are never-married-and the racial difference is over 25 points. However, at this same age, more than three-quarters of White women and over 60 percent of Black women have been in a union – here, the racial difference is only 15 percentage points.

Bramlett et al. (2002) [archived] reported data on marriage, cohabitation, divorce, and remarriage from the 1995 National Survey of Family Growth (NSFG). Consistent with the prior findings on marriage, the report finds that black women are less likely than whites to have ever cohabited or married:

Almost 28 percent of women 15–44 years of age have never married nor cohabited (table B). This percentage is considerably larger for young women and decreases as age increases. About 62 percent of women have ever been married, one-half of whom have ever cohabited and one-half of whom have never cohabited. The remaining 10 percent have cohabited, but never married. Non-Hispanic white women are more likely to have experienced both cohabitation and marriage, while non-Hispanic black women are more likely to have experienced neither cohabitation nor marriage (table B).

Table B of the report shows the exact rates of marriage and cohabitation. The table shows that about 40% of black women aged 15−44 have never cohabited or married, compared to just 25% of white and 28% of Hispanic women. I illustrated the data in the following chart:

Copen et al. (2012) [archived] reported similar findings using the 2006−2010 NSFG. The marital status of women by race/ethnicity were reported in Table 1. The following chart illustrates the data shown in Table 1:

Consistent with prior data, black women are much less likely to form a union than white and Asian women. The patterns for Hispanic women differ sharply by nativity. Foreign born Hispanic women have similar patterns as white and Asian women, whereas U.S. born Hispanic women have patterns that are more similar to black women. It would be interesting to also break out the data for Asian women by nativity, but that information was not provided in the study (it’s possible that sample sizes weren’t large enough to break out the data).

The patterns for men are mostly similar to the patterns for women (Table 2). One key difference is that black men have higher rates of marriage than black women, whereas men have lower rates of marriage than women for each other racial/ethnic group. In fact, compared to Hispanic men, black men have slightly higher rates of marriage and rates of being in a union.

The previous data show that a very small percentage of Asians are cohabiting. This has also been discovered in other studies of NSFG data. For example, Copen et al. (2013) [archived] used data from the 2006−2010 NSFG to report the probability of having a first premarital cohabitation by age among women. They found that roughly 70−80% of women from all ethnic groups except Asians had a first premarital cohabitation by age 30, whereas only 38% of Asian women had done the same (Table 2). In fact, most women of all ethnic groups except for Asians had a first premarital cohabitation by age 25, whereas 19% of Asian women had done the same. I plotted this data in the following chart:

Despite racial differences in rates of cohabitation and rates of marriage, the currently available data suggests very similar rates of cohabitation prior to marriage. Smock and Schwartz (2020) [archived] report the following:

Racial-ethnic differences in the percentage of adults who cohabit prior to marriage are small: The numbers are 70%, 77%, and 73% for non-Hispanic Whites, non-Hispanic Blacks, and Hispanics, respectively. The major point is that most people who marry are cohabiting first.

However, it would be interesting to see the data for Asians. The previous data suggests that Asian women are very likely to marry, but relatively unlikely to cohabit. This would imply that Asians would have lower rates of cohabitation prior to marriage.

Transitions

Conditional on cohabitation, there are racial disparities in rates of transition to marriage. These findings were reported by Bramlett et al. (2002) [archived] (which was described earlier) using the 1995 NSFG:

Table 9 and figure 6 show that the probability that a first premarital cohabitation becomes a marriage is higher for white women, lower for black women, and intermediate for Hispanic women (figure 6). The probability that the first cohabitation becomes a marriage within 5 years is 75 percent for white women, 61 percent for Hispanic women, and only 48 percent for black women (table 9).

Interestingly, family income seems to be associated with rates of transition to marriage for black women but not white women:

Among white women, there is only a nonsignificant 4 percentage point difference in the probability of the transition to marriage between the low-income and high-income groups, whereas the difference is 32 percentage points among black women (figure 7).

Copen et al. (2013) [archived] used the 2006−2010 NSFG to report racial differences in whether first premarital cohabiting unions transition into marriage within 3 years. After 3 years, about 31% of black cohabiting unions transitioned into marriage, compared to 44% of white cohabiting unions, 31% of U.S. born Hispanic cohabiting unions, and 42% of foreign born Hispanic cohabiting unions. Thus, cohabiting unions of white and foreign born Hispanics transition to marriage at considerably higher rates than cohabiting unions of black and U.S. born Hispanics.

Post-conception

There are also racial differences in the rates of marriage after a non-marital conception and before the birth of the child (sometimes called “shotgun marriages”). Compared to unmarried black women, unmarried white women who become pregnant are more likely to marry before their child’s birth. These findings were reported by Manning (2001) [archived] which analyzed racial differences in conception and childbearing among cohabiting couples using the 1995 NSFG. In this study, 44% of white mothers married before birth whereas only 13% of black mothers did the same (Table 2). The rate for Hispanic mothers was between the rate for black and white mothers (27%).

These racial disparities were also emphasized by the author:

In this article, we can distinguish between women who had children that were born during cohabitation and those who conceived children during cohabitation. This distinction is important because nearly three-fifths (59%) of women who conceived a child during cohabitation went on to give birth to that child during cohabitation (Table 2). Thirty-four percent of the women married before the birth of their child, while only 7% of women separated from their cohabiting partner before their child was born…the family type at birth sharply differs among racial and ethnic groups. The overwhelming majority of Hispanic (69%) and black (77%) women who conceived a child during cohabitation also gave birth during that cohabiting union, while only half of white women did so.

Note that the racial differences reported here actually underestimate the true racial disparities in shotgun marriages. That’s because the sample focused specifically on conceptions that occurred during cohabitation. But unmarried black mothers are probably less likely to conceive children during cohabitation compared to unmarried white and Hispanic mothers. Unmarried black mothers are more likely to be single (i.e. not cohabiting) than unmarried white and Hispanic mothers, as I show below.

Similar findings have been reported in more recent iterations of the NSFG. Lichter et al. (2014) used the 2006−2010 NSFG to study transitions to cohabitation and marriage among women who were unmarried at conception. The authors reported large racial differences in transitions after unmarried conceptions, particularly among women who were single (neither cohabiting nor married) at conception (Table 4).

  • Among women who were cohabiting at conception, about 18% of white women transitioned to marriage by the child’s birth, compared to 11% of black women and 7% of Hispanic women. Moreover, among these women, about 6% of white women transitioned to single at birth, compared to 12% of black women, and 5% of Hispanic women.
  • Among women who were single at conception, about 32% of white women transitioned to either marriage or cohabitation by the child’s birth, compared to 13% of black women, and 27% of Hispanic women. Thus, among women who were single at conception, white and Hispanic women were about 2-3 times as likely as black women to transition into a union before the child’s birth.

Similar findings were reported by Lichter et al. (2012) using the 2006−2008 data from the NSFG.

Union quality


This section describes racial differences in union quality. This involves racial differences in relationship stability, survival, infidelity, and marital well-being.

Summary of section

  • Unions with black women are less stable than unions with white women. Hispanic women actually tend to have similar or more stable unions than white women. Asian women have more stable unions than other racial groups. Unions with black women are about 2-3 times as likely as unions with Asian women to suffer some form of disruption.
  • Black people are much more likely to engage in sexual infidelity and extramarital sex than white people. Data on Hispanic and Asian individuals was either inconsistent or absent.
  • Black women report much lower levels of marital quality than white women. Hispanic women typically reported levels of marital quality between the levels for black and white women. Interestingly, some data indicate that black men report similar lower levels of marital quality as white men, although this is not consistent.

Union stability

There are large racial differences in marriage stability and survival. However, some data suggest that these differences only exist for women. In their review on divorce by race/ethnicity, Smock and Schwartz (2020) [archived] note that differences in marital dissolution between black and white women emerged in the 20th century and have increased over time:

Differences in marital dissolution between Black and White women, however, have increased. Before 1920, race differences in divorce were relatively small, but they increased substantially between the 1960s and 1990s (). Estimates from the late 2000s of the lifetime percentage of marriages that dissolve indicate that Black women’s marriages are more likely to dissolve than other women’s; 63% of Black women’s first marriages are estimated to dissolve within 20 years compared with 46% of White women’s, 47% of Hispanic women’s, and 31% of Asian women’s ().

Copen et al. (2012) [archived] reported racial differences in the probability that a first marriage remains intact. A marriage is defined as remaining “intact” if there is no disruption, where “disruption” is defined as either separation or divorce (widowhood is not considered as disruption since it is very rare in the sample). The report found that marriages for black women had the highest rate of disruption, Asian women had the lowest rate of disruption, and white and Hispanic women were about in the middle. The following figure shows the percentage of marriages remaining intact by year and race among women:

The biggest difference is between black and Asian women. After 20 years, Asian marriages are about twice as likely as black marriages to remain intact among women (69% vs 37%).

Surprisingly, there were no large racial differences in marriage survival among men.

Looking at Hispanic origin and race, Asian women (69%) and foreign born Hispanic men (70%) had the highest probability of first marriages lasting 20 years (Figure 4; Tables 5 and 6). Black women (37%) had the lowest chances of first marriages lasting 20 years, significantly lower than for white women (54%). There was no significant difference in the probability of first marriage lasting 20 years between white (54%) and black (53%) men. Foreign born Hispanic women and men (56% and 70%, respectively) had higher of probabilities of their marriage lasting 20 years compared with U.S.-born Hispanic women and men (47% and 48%, respectively).

So, while black women are much more likely to experience a marital disruption than white women, black men are no more likely than white men to experience such an event after 20 years. In fact, Hispanic men are more likely to experience a marital disruption than black men after 20 years (48% vs 53% probability of marriage remaining intact for 20 years). Unfortunately, data was not presented for Asian men.

Racial differences in marital stability were also reported in Raley et al. (2015) [archived]. Using the 2006−2010 NSFG dataset, these authors found that, among women aged 40−44, 53% of ever-married black women have experienced an unstable marriage compared to 41% of white women and 46% of Hispanic women (Table 4). When accounting for the fact that black women are much less likely to marry in the first place, this implies considerably large racial differences in rates of stable marriages, at least among women:

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.)

Now, let us focus specifically on divorce (rather than marital disruption generally which may include separation). Data on divorce by race is presented in a Census report conducted by Mayol-García et al. (2021) [archived]. The authors rely on a supplement of the 2014 Survey of Income and Program Participation (SIPP). Using this data, the authors illustrated the percentage of women whose first marriage ended in a divorce as follows:

Here, we see mostly typical racial patterns: Black women are most likely to divorce, Asian women are least likely to divorce, with whites and Hispanics somewhere in the middle. The one surprise here is that Hispanic women are less likely to divorce than non-Hispanic white women.

Racial differences in stability are found not just among marriages but also for cohabitations. Lamidi et al. (2019) [archived] found that first premarital cohabiting unions are less stable among black couples compared to white and Hispanic couples. Based on 2006−2013 data from the NSFG, the authors report that roughly 47% of cohabitations dissolve within 5 years for blacks, compared to 35% for whites, and just 25% for Hispanics (Table 2).

Sexual infidelity

There are significant differences in rates of infidelity by race and ethnicity. Before considering specific studies, I will first reference a systematic review on infidelity conducted by Haseli et al. (2019) [archived]. In the section on race and ethnicity, the authors note the following:

Although numerous studies have reported higher indelity behavior among blacks and African Americans, other studies do not support this conclusion. Higher rates of incarceration and other social contextual factors in these communities can lead to the unavailability of desired primary partners.

They cite 6 studies finding higher rates of infidelity for blacks but 2 “other studies” without this finding. The 2 “other studies” are Kuroki (2013) and Atkins and Kessel (2008). Both of these studies rely on data from the General Social Survey (GSS) to estimate frequency of marital infidelity. When considering the data on infidelity reported since 1991, the overall trend shows that blacks have higher rates of infidelity than whites (as shown below). Before reporting this evidence, let us consider the 2 “other studies” here that allegedly do not find higher rates of infidelity among blacks:

  • Kuroki (2013) analyzed the association between workplace sex ratios and marital infidelity in the 1998 GSS. The author regressed incidence of marital infidelity on a number of independent variables (including race). The regression models found that race (specifically, being black) was not significantly associated with marital infidelity after controlling for other variables such as marital status (e.g., whether one is married, divorced, separated, etc.) educational attainment, employment status, etc. However, this doesn’t show that blacks don’t have higher rates of infidelity simpliciter; it merely shows that blacks don’t have higher rates of infidelity after controlling for various factors.
  • Atkins and Kessel (2008) also used the 1998 GSS in order to analyze the relationship between religiosity and infidelity. Like the previous study, this study also regressed infidelity on a number of independent variables, such as race, education, income, religious attendance, etc. Each model found that black individuals had higher rates of infidelity (odds ratio for infidelity for non-black individuals was around 0.7 to 0.75, depending on the model, Table 3). However, these effects were statistically insignificant in one model (Model 1) and only significant at the 10% level in two other models (Model 2 and Model 3). The authors report the following: “There is a weak suggestion of differences in rates of infidelity between non-Black and Black participants, with fewer non-Black participants reporting infidelity.” So, contrary to the description by Haseli et al. (2019), this study does find higher rates of infidelity among black participants (although this was significant only at the 5% in some models). In addition, just like Kuroki (2013), this investigated whether blacks have higher rates of infidelity after controlling for various factors, rather than whether blacks have higher rates of infidelity simpliciter.

Now it is worth checking whether there are racial differences in infidelity in the raw GSS data, without controlling for extra factors. Also, it is worth reviewing GSS data across all waves, instead of just the data from a particular year. This has been investigated specifically by Djamba and Kimuna (2020) [archived]. These authors used 15 waves of GSS data to investigate racial and gender differences in extramarital sex in the United States between 1991 and 2018. These waves of the GSS asked ever married participants the following question: “Have you ever had sex with someone other than your husband or wife while you were married?” The findings show consistently higher rates of infidelity for black respondents compared to white respondents:

Interestingly, men are also considerably more likely to report extramarital sex than women for each year:

Considering all waves of data together, about 24% of blacks reported extramarital sex compared to just 17% of white respondents (see Table 1). So blacks reported about 40% higher rates of extramarital sex than whites. The results were described as follows:

Racial differences between blacks and whites have also been consistent in the last three decades (Figure 2). In 1991, 13.46% of white participants and 25.49% of black participants reported engaging in EMS. In 2018, the rate for blacks who reported engaging in EMS had decreased by one percent, whereas, the rate for whites increased by a little more than two percent (Table 1). Further, our findings show that individuals who were divorced and those who were separated reported higher rates of EMS than both currently married and widowed participants (Table 1).

These racial differences have also been reported in other nationally representative surveys. One additional study reporting these findings comes from Treas and Giesen (2000) [archived]. The researchers focused on sexual infidelity among both married and cohabiting Americans using the 1992 National Health and Social Life Survey (NHSLS). The researchers found that blacks reported much higher rates of extramarital sex than whites:

Gender and race were statistically significant. All things considered, being male increased the odds of having engaged in extramarital sex by 79%. Being African-American raised them by 106%, even though education controlled for racial differences in socioeconomic status. showed a weak negative association. Both frankness and marital duration (i.e., exposure time) showed the expected positive relationships.

The authors concludes with the following summary:

Although epidemiological research consistently reports men to be at higher risk of infidelity than women, studies have not usually included indicators of sexual values and tastes. When we controlled for interest in sex and permissiveness of sexual values, we found that the main effects of gender were markedly reduced or even eliminated. Consistent with prior research, we found that being African-American was positively associated with multiple sex partners, even when educational attainment (an indicator of socioeconomic status) and other variables were controlled. The persistence of this effect points to the need for further research to clarify the role of race. Because we found the sexual opportunity structure to be important in understanding sexual behavior, racial differences in the sex ratio might influence the likelihood of having multiple partners.

Whisman and Snyder (2007) also studied infidelity in the 1995 NSFG, focusing on the nearly 5,000 women who were married for at least 13 months. Sexual infidelity was operationalized by asking these women how many men they had sexual intercourse with. Participants who reported 2 or more sex partners were coded as having engaged in sexual infidelity.

Note: this measure will have some error because of the following two reasons: (1) this measure overcounts infidelity insofar as it counts extramarital sex acts that are permitted by one’s spouse, and (2) this measure undercounts infidelity insofar as it excludes extramarital sex acts with same-sex partners and extramarital sex acts by women who did not have sex with their husband.

Nevertheless, the results showed that the odds of sexual infidelity reported by black women were about 2.5 to 3.35 times the odds reported by non-black women, even after controlling for a few confounding factors (Table 2). The actual racial difference depends on whether the data was collected via face-to-face interviews or the computer assisted interviews (A-CASI).

The following is included in the summary of the findings:

As can be seen in this table, for both methods of interview, the probability of sexual infidelity was (a) significantly and positively associated with race (i.e., being Black), lifetime sexual partners, childhood sexual abuse, and premarital cohabitation; and (b) was significantly and negatively associated with religiosity.

Marital well-being

Several studies also show large racial differences in marital well-being. For example, Broman (1993) investigated this topic in the aptly titled study “Race differences in Marital Well-Being”. The author analyzed the relationship between race and marital well-being among 2,059 married adults in the first wave of the ACL which was conducted in 1986. Marital well-being was assessed using two measures: marital harmony and marital satisfaction. Marital harmony was measured based on the degree to which respondents agreed with the following statements:

  • “There is a great deal of love and affection expressed in our marriage”.
  • “My spouse doesn’t treat me as well as I deserve to be treated”.
  • “I sometimes think of divorcing or separating from my spouse”.
  • “There have been things that have happened in our marriage that I can never forgive”.

Lastly, marital satisfaction was measured based on their answer to the question “Taking all things together, how satisfied are you with your marriage?”. Blacks reported lower marital quality based on both measures:

Table 1 presents a regression of marital well-being measures on predictor variables. The results bear out our hypothesis; blacks are significantly less likely to feel that their marriages are harmonious, and are significantly less likely to be satisfied with their marriages. Age, sex, and parental status are also significant predictors.

Interestingly, for marital satisfaction specifically, the author found significant race by sex interactions. That is, the black-white gap in marital satisfaction is purely due to the fact that black women report lower rates of marital satisfaction than white women. Black men did not report lower rates of marital satisfaction than white men.

Several writers have shown how the sex of the respondent is a major factor in feelings about marriage, and I investigated this possibility by estimating race by sex interaction terms and testing their significance. Statistically significant interactions were found for both measures. Table 3 presents results of a multiple regression analysis of marital satisfaction for men and women (because results are not significant for marital harmony, they are not shown in this table). There is a significant race by sex interaction, with the race difference in marital satisfaction present only for women. It is black women (in comparison to white women) who are less satisfied with their marriages.

Adelmann et al. (1996) also found similar racial disparities in marital quality using the same dataset.

Broman (2005) continued investigating racial differences in marital quality using additional waves from the ACL survey. This author included data from waves 2 and 3 of the ACL, which were gathered in 1989 and 1995, respectively. Marital quality was measured as a composite of the 5 questions described in Broman (1993) (the 4 questions that constituted marital harmony and the 1 question concerning marital satisfaction). In addition, this study also included 6 questions about spousal behavior which are described as follows:

Six items are used to measure spouse behavior. The first three ask whether the spouse is too critical, is willing to listen, and makes one feel loved. These three measures were coded from 1 to 5, where 1 = strongly disagree and 5 = strongly agree. The next three items ask whether the spouse pushes, slaps, or hits the respondent, has extramarital affairs, and wastes money. These last three items are coded from 1 to 3, where 1 = never and 3 = often.

The results showed that blacks were significantly more likely to report many negative spousal behaviors and lower marital quality:

Blacks are more likely to report the negative spousal behaviors of the spouse having affairs, hitting, and wasting money. Blacks are less likely than are Whites to say that their spouse makes them feel loved, and they report lower marital quality than do Whites. All race differences are significant at the .05 level except for sex, spouse critical, and spouse listening.

Bulanda and Brown (2007) [archived] examined racial and ethnic differences in marital quality using a different dataset, the National Survey of Families and Households (NSFH). The authors use the 1987−88 and 1992−94 waves of the NSFH. The authors limited the sample to white, black, and Mexican Americans in intra-racial marriages, resulting in a final sample size of 6,231 respondents. Marital quality was measured with 5 different dimensions: marital happiness, marital interaction, marital disagreements, marital problems, and perceived marital instability. The authors found that blacks reported lower marital quality than whites along each dimension. Unlike other findings reported here, Mexican Americans were not significantly different from whites along most dimensions.

Weighted means and unweighted standard deviations of all variables are shown in Table 1. Mexican Americans report significantly less marital interaction and fewer marital problems than Whites, on average, whereas there are no differences between these two groups on the other three dimensions of marital quality or in the percentages who divorce. Blacks have significantly lower levels of marital happiness and marital interaction, on average, and higher levels of marital disagreements, problems, and perceived instability than Whites. Relative to Blacks, Mexican Americans report higher levels of marital happiness, fewer marital problems, and less perceived instability, on average.

Some studies have also reported racial differences in marital quality over time. For example, Corra et al. (2009) [archived] use General Social Survey (GSS) data to outline trends in marital happiness by race from 1973 to 2006. Marital happiness is measured based on the GSS questionnaire item “Taken things all together, how would you describe your marriage” with answers ranging from “very happy” to “not too happy”. Consistent with prior studies, these authors found significant racial differences in marital happiness for every year of data:

The general finding is that men report higher levels of happiness than women, and whites report higher levels of happiness than blacks. However, in recent years, black women have strangely reported equal or slightly higher levels of happiness compared to black men. It’s unclear if this is just an outlier or the beginning of a trend.

Another study analyzing trends racial differences in marital quality comes from James (2014) [archived]. This author analyzed whether racial differences in marital quality among women persist over the duration of a marriage or relationship. The data was gathered from the 1979 National Longitudinal Survey of Youth (NLSY79). The sample used in this study was limited to women who were in first marriages starting in 1992 (the NLSY79 did not ask about marriage prior to 1992). Thus, the final sample was limited to about 2,640 women.

Marital quality was measured based on marital happiness, marital communication (how often respondents laughed together, calmly discussed, or talked about their day with their spouse), and marital conflict (how often respondents reported arguing about various matters). The results demonstrated that marital happiness and marital communication both declined rapidly over the duration of one’s marriage (Figure 1). When comparing the results by race, black women report lower marital quality than white women along each dimension at the beginning of their marriage, with Hispanic women somewhere in between. Over time, marital quality declines for each racial group. However, for marital happiness in particular, the decline is particularly strong for white women, which results in white women reporting similar levels of happiness as black women and lower levels than Hispanic women. The trajectory by race are presented in Figure 4:

Sexual behaviors


This section describes racial differences in sexual behaviors. This involves racial differences in rates of teenage sexual behavior, age of sexual debut, and number of sexual partners.

Summary of section

  • The percent of teenagers who have ever engaged in sexual intercourse is greatest for black teenagers, followed by Hispanic teenagers, white teenagers, and Asian teenagers (in that order). However, many of these racial disparities have declined significantly in recent years.
  • The age of sexual debut is youngest for blacks, followed by Hispanics, whites, and Asians (in that order). Again, these disparities have diminished sharply in recent years. The racial disparities are particularly large among males.
  • The data on the number of sexual partners is a bit mixed. Black people typically report more sexual partners than white people, particularly among males. Asians report the fewest sex partners. The data for Hispanics seems inconsistent and depends on the dataset.
  • White women are more likely than black and Hispanic women to use pills and other hormonal birth control. Blacks are more likely to use condoms than whites.
  • Racial disparities in contraceptive methods are particularly great for unmarried women, where unmarried black and Hispanic women are much less likely to use effective contraceptive methods than unmarried white women.

Teenage sexual behavior

I reported racial differences in teenage sexual behavior in my previous post which investigated racial differences in rates of risky sexual behaviors among high school students. The data shows large racial differences in the percent of students who reported ever having sexual intercourse, although the gaps have diminished sharply in recent years.

There are also very large racial differences in the percent of students reporting sexual intercourse before age 13. In particular, black students reported rates of sexual intercourse before age 13 that was several times the rate of other racial groups in the 1990s. These differences have nearly closed in recent years.

For more racial differences in sexual behaviors, see my previous post.

Age of sexual debut

Many studies have reported racial disparities in age of sexual debut. The typical finding is that blacks, on average, begin having sex earlier than do whites, who begin having sex before Asians. However, these patterns sometimes depend on the sex of the individuals. In one review of 10 years of longitudinal studies on adolescent sexual behavior in the United States, Zimmer-Gembeck and Helfand (2008) [archived] reported that black males, but not black females, initiated sexual intercourse at earlier ages than their non-black counterparts:

Compared to white adolescents, findings showed (1) earlier onset of sexual intercourse for Black males, but not females, (2) later onset for Asian American adolescents and (3) average age of first intercourse for Hispanic adolescents that was similar to white adolescents. Overall, 11 of 15 studies found associations between age of first sexual intercourse and race/ethnicity, with larger studies more likely to support racial/ethnic differences in age of first intercourse. A Black adolescent male was, on average, 2.8 times more likely to have a history of sexual intercourse when compared to white adolescents. When boys and girls were examined separately, Black girls were not found to initiate sexual intercourse significantly earlier than girls from other racial/ethnic backgrounds.

However, one problem with this review is that it includes samples that are not nationally representative. For example, one of the studies focused on adolescents in Los Angeles. This problem is also mentioned by Cavazos-Rehg et al. (2009) [archived], which note that much of the existing research on this topic relies on non-representative datasets:

Examining differences in sexual debut across racial/ethnic groups is important to improve understanding on the effects of culture on sexuality. A vast amount of published data has already examined sexual debut and the data suggests that African Americans tend to initiate sexual intercourse at an earlier age than Caucasian, Hispanic and Asian youth, while Asians experience sexual debut at a later age when compared with these racial/ethnic groups [5–6]. However, most of the findings are based on community and/or convenience samples with relatively few studies utilizing nationally representative samples that support generalization [7]. In fact, no studies with sample sizes large enough to sufficiently characterize and compare sexual debut across these four major racial/ethnic groups in the general population were found in the existing literature within the last 10 years. This is particularly true for Asians due to the small sample of this population in national data systems.

In this same study, the authors investigated racial differences in age of sexual debut between 1999 and 2007 using the nationally representative Youth Risk Behavior Surveillance System (YRBSS). In this nationally representative sample, the authors found large racial differences in age of sexual debut, among both males and females. The results were presented as follows:

The findings show that roughly half of black males (42%) had engaged in sexual intercourse before their 14th birthday. By comparison, only 11% of white and 7% of Asian males had done the same.

Blum et al. (2000) [archived] also studied racial differences in sexual intercourse (and other risky behaviors) in the National Longitudinal Study of Adolescent Health. Consistent with the previous study, there were large racial differences in the percent of students reporting a history of sexual intercourse. Among students in the 7th to 8th grade, about 37% of black respondents reported having sexual intercourse compared to just 16% and 11% for Hispanic and white respondents, respectively (Table1). Among students in the 9th to 12th grade, about 67% of black respondents reported having sexual intercourse compared to about 46% for both whites and Hispanics.

It should be noted that the findings from the previous two studies are based on data from before 2010. My previous post suggests that most of the racial gap in age of sexual debut has mostly disappeared within the last few years. Data from more recent iterations of the NSFG has also found smaller racial differences in age of sexual debut, particularly among females. For example, Abma and Martinez (2023) [archived] uses data from the 2015−2019 NSFG to report racial differences in the probability of first sex by age. The probability of having had first sex by age is plotted in the following charts for females:

The same chart for males is as follows:

Thus, black respondents reported modestly higher rates of having sex at each than white respondents. For Hispanics, the gaps depended on sex. Hispanic males reported similar rates of sex as black males, whereas Hispanic females reported similar rates as white females.

Number of sex partners

I reported racial differences in number of sex partners among teenagers in my previous post which investigated racial differences in rates of risky sexual behaviors among high school students in the YRBSS. The data shows large racial differences in the percent of students who reported 4+ sex partners, although the gaps have diminished sharply in recent years.

Other analyses of the YRBSS data has also found that black adolescents (specifically black males) reported a larger number of sex partners than white adolescents.

  • Valois et al. (1999) analyzed the 1993 sample from this dataset and found that 56% of black males had 4+ sex partners, compared to 26% of black females, 25% of white males, and 15% of white females (Table 1).
  • Santelli et al. (1998) [archived] also examined sex partners in a larger sample. They analyzed a sample of 8,450 individuals aged 14−22 in the 1992 NHIS and the Youth Risk Behavior Survey (YRBS). The authors found higher number of sex partners for black respondents and Hispanic males: “Black and Hispanic respondents were significantly more likely than their white peers to report six or more lifetime partners (odds ratios, 1.4–2.6), with one exception: Hispanic females had a reduced probability of this outcome” (page 274).

Data from different iterations of the National Survey of Family Growth (NSFG) have also reported more sexual partners for black adults aged 15-44 compared to other racial groups.

  • Leichliter and Aral (2009) [archived] analyzed the 1995 and 2002 NSFG and found that black women were more likely than white and Hispanic women to report having multiple sex partners in the past year. In 1995, black women were about twice as likely to report multiple sex partners as white women (20.4% vs 10.1%). The gap had reduced in 2002, when black women were just about 50% more likely to report multiple sex partners (16.6% vs 11.4%).
  • Harper et al. (2017) [archived] analyzed data from the 2002 and 2011−2013 NSFG and again found racial differences in number of lifetime sexual partners. For example, consider the following findings from the 2011−13 cycle:
    • Hispanic women reported a lower median median number of lifetime opposite-sex partners (2) than white (4) and black women (5) (Table 1).
    • Black men reported a median of 9 sex partners, compared to just 5 for white men, and 4 for Hispanic men (Table 2).
    • Interestingly, there were no racial differences in the median number of opposite-sex partners within the past year (Tables 3-4), with all groups reporting the same median of 1.

Interestingly, the NSFG data suggests fewer sex partners for Hispanics than whites. This is in contrast to the YRBS data which shows greater sex partners. One explanation for this gap may be that the NSFG data includes more foreign-born Hispanics, who may be less likely to have more sex partners.

One inconsistent finding comes from Kan et al. (2010) [archived]. These authors examined predictors of changes in number of sexual partners among adolescents and young adults. They analyzed data from the National Longitudinal Study of Adolescent Health. During each wave, participants were asked to report their sexual relationship history (i.e. relationship involving sexual intercourse) over some period prior to the interview. The authors used these responses to calculate the number of sexual partners per year. Black and white respondents reported very similar numbers of sexual partners; for both groups, the average number of sexual partners increased from 0 at age 13 to around 1 at age 19 (Figures 1-2).

It’s unclear why the results here differ from the other studies. One possibility is that the measure of sexual partners was based on the number of sexual relationships. Perhaps sexual encounters outside of the context of a serious relationship were not considered to be sexual relationships by the respondents. Another strange result of this study is that female respondents reported more sexual partners than male respondents (Figures 1-3), which contradicts all of the previous studies. Perhaps this is explained by the same possibility earlier: maybe male respondents are more likely to have sex outside of the context of a serious relationship and are therefore less likely to include this as a “sexual relationship” when answering the survey.

Contraceptive use

I reported racial differences in contraceptives among teenagers in my previous post which investigated racial differences in rates of risky sexual behaviors among high school students in the YRBSS. The data shows some differences in the types of contraceptives used by race/ethnicity. For example, blacks have been historically more likely than whites and Hispanic to use condoms, although the gaps have closed or reversed in recent years.

Another racial difference is that whites are much more likely than blacks and Hispanics to use birth control pills. These gaps have remained stable at least until recent years.

Other surveys have also reported similar racial differences in types of contraceptives used. For example, consider a 2016 CDC report [archived] on Health in the United States. The report is fairly massive and covers many different topics. However, for my purposes, the focus is on Table 8 which investigates contraceptive use in the past month among women aged 15−44 (to load just this table without the full report, see here [archived]).

The data on contraceptive use is gathered from different iterations of the National Survey of Family Growth (NSFG). The first thing to note from Table 8 is that racial differences in any contraceptive use are not particularly large (see page 2 of Table 8). Among women who use contraception or are sexually active in 2011−2015, 91.2% of white women aged 15−44 use contraception, compared to 86.5% for black women, and 87.1% for Hispanic women. Here are the trends since 1995:

Note: when constructing the line chart, I assigned the values for the 2006−2010 and 2011−2015 datasets to the years 2010 and 2015, respectively, on the x-axis.

For whatever reason, contraceptive use dropped for black women from 90% in 1995 down to 82.8% in 2006−2010, and then back up to 86.5% in 2011−2015.

There are also fairly large racial disparities in the methods of contraceptives used (see page 5 of Table 8). In 2011−2015, black women were more likely than other groups to undergo female sterilization, use condoms, and use an injectable. By contrast, white women are more likely to use birth control pills and are more likely to have a partner who has undergone male sterilization. Here are the rates of contraceptive use by race for the main contraceptives methods in 2011−2015:

Interestingly, some of these racial differences have not always been present. For example, in 1982 black and Hispanic women were more likely to use birth control pills than white women, but now that pattern has reversed:

Furthermore, black and Hispanic women were less likely to use condoms in 1982, but that pattern has reversed as well:

Other studies investigating racial differences in contraceptive use in the NSFG have reported similar findings (e.g. Kramer et al. 2018, Welti et al. 2011, and Daniels et al. 2015). Daniels and Abma (2020) [archived] reported data on racial differences in contraceptives among women based on the 2017−2019 NSFG. This recent data shows that racial differences in female sterilization have reversed in recent years. However, many of the other racial gaps persist for other methods.

Some studies have examined racial differences in contraceptive use using other samples, but these have not used nationally representative samples of fertile women to my knowledge (Kusunoki et al. 2016, Gaydos et al. 2010, Shih et al. 2011). Nevertheless, such studies typically report similar findings here (e.g., black women are more likely to undergo sterilization and use condoms, white women are more likely to use birth control pills, etc.).

Other studies show that racial differences in contraceptives are particularly large among unmarried women. Using the 2006−2010 NSFG, Kim and Raley (2015) [archived] reported the percent of single and cohabiting women who used no contraception, very effective contraception, and other contraception by race/ethnicity. Very effective contraceptive methods include sterilization, IUD, pill, and other hormonal methods. Other methods include male and female condoms, withdrawal, and other methods. The authors found that white women were considerably more likely to use very effective contraceptive methods than black and Hispanic women.

Pregnancies


This section describes racial differences in pregnancies. This involves racial differences in relationship status at conception, rates of unintended pregnancies, and abortions.

Summary of section

  • The percentage of women unmarried at conception is greatest for black women, followed by Hispanic women, white women, and Asian women (in that order). A similar rank-order pattern holds when considering the percent of women who are neither married nor cohabiting at conception. In fact, nearly half of black women are neither married nor cohabiting at the time of conception.
  • Pregnancies of black women are substantially more likely to be unintended (either unwanted or unintended) than pregnancies of white women. The percentage of pregnancies that are unintended for Hispanic women is between the percentage for black and white women. The intended pregnancy rate for black women is actually lower than the rate for white women, whereas the unintended pregnancy rate for black women is 2-3 times the rate for white women.
  • Black women perform more abortions than women of any other racial/ethnic group, accounting for 39% of all abortions despite the fact that black women are outnumbered by both white and Hispanic women. The black abortion rate is about 4 times the white abortion rate and 2 times the Hispanic abortion rate.

Relationship status

Data from nationally representative surveys show that unmarried black women who give birth tend to not be neither married nor cohabiting at the time of conception. This is revealed in Guzman et al. (2010) [archived], a study described earlier which utilized the Early Childhood Longitudinal Study−Birth Cohort. The authors found that, in this sample of mothers who gave birth in 2001, the plurality of black women who gave birth were neither married nor cohabiting at the time of conception (47.7%, Table 1). A much smaller percent of black mothers were cohabiting (27.8%) or married (24.4%).

Lichter et al. (2014) [archived] also found racial differences in relationship status at conception among women in the 2006−2010 NSFG. The authors focused on investigating the most recent live birth occurring after 1998 among the women in the sample. They reported that black women were about 3 times as likely as white women to be neither cohabiting nor married at the time of conception (49% vs 16%). On the other hand, white women were about 3 times as likely as black women to be married at the time of conception:

Intendedness

There are large racial differences in rates of unintended pregnancies. As with other undesirable outcomes, blacks and Hispanics have higher rates of unintended pregnancies than whites and Asians. This has been found several times in different iterations of the National Survey of Family Growth (Finer and Henshaw 2007, Finer and Zolna 2011, Finer and Zolna 2014, Finer and Zolna 2016). Before presenting the data, let us clarify some terms. An unintended pregnancy is a pregnancy that is either mistimed or unwanted, which are defined as follows:

  • Mistimed: a pregnancy that occurred earlier than the mother wanted.
  • Unwanted: a pregnancy that occurred at a time when the mother did not want any (more) children.

That said, in the most recent study from the NSFG, Finer and Zolna (2016) [archived] reported that in 2011 black women had the highest unintended pregnancy rate (79 per 1,000 women aged 15−44), followed by Hispanic women (58) and white women (33) (Table 1). In addition, the percentage of pregnancies that were unintended was greatest for black women (64%), followed by Hispanic women (50%) and white women (38%). This pattern has been mostly present for every year that data has been collected, as illustrated with the following graph:

In each year except for 1981, blacks had the highest rates of unintended pregnancy, followed by Hispanics, and then whites (in that order).

Interestingly, black women actually have lower rates of intended pregnancies than white women. For example, Finer and Zolna (2014) [archived] used the NSFG data to show that the intended pregnancy rate for black women was 40 per 1,000 women in 2008, compared to 51 per 1,000 for white women.

In the most recent study, Finer and Zolna (2016) did not report intended pregnancy rates in 2011. However, these rates can be easily calculated by simply subtracting the unintended pregnancy rate from the total pregnancy rate, which are shown in the following chart. The intended pregnancy rates in 2011 were similar to the rates in 2008.

To my knowledge, there have been no other studies reporting racial differences in unintended pregnancies in nationally representative samples. Other studies typically rely on small samples and/or samples from specific areas of the country (PRAMS, Kemet et al. 2018).

Abortions

There are also large racial differences in rates of abortion. Similar to other patterns, black women are much more likely to have an abortion than white women, with Hispanic women somewhere between the two. The best source for this data comes from the annual “Abortion Surveillance” reports by the CDC. Each year, the CDC requests abortion data from the central health agencies of each state in the country. In the most recent report, “Abortion Surveillance — United States, 2020” by Kortsmit et al. (2022) [archived], over 600,000 abortions were reported to the CDC for 2020. The report had the following findings:

  • Black women had the highest abortion rate at 24.4 per 1,000 women aged 15−44, followed by Hispanic women at 11.4 per 1,000 women, and white women at 6.2 per 1,000 women (Table 6).
  • Black women had the highest abortion ratio at 426 per 1,000 live births, followed by Hispanic women at 173 per 1,000 births, and white women at 118 per 1,000 births.
  • In fact, the black abortion rate was so high that there were more abortions for black women in 2020 than there were for white women (130,538 vs 108,904) despite the black population being much smaller. In other words, despite accounting for about 14% of women, black women accounted for 39.2% of all abortions in 2020, more than women of any other race.

In fact, the high abortion rate for black women seems to partly explain why blacks do not have much higher birth rates than white women. This can be illustrated by research conducted by Curtin and Kost (2015) [archived] which analyzed changes in pregnancy from 1990 to 2010. The authors used data from the same Abortion Surveillance System released by the CDC that was mentioned earlier. The following useful chart shows how pregnancy rates translate to birth rates, abortion rates, and fetal loss rates for each racial group since 1990:

As you can see, the racial gap in pregnancy rates is typically much greater than the racial gap in birth rates. For example, in 2010, the pregnancy rate for black women was 60% higher than the pregnancy rate for white women (135.1 vs 84.1), but the birth rate for black women was only 13% higher (66.6 vs 58.7). The biggest cause of this difference is that such a large portion of black pregnancies are aborted. In fact, in 1990, among pregnancies that did not result in fetal loss, nearly half of the pregnancies for black women were aborted.

The racial disparity in abortion rates is mostly driven by abortions by unmarried women. Unmarried black women have substantially more abortions than unmarried white women. This is shown in a CDC report by Ventura et al. (2001) [archived] which studied trends in pregnancy rates from 1976 to 1997. This abortion data was also based on the same abortion surveillance system described earlier. When breaking down abortion rates by marital status and race, the results were as follows:

There are a few interesting findings to note here:

  • The abortion rate for unmarried black women is extremely high. In fact, it is so high that the abortion rate for unmarried black women is greater than the birth rate for married black women. This pattern is not present for any other racial group.
  • Among unmarried women, the abortion rate is typically very close to the birth rate for each racial group, except for Hispanics where the abortion rate is considerably lower. In other words, conditional on pregnancy for unmarried women, black and white women are similarly likely to abort vs delivery the pregnancy.

This implies that the higher abortion rate for black women compared to white women is primarily driven by the higher pregnancy rates for unmarried women. It is also secondarily driven by the fact that black married women are more likely to abort a pregnancy than white married women.

Similar findings were reported in a more recent report by Ventura et al. (2012) [archived].

There’s also data from different cycles of the NSFG showing racial disparities in the percent of unintended pregnancies ending in abortion (Finer and Zolna 2011, Finer and Zolna 2014, and Finer and Zolna 2016). The data show that the percentage of unintended pregnancies ending in abortion is similar for white and Hispanic women, but somewhat higher for black women. For example, in 2011, Finer and Zolna (2016) [archived] report that 50% of unintended pregnancies for black women ended in abortion, compared to 40% for Hispanic women, and 36% for white women (Table 2).

Fertility rates


This section describes racial differences in rates of reproduction. This concerns racial differences in rates of fertility, parenthood, multi-partner fertility rates, and fertility rates by level of parental skill.

Summary of section

  • Fertility rates in recent years are fairly similar for all racial/ethnic groups. This is mostly due to recent drops in fertility among black and Hispanic women. The one exception to this pattern is Native American women, who have significantly lower fertility rates than other groups.
  • There are similar probabilities of becoming a parent for men and women of each racial/ethnic group. About 80 to 90% of women of all groups become a parent by age 40, compared to about 70 to 80% of men.
  • There are large racial differences in multi-partner fertility (MPF). Black adults are several times as likely as white adults to have children with multiple partners. The MPF rate for Hispanic adults is between the MPF rate for blacks and whites.
  • The negative correlation between parental skill (e.g., IQ, education, etc.) is particularly strong for black and Hispanic women compared to white women.

Raw fertility rates

Birth rates (births per 1,000 population) for all racial groups have sharply decreased since their peak in the mid 20th century. Birth rates by race have been plotted in a previous post.

Expectedly, fertility rates (births per 1,000 women aged 15−44) have also sharply decreased over the same time period. Historically, black and Hispanic women have had considerably higher fertility rates than non-Hispanic women. However, most racial differences have been eliminated in recent years. Fertility rates for Native Americans have experienced particularly strong declines.

The convergence of fertility rates by race/ethnicity was also described by Sweeney and Raley (2014) [archived]:

Fertility rates among white, black, and Hispanic women converged considerably since 1990. The total fertility rate (TFR), a common measure of completed childbearing, corresponds to the average number of children a hypothetical woman would be expected to have during her lifetime, given age-specific fertility rates prevailing in a particular year. TFRs for 1990 suggested that white women would be expected to have a total of 1.9 births in her lifetime, compared to 2.5 births among black women and 3.0 births among Hispanic women (see Table 1). Between 1990 and 2012, however, the TFR declined more rapidly for black and Hispanic women than for white women. Much of the decline for black and Hispanic women reflected reductions in childbearing among young women and teens. Fertility rates also fell during this period among white teens and young adults, but these declines were not as pronounced as for black or Hispanic women, and were also offset by white women’s greater increases in childbearing at older ages (see Figure 2).

The same authors also point out that the higher fertility rate for Hispanic women is primarily driven by foreign born Hispanics:

 Within the population of Hispanic women, fertility rates tend to vary by country of origin and nativity status. For example, contemporary fertility rates are lower for Cuban and Puerto Rican women than for Mexican or other Hispanic women. U.S.-born Hispanic women currently bear children at about the same rate as other women born in the United States (). The difference between Hispanic and Non-Hispanic women in completed fertility reflects the relatively high fertility of immigrant Latinas (), although fertility rates among foreign-born women living in the United States began to fall dramatically in 2007 ().

Another discussion on racial differences in fertility is found in Guzzo and Hayford (2020) [archived].

Rates of parenthood

There are modest racial differences in the percentage of individuals who ever had a biological child. About half of women and about 40% of men aged 15−44 have ever had a biological child (Martinez et al. 2018 [archived]).

The biggest racial disparities are the higher rates of parenthood for Hispanics and the lower rates for Asians. However, when focusing only on native Hispanics, Hispanics actually have slightly lower rates of parenthood than whites. It’s also possible that rates of parenthood for native Asians would be closer to the national average, but unfortunately the data for Asians is not broken out by nativity.

The previous statistics do not control for age, so they may be slightly misleading because different racial groups have different average ages. When looking at the percentage of older adults who have ever had a biological child, there are no large differences by race. This is demonstrated in a CDC report by Martinez et al. (2023) [archived] which investigated fertility based on the 2015−2019 NSFG. Figure 1 shows the percentage of women aged 40−49 who ever had a biological child by race/ethnicity.

Table 5 shows additional data, including Asians and breakouts for Hispanics by nativity:

As you can see, by age 40, there are no large racial differences in probability of having a child. About 80-90% of women have had a child by age 40 for all racial groups. About 70-80% of men have had a child by age 40, with the exact value depending on the race/ethnicity.

While there are no large racial differences in parenthood by age 40, there are large differences among younger ages. That’s because there are large racial differences in average age of first birth, which is the topic of later sections.

Multi-partner fertility

There are also large racial differences in multi-partner fertility (or MPF). Multi-partner fertility occurs when a person has biological children with multiple spouses. Several sources show that blacks have significantly higher rates of MPF than whites, with Hispanics falling somewhere in between. Stykes and Guzzo (2018) [archived] investigated MPF using different measurement approaches in both the 2014 SIPP and the 2011−2015 NSFG. I will focus on the measures that use what the authors call the direct approach in the SIPP and the semi-direct approach in the NSFG. They are defined as follows:

  • Direct: In the SIPP, respondents over 15 years old with at least 2 biological children were asked “Do all of your biological children share the same biological mother/father?”. Respondents who answered no were considered to have MPF.
  • Semi-direct: In the NSFG, female respondents were asked about the history of their cohabiting and marital unions. For each union, women were asked whether they had any biological children, and, if so, how many. Women were recorded as having MPF if they reported children with more than 1 coresidential partner or if they reported children with only 1 coresidential partner but reported 2+ children overall. Note: this method under-estimates MPF because it doesn’t account for mothers who had no children with coresidential partners but who had 2+ children with non-coresidential partners (because the data does not have information to determine the unique number of non-coresidential partners).

Using these methods, the authors show large racial differences in MPF:

Thus, among mothers with 2+ children, about 45% of black women have MPF, compared to about 25% and 35% of white and native Hispanic women, respectively. Similar racial disparities were also reported in other analyses of the NSFG (Logan et al. 2006, Guzzo and Furstenberg 2007, Manlove et al. 2008) and the SIPP (Monte 2019). For example, Logan et al. (2006) [archived] show even larger racial disparities in MPF among American men:

Similar racial disparities in MPF in the Fragile Families and Child Wellbeing Study (FFCWS) (Carlson and Furstenberg 2006).

Parental skill

Most people are probably aware that parental skill is negatively associated with fertility. That is, more skilled parents tend to have lower fertility rates (I’m using the term “skill” fairly broadly, to include cognitive ability and educational attainment). However, many may not know that the strength of the association between skill and fertility varies by race/ethnicity. In particular, the negative association between skill and fertility is stronger for black women than white women.

Meisenberg (2010) [archived] used the NLSY79 to examine the relationship between fertility and IQ and educational attainment by race. The author reported the correlations between number of children and IQ/education by race as follows:

As you can see, among women, the negative association between skill level and number of children is much stronger for black women than white women. Among men, the negative association is actually weaker among black men compared to white men, although the differences are not large. For example, the magnitude of the correlation between highest degree attained and number of children among black women is about 80% greater than the magnitude of the same correlation among white women (r = −.324 vs r = −.182). Another interesting finding is that marital duration is much more strongly correlated with number of children among white participants compared to black participants.

Instead of just looking at the correlation coefficients, it may be useful to look at charts showing that mean number of children within each IQ range by race and sex:

As you can see, as the IQ range increases from <70 to 120−130, there is a rather sharp drop in mean number of children by black women. In fact, the mean number of children by black women with the lowest IQs (IQ < 70) seems to be about 5 times the mean number of children by black women with the highest IQs (IQ between 120 and 130). By contrast, the drop in mean number of children among white women is much more modest. The mean number of children by white women with the lowest IQs (IQ < 70) seems about 20-30% greater than the mean number of children by white women with the highest IQs (IQ > 130).

As the IQ range increases after 90-100, there also seems to be a bigger drop in mean number of children by black men compared to white men. However, the mean number of children by black steadily increases as the IQ range increases from <70 to 90-100. This may explain why Table 1 shows that the negative correlation between IQ and number of children is weaker for black men than white men (r = −.049 vs r = −.089).

Hamilton (2021) [archived] also reported fertility rates by education and race/ethnicity based on CDC data which contains 100% of all births to residents of the United States. The data shows that whereas black women have higher fertility rates than white women among women with low education, black women have lower fertility rates than white women among women with high education. For example, the total fertility rate among high school graduates is 1,823 births per 1,000 women for white women, compared to 2,381 for black women (Figure 5). However, the total fertility rate among bachelor’s degree holders is 1,307 births per 1,000 women for white women, compared to 1,111 for black women. The following chart visualizes how birth rates relate to education by race/ethnicity:

Zang et al. (2022) [archived] have also reported fertility rates by education and race/ethnicity among women in the 2006−2017 NSFG. They found similar results as the previous studies: education is negatively associated with fertility among all racial/ethnic groups, but the relationship is stronger among black and Hispanic women compared to white women. They plotted total fertility rates by education and race/ethnicity as follows:

The authors also noted that the reversed black-white gap in fertility among college educated women is primarily explained by the fact that black women are less likely to have second births. They summarized the results as follows:

This study examines completed fertility by the intersection of race/ethnicity and educational attainments for a cohort of U.S. women who are just finishing their reproductive years. We find that racial/ethnic disparities in fertility levels exist not only among women without a college degree, as documented in prior research examining earlier cohorts (Johnson 1979; Yang and Morgan 2003), but also among college educated women. Compared to their White counterparts, Black and Hispanic women without a college degree have higher fertility. However, among college educated women, Black women have lower fertility levels whereas Hispanics have higher fertility levels than Whites. The difference in the TFRs between Black and White college educated women is mainly driven by the smaller proportion of Black mothers (more than 10 percentage points less) having second births. We find little evidence that the observed racial disparities in TFRs across educational levels are driven by fertility timing differences. These results confirm our hypothesis that education plays different roles in shaping fertility levels across races/ethnicities.

Circumstances of births


This section describes racial differences in circumstances at the time of birth. This includes age of parents, the intendedness of the births, the marital status of the parents, and whether the parents were cohabiting.

Summary of section

  • Black and Hispanic women tend to be younger than white women at their first birth. Asian tend to be older than women of other racial groups at first birth.
  • The percentage of births to unmarried mothers is greatest for black women, followed by Native American women, Hispanic women, white women, and Asian women (in that order). The married and unmarried birth rates are similar among black and Hispanic women, whereas the married birth rate is substantially greater than the unmarried birth rate for white and Asian women. In fact, in recent years, there are more yearly births to married Asian women than there are to married black women.
  • The percentage of births to unmarried and non-cohabiting women is greatest for black women, followed by Hispanic women, and white/Asian women (in that order). Nearly half of black women are neither married nor cohabiting when giving birth, which is about 3-5 times the rate for women of other racial/ethnic groups.
  • Births to black women are substantially more likely to be unintended (the result of an unwanted or mistimed pregnancy) than births to white women. The percentage of births that are unintended for Hispanic women is between the percentage for black and white women. The unintended birth rate (rates of unwanted/mistimed births) to black women is about 2-3 times the unintended birth rate to white women. The unintended birth rate for Hispanic women is between the rates for blacks and whites.

One important finding is that, while there are only small racial differences in overall fertility rates (see above), there are large racial differences in the distribution of births under desirable vs undesirable circumstances. For example, black women have lower birth rates (in terms of births per woman) under desirable circumstances (e.g., married births, intended births, etc.) than white women, but they have higher birth rates under undesirable circumstances (e.g., unmarried, non-cohabiting, unintended/unwanted births, etc.).

Most of the information presented in this section has also been described by Sweeney and Raley (2014) [archived] in their review on differences in childbearing patterns by race/ethnicity.

Parental age

There are large racial differences in average parental age at birth. Blacks and Hispanics tend to be younger than whites and Asians at the time of their child’s birth. Martinez et al. (2023) [archived] reported that whereas 58% of black women aged 25 have had a child, only 37% of white and 17% of Asian women have had a child by this age (Table 5). By age 20, black women are about twice as likely as white women to have had a child (34% vs 17%) and about 7 times as likely as Asian women to have done so (34% vs 5%).

Mathews and Hamilton (2016) [archived] have also illustrated racial differences in mean age at first birth in a CDC data brief on mean age of mothers. The researchers found that the mean age at first birth has increased in recent years for mothers of all racial/ethnic groups. The racial/ethnic differences in maternal age was consistent with the data presented earlier:

More recent trends show that the mean age of mothers has increased even more in recent years. In the most recent CDC report on births in the United States by Osterman et al. (2023) [archived], the mean age of first births in 2021 was 31.2 years for Asian women, 28.1 years for white women, 25.5 years for black women, and also 25.5 years for Hispanic women (Table 11). Thus, the mean age at first birth for black and Hispanic women is about the same as the mean age for whites in 2000. The following chart shows the birth rates by race/ethnicity and age of mother in 2021.

There were similar results in the CDC report on births in 2019 by Martin et al. (2021), so these numbers are not largely inflated due to COVID.

Much of the higher mean age for Asian mothers is likely to be explained by their foreign-born status. For example, in the CDC report on births in 2003 by Martin et al. (2005) [archived] (which is the most recent report I found that broke out the Asian data by nativity), only 3.5% of Asian mothers were under the age of 20 at birth, compared to 9.4% for white mothers (Table 13). However, among mothers born in the United States, 11.5% of Asian mothers were under 20 years old compared to 9.4% of white mothers.

Marital status

Unsurprisingly, there are also large racial differences in births by marital status. Most people are probably already familiar with the fact that there are large racial differences in the percent of births to unmarried mothers (i.e. “out-of-wedlock” birth rate). I cited trends on rates of unmarried births by race in a previous post:

There are some interesting findings to note when comparing the absolute numbers of births by marital status and race. I’ll focus on these numbers from the most recent CDC report on births in the United States by Osterman et al. (2023) [archived]. Table 11 indicates the total number of births and the percent of births to unmarried mothers by race/ethnicity. This allows easy calculation of the total number of married births and the total number of unmarried births. I used these calculations to create the following chart:

Unsurprisingly, most of the births to black women are unmarried whereas most births to whites and especially Asians are married. There are a few other interesting findings to note just from the absolute numbers:

  • The absolute number of births to married Asian women exceeds the absolute number of births to married black women (187k vs 155k), despite the fact that black women greatly outnumber Asian women and the fact that black women have slightly greater overall fertility rates.
  • The absolute number of births to unmarried black women is fairly close to the absolute number of unmarried births to white women, despite the large gap in population between black and white women.

It should be noted that the extremely low non-marital birth rate for Asian women is likely to be explained by their foreign-born status. For example, consider the CDC report on births in 2003 by Martin et al. (2005) [archived], which is the most recent report I could find that broke the data out by nativity. In this report, only 15% of Asian mothers who gave birth were unmarried at the time of birth, compared to 29.4% of white mothers (Table 13). However, among mothers born in the United States, 31.6% of Asian mothers were unmarried compared to 25.5% of white mothers.

Interestingly, married black women actually have lower birth rates than married women of other major racial/ethnic groups, including Asian and white women. For example, in a report on pregnancy outcomes, Ventura et al. (2012) [archived] report that the birth rate in 2008 for married women aged 15−44 was 69.7 births per 1,000 women for black women, compared to 88.0 for Hispanic women, and 87.5 for white women (Table 5). In a previous post, I presented data on married birth rates by race from 1940 to 2010. The data shows that the white married birth rate did not begin exceeding the black married birth rate until the 1970s.

There are a few interesting findings to note here:

  • The married birth rate is lowest for black women. This shows that the higher overall birth rate for black women compared to white and Asian women is entirely driven by unmarried births.
  • The married and unmarried birth rates are similar among black and Hispanic women, whereas the married birth rate is substantially greater than the unmarried birth rate for white and Asian women.

Married black women have lower birth rates than married white and Asian women even when looking at specific age groups. According to a CDC report [archived] on birth rates for married women from 1950 to 2003 (the latest data that I could find on age-specific birth rates for married women), the birth rate in 2003 for married women aged 20−24 years was 200.1 per 1,000 women for non-Hispanic white women, 181.9 for black women, 207.1 for Asian women, and 220.0 for Hispanic women. For married women aged 25−29, the rates were 166.1 for non-Hispanic white women, 119.9 for black women, 167.9 for Asian women, and 149.9 for Hispanic women. For historical data on age-specific birth rates for married women by race/ethnicity, see Ventura and Bachrach (2000) [archived].

Cohabitation status

The non-marital births to black mothers predominantly involve women who are not cohabiting. For example, Martinez et al. (2023) [archived] shows that most black mothers (58.9%) were never married and not cohabiting at the first birth. By contrast, only 14.9% of white mothers were in a similar situation. Thus, black mothers were about 4 times as likely as white mothers to be never married and not cohabiting at the time of their first birth.

The graph does not include Asian mothers due to the low sample size of Asian non-marital births. However, Table 7 shows that 87.8% of Asian mothers were married or formerly married at the time of their first birth. Thus, the percent of Asian mothers who were never married and not cohabiting would have to be some fraction of 12.2%, much lower than the percent for the other groups.

That report only shows the relationship status of mothers at the time of their first birth. It does not show relationship status for all births. This information was provided in a report of an earlier cycle of the NSFG: Martinez et al. (2012) [archived] used the 2006−2010 NSFG to show that nearly half (46%) of black mothers were neither married nor cohabiting at the time of birth in 2006−2010, compared to just 9-16% for other racial groups. Thus, black mothers were about 3-5 times as likely as other groups to be neither married nor cohabiting when giving birth.

This data also show that most unmarried Hispanic and white mothers were cohabiting at the time of delivery, whereas the opposite is true for black women. Similar results were reported in the 2006−2008 NSFG by Lichter (2012).

After forming a cohabiting premarital union, there are also racial differences in the probability of a pregnancy occurring within that union. For example, Copen et al. (2013) [archived] shows that the probability of a pregnancy within 2 years of a first premarital union was about 40% for black and U.S. born Hispanic women, compared to about 24% for white and Asian women (Table 4). Note that this only includes pregnancies that lead to live births. Thus, even though cohabiting unions for black women are less stable and less likely to transition to marriage (see above), cohabiting unions for black women are nevertheless more likely to lead to a birth.

The previous section showed that married and unmarried birth rates are similar among black and Hispanic, unlike white and Asian women. We can also compare birth rates between cohabiting and single women (where single means neither cohabiting nor married). This information was reported by Kim and Raley (2015) [archived] using the 2006−2010 NSFG. They plotted fertile pregnancy and expected pregnancy rates for cohabiting and single women as follows:

Note: “fertile pregnancies” are pregnancies that result in birth (as opposed to abortion or miscarriage). The authors calculated “expected” pregnancy rates based on information about contraceptive use by race/ethnicity and published estimates of pregnancy risk for each contraceptive method.

As you can see, the fertile pregnancy rate for cohabiting white women is about 3 times the rate for single white women. However, the fertile pregnancy rate for cohabiting black women is roughly equal to the rate for single black women. For U.S. born Hispanics, the ratio of cohabiting rates to single rates falls somewhere between the ratios for whites and blacks. Furthermore, the fertile pregnancy rate for single black women is about 4 times the rate for single white women. Thus, relative to whites and to a lesser degree Hispanics, a much larger portion of non-marital births for black women are contributed by women who are not even cohabiting.

In fact, these authors conclude that the pregnancy rate among single women (not cohabiting women) is the biggest contributor to racial differences in unmarried birth rates. By contrast, post-conception marriages (i.e. shotgun marriages) explains a relatively small portion of the gap:

Table 3 presents the results of the decomposition of the non-marital fertility rates. Each number represents the proportion of the racial-ethnic difference that is due to a specific factor, and these numbers sum to 1. For example, the 0.64 in the non-marital pregnancy rate among singles row indicates that 64% of the black-white difference in the non-marital fertility rate is due to black-white differences in the non-marital pregnancy rate among sexually active single women. The higher levels of sexual activity among black singles accounts for 17 % of the gap, but this is somewhat offset (− 4%) by lower levels of cohabitation among blacks. Higher fertile pregnancy rates among cohabiting blacks also contribute slightly (14%) to the black-white gap. Overall, the results indicate that by far the biggest contributor to the black-white gap in the non-marital fertility rates is differences in the non-marital pregnancy rate among sexually active single women. In contrast, post-conception marriage contributes very little (overall 7+1=8%) to black-white differences in non-marital fertility.

The black-white gap in non-marital fertility rate is also partially explained (15%) by the black-white gap in pregnancy rates among sexually active cohabiting women. But a much larger portion of the gap is explained by the black-white gap in pregnancy rates among sexually active single women.

The previous data has focused on the relationship status of women at the time of birth. It is also interesting to note the relationship status of men as well. As with women, there are also large racial differences in the relationship status of men at the time of birth. Martinez (2015) [archived] used data from the 2002, 2006−2010, and 2011−2013 NSFG to examine non-marital first births among fathers from 1980 to 2009. The author reported data on the relationship status of fathers at first birth by race/ethnicity (Figure 2).

The non-marital birth rates presented are not surprising. About 70% of births to black fathers were non-marital, compared to about 50% of births to Hispanic fathers, and about 30% of births to white fathers. This is all consistent with all of the data presented thus far.

However, we can also calculate the percent of births to single (i.e. neither married nor cohabiting) fathers as well, since cohabiting births are just a subset of non-marital births. For example, among first births in 1980−1989, 54% of black fathers (77% – 23%) were single at the time of their first birth, compared to 23% of Hispanic fathers, and 15% of white fathers. The numbers have decreased for all racial groups among births in 2000−2009: 36% of black fathers were single at the time of their first birth, compared to 16% of Hispanic fathers, and just 6% of white fathers.

Thus, when black men have their first child, it is very common for them to neither be married to or cohabiting with the mother of the child. By contrast, this is relatively rare for Hispanic and especially white men.

Intendedness

There are large racial differences in rates of unintended and unwanted births. The definitions of “unintended” and “unwanted” are based on the definitions provided in the earlier section on unintended pregnancies. An unintended birth is a birth that is the product of an unintended pregnancy, whereas an unwanted birth is a birth that is the product of an unwanted pregnancy.

The first source on racial differences in births by intendedness comes from the National Survey of Family Growth. Mosher et al. (2012) [archived] examined the 2006−2010 NSFG and found that black women had the highest percent of unintended births, followed by Hispanic women, and white women (in that order):

Thus, the percent of births that were unintended among black women was over twice the same percent for white women (45% vs 20%). Similarly, the percent of births that were unwanted was also over twice as high for black women compared to white women (23% vs 9%).

Similar findings can be found in some of the previously cited studies that focused on unintended pregnancies (rather than unintended births) in different cycles of the NSFG (Finer and Zolna 2016, Finer and Henshaw 2007). For example, Finer and Zolna 2016 [archived] showed that the rate of unintended pregnancies ending in birth in 2011 was about twice as high for black women as for white women (33 vs 17 pregnancies per 1,000 women aged 15−44, Table 2). The rate for Hispanic women was about the same as the rate for black women.

The second source comes from the Early Childhood Longitudinal Study−Birth Cohort (ECLS−B). Guzman et al. (2010) [archived] examined a sample of about 9,000 from the ECLS−B who could provide information on childbearing intention and relationship status. The study found significant racial/ethnic differences in childbearing intention. The authors described the main findings as follows:

In the full sample, 43% of births were reported as unintended, including 30% that were mistimed and 13% that were unwanted (Table 2). As we had expected, the proportion of births that were unintended was higher among mothers who had been cohabiting (58%) or outside of a coresidential union (72%) at the time of conception than among mothers who had been married (28%), and women who had been outside of a coresidential union were more likely than women in the other two groups to report a birth as unwanted (30% vs. 7–16%). The majority (66%) of births to black mothers were characterized as unintended, compared with roughly one-third of births to whites (36%) and Asians (33%), and about one-half of those to Hispanics (46%) and women of other races and ethnicities (53%). Additionally, blacks reported that a higher proportion of births were unwanted than did mothers of all other races and ethnicities (33% vs. 9–18%).

I illustrated the full results in the following chart:

One interesting finding is that the racial differences in relationship status only partially explain racial differences in childbearing intentions. Among cohabiting women, 60−65% of births were unintended for all racial groups except foreign-born Hispanics (Table 2). Also, among women who are neither married nor cohabiting women, 70−75% of births were unintended for all racial groups (again, excluding foreign-born Hispanics). However, there are still some key racial differences:

  • Even among these unmarried women, there are large racial differences in rates of unwanted births. For example, among women who are neither married nor cohabiting, 42% of births to black women were unwanted compared to just 22% of births for white women (Table 2).
  • Among married women, there are large differences in rates of both unwanted and unintended births. Among married women, black women have the highest rate of unintended births (48%), followed by Hispanic women (35%), Asian women (26%), and white women (24%) (Table 2). Similarly, married black women have the highest rate of unwanted births (19%), followed by Hispanic women (7.6%), Asian women (7.0%), and white women (5.2%). In fact, the rate of unwanted births for married black women is greater than the rate for cohabiting white women (15%), and is nearly as great as the rate for white women who are neither married nor cohabiting (22%).

The 3rd source comes from The 1979 National Longitudinal Survey of Youth (NLSY79). Musick et al. (2009) [archived] investigated education differences in intended and unintended fertility. The participants of the survey were interviewed either annually or every other year until 2004, when they were 42 years old on average, which is very near the end of child-bearing years. The authors used the participants’ birth histories to estimate completed fertility rates disaggregated by education and race (completed fertility rate is the number of births per woman). I illustrated the overall findings by race (without disaggregating by education) in the following chart:

The authors described these findings as follows:

Table 1 shows the average number or intended, mistimed and unwanted births to white and black women in 2004, at the last year of the survey. Recall that respondents are 39–46 years old at last interview, i.e., at or very near the end of their child-bearing years. These are thus reasonable estimates of completed child-bearing for this cohort. The first column of Table 1 shows average fertility levels pooled across education groups. As is well documented (Yang and Morgan 2003), whites have lower overall fertility than blacks, although neither group is far from replacement level, with whites averaging 1.85 children per woman and blacks, 2.18. The share of all births that are unintended differs by race, with 23 percent (i.e., . 42/1.85) mistimed and 4 percent unwanted among whites, compared to 38 percent and 18 percent, respectively, among blacks.

The exact results of these three different samples have some differences. For example, the percent of unwanted births is about equal to the percent of mistimed births in the NSFG for all groups, but the percent of unwanted births is much lower in the NLSY79. Also, the percent of intended births is relatively low in the ECLS−B compared to the NLSY79 and NSFG. However, there is a consistent pattern of racial disparities across all samples: black women are consistently 2−4 times as likely as white women to have an unwanted or mistimed birth.

Families with children


This section describes racial differences in families with children. This includes racial differences in single parenthood, presence of half-siblings, and rates of transition.

Summary of section

  • The single-parent household rate is greatest for black children, followed by Hispanic children, white children, and Asian children (in that order). Roughly half of black children live in a single parent household, which is about 3-4 times the rate for other children.
  • The percentage of children living with half siblings is greatest for black children, followed by Hispanic children, white children, and Asian children (in that order). Black children were about 2-3 times as likely to live with a half sibling as white children.
  • Black children tend to experience more family transitions throughout their childhood than white children. Hispanic children tend to experience transitions at rates similar to whites, although they sometimes experience transitions at rates between the white and black rate depending on the dataset.

Racial differences in family structure are so great that the absolute number of certain undesirable family structures is greater for blacks than whites, and the absolute number of certain desirable family structures is greater for Asians than blacks, despite the large gaps in the respective population sizes. For example, there are more family groups headed by a solo never-married black mother than there are family groups headed by a solo never-married white mother. As another example, there are more married Asian couples with children than there are married black couples with children. Again, these are absolute numbers, not percentages or rates.

Single parenthood

Unsurprisingly, there are large racial differences in rates of single parenthood. Using data from the CPS, Anderson et. al (2022) [archived] reported percentage of children living in different household structures by race/ethnicity. For example, the percent of children living in two married households in 2019 was about 38% for black children, compared to 61% for Hispanic children, 75% for white children, and 85% for Asian children (Figure 2a). By contrast, black children were much more likely than other children to live with just one parent. About 46% of black children lived with their mother only compared to 24% for Hispanic children and <15% for whites and Asians.

When accounting for children living with their father only, this data shows that most black children live with a single parent (50.8% = 45.7% + 5.1%). This data closely matches analysis of 2017 CPS data by Pew Research.

Another interesting finding is that the vast majority of children living with a single parent are not living with a cohabiting parent. Instead, most of these children are living with a single (i.e. non-cohabiting) mother. Other studies have also demonstrated that most children with unmarried parents are living with single parents as opposed to cohabiting parents (see Manning et al. 2014 analyzing SIPP data, Table 3).

Vespa et al. (2013) [archived] showed that, among family groups maintained by white mothers in 2012, 47% (2,506 out of 5,371) of mothers were divorced and 31% (1,649 out of 5,371) were never married (Table 6). Among family groups maintained by black mothers, the reverse was true: 20% (692 out of 3,545) of mothers were divorced and 64% (2,262 out of 3,545) were never married (Table 6). So, not only are black families more likely to be maintained by a single mother than white families, black single mothers are vastly more likely to be never married than white single mothers. In fact, the absolute number of solo never-married black mothers is considerably greater than the same number for white mothers (2,262 vs 1,649), despite the fact that white women greatly outnumber black women. The following chart shows an illustration of the same data:

Thus, racial disparities in the percentage of children in never-married single-parent households are far greater than the racial disparities in the percentage of children in single-parent households generally. That was also covered in a previous post.

Racial disparities in marital structure are so large that the absolute number of Asian married family groups with children exceeds the absolute number of black married family groups with children (2,094 vs 2,189, see Table FG10 of America’s Families and Living Arrangements: 2023). This is despite the fact that the black population is over twice the size as the Asian population (Census 2023). The following chart shows the number of family groups with children under 18 disaggregated by family structure and race/ethnicity:

The same data is presented as a visualization on the census site here (though the stability of this link is not clear).

Half-siblings

There are also large racial differences in the presence of half-siblings. This is to be expected given the large differences in multi-partner fertility. Anderson et al. (2022) [archived] used data from the 2014 SIPP to show the percent of children living with half siblings by race/ethnicity in 2014 (Table 3). The biggest gap is between black and Asian children: about 31% of black children were living with half siblings compared to just 3% of Asian children. The percentages for other racial/ethnic groups are presented as follows:

Interestingly, the same data shows that the absolute number of children living with a never-married solo mother and half siblings is 3 times larger for black children compared to white children (1,718k black children vs 496k white children, Table 3).

It should be noted that these estimates for the prevalence of half-siblings are much greater than the estimates from some previous reports of the SIPP data, such as Manning et al. (2014). Anderson et al. note that this difference stems from the fact that the 2014 redesign of the SIPP used more accurate measures of half-siblings.

The first wave of the 2014 SIPP panel estimated that 12.5 million children (17.1 percent) under the age of 18 lived with at least one half sibling. This is notably higher than the earlier estimate of 8.0 million children (10.8 percent) published from the 2008 SIPP panel. The 2008 estimates were derived from responses to questions asking how each household member was related to everyone else in the household. A technical term such as half sibling is seldom used in daily life and was likely reported less frequently than the other response categories listed for siblings. In the 2014 redesign of the SIPP instrument, deriving half sibling relationships from the fertility history information reported for coresident parents provides more detailed information than was previously available, yielding a higher estimate of children living with siblings with whom they share one biological parent. Some of the changes shown are a result of changes in measurement and do not reflect true changes in the underlying population.

The previous data shows the percent of children living with half siblings. It may be more appropriate to look at the percent of children with half siblings among children with at least 1 sibling. Amorim and Tach (2019) reported this information from both the NLSY79 and NLSY97. The authors found large differences in the presence of half-siblings by race/ethnicity. The biggest racial gap in presence of half siblings was between black and white children in the NLSY79 cohort, as 46% of black children vs 17% of white children had a half-sibling (among children with 1+ siblings, Table 4). The results for each race/ethnicity and cohort are presented as follows: 

Measures of half sibling prevalence in this study has some advantages over the prevalence reported in other studies:

  • This study estimated the presence of half siblings regardless of whether the half siblings resided in the same household.
  • This study estimated the presence of half siblings among children at any point before turning 18 years old, as opposed to other studies which measure the presence at some point in time.

Racial differences in the presence of half siblings were also reported by Cancian et al. (2011). The authors used the KIDS dataset, which contains records on 86% of all non-marital children born in Wisconsin. The authors focused on roughly 8,000 non-marital children born in 1997 who were their mother’s first child. They then estimated what percent of these children acquired new half-siblings over the first 10 years of their life. They found that black children were much more likely to acquire half siblings than children of other races.

For example, among children with two black parents, about 72% of children acquired a new half sibling by age 10, compared to 50% of children with two Hispanic parents, and 42% of children with two white parents (Table 3). In fact, 29% of children born to two black parents acquired a new half sibling on both their father’s side and their mother’s side, compared to 13% of children to two Hispanic parents, and 10% of children to two white parents.

Transitions

Unsurprisingly, there are also racial differences in changes in family structure over time. For example, Daniels et al. (2017) [archived] used the FFCWS to study racial differences in transitions to marriage among mothers following a non-marital birth. Before considering the findings on transitions to marriage, the authors noted racial differences in cohabitation among unmarried mothers:

Around 48% of mothers with a nonmarital birth were cohabiting with their baby’s father. We also observed significant racial differences since 61% Whites, 59% Hispanics and only 35% Blacks cohabited with the baby’s father. This statistic significantly decreased to 12% by year nine suggesting that majority of the relationships either dissolved or transitioned into marriage.

Thus, white mothers were nearly twice as likely as black mothers to cohabit with the baby’s father following a non-marital birth. This is consistent with the data presented earlier.

Now, consider racial differences in rates of transition to marriage. The following chart shows the percentage of mothers who married the father of their child following a non-marital birth.

As you can see, black mothers were significantly less likely than white and Hispanic mothers to marry their child’s father. Whereas about 15-20% of black mothers married their child’s fathers 9 years after the child’s birth, about 35-40% of white and Hispanic mothers did the same.

Here is a similar chart for fathers:

The results for fathers are similar to the results for mothers, except now there is a significant white-Hispanic gap for fathers which was not present for mothers. Whereas about 50% of white fathers married the mother 9 years after a non-marital birth, only about 20% of black fathers did the same.

Brown et al. (2016) [archived] also reported racial differences in family transitions of children from birth to age 12 in the 1995 and 2006−2010 NSFG. Transitions were defined as either entering directly into a marital or cohabiting union or dissolving into a marital/cohabiting union (moving from a cohabitation to a marriage was not treated as a family transition). The authors found that black children experienced nearly twice as many transitions as white children by age 12 in the 2006−2010 dataset (1.63 vs 0.87). The number of transitions by age and race/ethnicity in 2006−2010 are plotted as follows:

The authors also presented information on life course trajectories through age 12 by race/ethnicity and by marital status of parents at birth (Table 3).

  • Among children born to married mothers in the 2006−2010 NSFG, 55% of parents of black children remained married by age 12, compared to 66% of parents of Hispanic children, and 70% of parents of white children. Married parents of black children were about twice as likely to split and never repartner as parents of white and Hispanic children (31% vs 14-15%).
  • Among children born to cohabiting mothers, 50% of parents of black children transitioned to marriage, compared to 61% of parents of Hispanic children, and 71% of parents of white children. Cohabiting parents of black children were also about twice as likely to split as parents of white and Hispanic children (46% vs 21-26%).
  • Among children born to single mothers, 25% of mothers of black children remained single through age 12, compared to 22% of mothers of Hispanic children, and 8% of mothers of white children.

Similar findings were reported in Copen et al. (2013) and Raley and Wildsmith (2004) using earlier or more limited NSFG data.

Description of datasets


Datasets gathered by the U.S. Census Bureau

  • Decennial census dataThe decennial census [archived] is conducted every 10 years to count every person living in the United States. Respondents are asked a short list of questions concerning questions such as age, sex, race/ethnicity, etc.
  • The American Community Survey (ACS). The ACS [archived] is conducted each month and is sent to a smaller sample of respondents (about 3.5 million addresses). Respondents are asked about a broader set of topics, including education, employment, transportation, etc. The U.S. Census Bureau describes [archived] it as “the premier source for detailed population and housing information about our nation”.
  • Survey of Income and Program Participation (SIPP). The SIPP [archived] is a continuous series of panels conducted by the U.S. Census Bureau. Each panel follows a large nationally representative sample of households over a four-year period. The survey provides information on income, employment, household composition, and government program participation. The Census Bureau describes the SIPP as “a leading source of data on economic well-being, family dynamics, education, wealth, health insurance, child care, and food security.”
  • The Current Population Survey (CPS). Sponsored jointly by the U.S. Census Bureau and the Bureau of Labor Statistics, the CPS [archived] provides monthly labor force statistics about the United States population. It is one of the “oldest, largest, and most well-recognized surveys in the United States”. Supplementary surveys are sometimes  conducted which include information on other topics. Some of the studies below will reference the following CPS supplements:

Datasets gathered by the CDC

  • CDC “Births” data. Each year, the CDC reports summaries of data collected from birth certificates in the United States. In the 2021 report, data was reported on the 3.66 million births that occurred that year. Information is presented on a variety of factors such as maternal age, race/ethnicity, marital status, etc.
  • CDC “Abortion Surveillance” data. Each year, the CDC requests and analyzes abortion data from the central health agencies of each of the 50 states, the District of Columbia, and New York City. The CDC releases this data each year in their “Abortion Surveillance” reports. In the 2021 report [archived], about 625,000 abortions were reported from 48 reporting areas.
  • The National Survey of Family Growth (NSFG)The NSFG [archived] has periodically collected information on fertility, relationship status, family life, etc. across nationally representative samples of specified portions of the United States population. The first 2 cycles of the NSFG (conducted in 1973 and 1976) focused specifically on ever-married women aged 15-44. Cycles 3-5 (conducted in 1982, 1988, and 1995) expanded the sample to include women regardless of marital experience. Cycle 6 (conducted in 2006) expanded to include men. Starting in 2006, the NSFG shifted from periodic surveys to continuous interviewing, where fieldwork was performed each year. A list of all publications using the NSFG can be found here [archived].
  • The Youth Risk Behavior Surveillance System (YRBSS). The YRBSS [archived] is a set of surveys conducted by the CDC which track risky behaviors among high school students. For my purposes, I will focus on the results from the Youth Risk Behavior Survey (YRBS), which are surveys administered every other year between 1991 and 2021 to nationally representative samples of students in grades 9-12 (see this report on methodology [archived]). The surveys include questions on activities such as drug use, dietary behaviors, sexual behaviors, etc.
  • The National Health Interview Survey (NHIS). The NHIS [archived] has been conducted continuously since 1957 to collect information on the health of the country. Each month’s sample is a nationally representative portion of the civilian noninstitutionalized population of the United States.

Datasets gathered by other government agencies

  • The National Longitudinal Survey of Youth (NLSY). The 1979 National Longitudinal Survey of Youth [archived] (NLSY79) and the 1997 National Longitudinal Survey of Youth [archived] (NLSY97) are 2 nationally representative longitudinal studies published by the U.S. Bureau of Labor Statistics for 2 cohorts of Americans. Both surveys collected data on several thousand subjects (approximately 10,000), beginning when the participants were in their youth. In the NLSY79, participants were initially interviewed in 1979, when they were aged 14 to 22. In the NLSY97, participants were initially interviewed in 1997, when they were aged 12 to 17. Both cohorts were re-interviewed in periodic follow-up waves (every 1-2 years).
  • The Early Childhood Longitudinal Study-Birth Cohort (ECLS−B). The ECLS−B [archived] is a longitudinal study of U.S. children conducted by the National Center for Education Statistics (NCES) (agency of the U.S. Department of Education). The study follows a nationally representative sample of approximately 10,700 children born in the year 2001 and examines various aspects of their development, including health, early care and education, and family environment.

Other datasets

  • The General Social Survey (GSS). The GSS [archived] is a nationally representative survey of U.S. adults. The survey has periodically collected data from 1972 until now. Each survey contains questions on topics such as demography, behavior, attitudes, well-being, etc.
  • The National Longitudinal Study of Adolescent Health (Add Health). Add Health [archived] is a longitudinal study of a nationally representative sample of adolescents in the United States. The initial wave included over 20,000 adolescents in grades 7-12 during the 1994-95 school year. The adolescents were followed for 5 waves, with most recent data in 2016-2018. It is described [archived] as “the largest, most comprehensive longitudinal survey of adolescents ever undertaken”.
  • The American’s Changing Lives survey (ACL)The ACL [archived] is the longest ongoing national study of the factors that influence health outcomes throughout adulthood. The initial study began with a survey of 3,617 adults in 1986. Participants have been re-interviewed in follow-up waves in 1989, 1994, 2001, 2011, 2019.
  • The Fragile Families and Child Wellbeing study (FFCWS). The FFCWS [archived] studies a sample of about 5,000 children born in large U.S. cities (population over 200,000). The sample consists of children born between 1998 to 2000. Follow-up interviews were collected in later waves when children were about 1, 3, 5, 9, 15, and 22 years of age. The weighted sample is representative of births in large U.S. cities.
  • The 1992 National Health and Social Life Survey (NHSLS). The NHSLS [archived] gathers data on sexual behaviors of about 3,500 adults in the United States. The survey is nationally representative of English-speaking adults aged 18-59.
  • The National Survey of Families in Households (NSFH)The NSFH [archived] is a national sample of about 13,000 respondents from about 9,600 households interviewed during the first wave in 1987−88. The survey collects information on respondents’ history of marriage, cohabitation, education, fertility, etc. Data for later waves were gathered in 1992−94 and 2001−02.

 

1 comment on Racial differences in relationship and childbearing patterns

Leave a Reply

Your email address will not be published. Required fields are marked *