|Fact Sheet Marriage and Education|
WOMEN'S EDUCATION AND THEIR LIKELIHOOD OF MARRIAGE:
A HISTORIC REVERSAL
A FACT SHEET PREPARED FOR
THE COUNCIL ON CONTEMPORARY FAMILIES
Paula England and Jonathan Bearak, New York University
Historically, women who graduated from college were far more likely than any other group of women -- whether high school dropouts, high school graduates, or women with some college - to remain single their entire lives. As late as 1950, a quarter of white female college graduates 40 years of age had never married, compared to only 7 percent of their counterparts without a college degree. (See this CCF Report.) But what has happened since women have been completing college and obtaining advanced degrees at much higher rates, and since divorce has become easier to obtain?
On the other hand, poorer and less-educated men and women are most apt to have children outside marriage while men lacking high school or college degrees are sometimes seen by themselves and their partners as less "marriageable" because stagnating wages have made them less reliable breadwinners (see Appendix A). The latter suggests the less educated are less likely to be married.
The key to resolving this puzzle is to realize that education affects not only the timing of marriage among those who eventually marry, but also whether people ever marry, and whether they divorce. As we'll see, these effects come into play at different ages for men and for women at different educational levels, and they also vary by race.
To show how the education differences in marriage rates vary by age, we have graphed marriage rates separately by age for each of four education groups-those who ultimately finished less than high school, just high school, some college, and four years of college. We show the education-by-age graphs separately for blacks and whites, as well as for all races combined, and separately for men and women. Our data come from a national sample of people who were born between 1958 and 1965 and therefore turned 20 between 1978 and 1985, a period of rapid cultural and socioeconomic change. These individuals were followed for many years, allowing us to know their marital status at every age through 45. Most people who ever marry have done so by age 45, so, roughly speaking, we are able to see what percent of those in each group will ever marry.
Figure 1 shows the percent of men or women who have ever married by each age, separately for each group by their ultimate education. In Figure 2, we look at the same thing except that we show what percent are "currently married" rather than what percent have "ever married." The difference between who is "currently" and "ever" married for any group is that those who have married but are now divorced or widowed and not remarried are included in the "ever married" but not the "currently married" figures. (More details on the data and our statistical analysis can be found in Appendix B.)
Educational Differences in When or If Individuals Marry
Getting a lot of education delays marriage, but for women and men born after 1958, unlike for women born early in the 20th century, the well educated are just as or more likely to ultimately marry as any other group. In Figure 1 we see that up to sometime in the 20s, the more education you have, the less likely you are to have married, with college graduates the least likely and those with less than a high school education the most likely. But somewhere between the mid 20s and late 30s a catch-up of the more educated occurs. Although people who get more education typically wait till after they complete schooling to marry, making their marriages later, they are just as likely to ultimately marry by about age 40. In all education groups, roughly 75-90% will ever marry.
The patterns for whites look about the same as the patterns for the whole sample. But among black women, those with a college degree not only catch up to the marriage rates of the less educated, but far surpass them. Overall, black women have lower odds of ever marrying than white women, but education affects ultimate marriage rates by the 30s and 40s much more substantially for blacks than whites. The only group of women where a majority have not married by age 45 is black women with no college education.
The overall pattern for men is similar. Like women, men who complete fewer years of schooling are more apt to marry at young ages, but like women, higher-educated men catch up later. However, men's results differ from women's in three key ways. First, men typically marry a couple of years later than women, reflecting the convention that men in first marriages are slightly older than their wives. Second, higher-educated men catch up to lower-educated men sooner than higher-educated women catch up to lower-educated women. Finally, whereas black women with some college are as likely to marry by age 45 as black women who graduate from college, there is a notable difference between black men who do and do not finish college, with college graduates much more likely to marry than men with only some college.
On average, almost half of American marriages will end in divorce, but divorce rates are lowest for college graduates, and the education gap in divorce has increased in recent decades. (See McLanahan, Sara. 2004. "Diverging Destinies: How Children Fare Under the Second Demographic Transition." Demography. 41(4): 607-627.) There are several reasons for this. Less educated individuals typically marry at an earlier age, which is associated with higher divorce rates. Additionally the lower incomes and greater economic insecurity of those with less education increase stress, affecting divorce.In Figure 2, we look at the percent of people who are currently married at each age, separately for each education group. These figures differ from the "ever married" percents in Figure 1, in that those who have married but are divorced and have not remarried appear in Figure 1 as married but not in Figure 2 as currently married.1 Thus, the "currently married" figures are influenced by educational differences not only in who marries, but also in who has divorced.
Looking first at women of all races, we see that after age 30 college graduates pull ahead in the percentage currently married - even though they don't catch up with the two education groups right below them in the percentage ever married until their late 30s. By age 40 college graduates are significantly ahead of the two education groups right behind them in the percentage currently married, while female high school dropouts remain way behind. About 75% of college graduates are married at age 40, compared to about 70% of those who attained high school or some college and only about 60% of those who didn't complete high school. In yet another historical reversal, women with less than a high school education are the least likely to be currently married at all ages after 30.
The figures for white women look quite similar to those described here for all women combined. But for black women, the percent currently married varies more dramatically by education than for whites. This is partly because, as we saw above, education differences are larger for blacks than whites in whether they ever marry by age 40. But there are also education differences in divorce for blacks, as for whites, and the higher divorce rates of less-educated black women also contribute to the difference in how many are currently married.
Educational differences in the percent currently married are larger for men than women. Overall, as we saw for women, black men have much lower marriage rates than white men. The only exception is black male college graduates by age 35. By age 45, about 75% of college-educated black men are married, whereas less than 55% of college-educated black women are married at the same age.
Men who attain less than a high school education are much less likely to be married than any other education group of men at most ages. Among white men, a similar percentage of all education groups except those who do not complete high school are married at age 30, whereas among black men, college graduates have much higher and high school dropouts much lower percents married compared to the middle education groups.
In sum, for both men and women, past age 30 or 35, education differences in who is currently married-favoring the well educated-are much larger than differences in who has ever married. The added educational differentials in who is currently married come from educational differences in divorce rates.
Early in the life cycle, those who ultimately get more education are less likely to have married than their less educated counterparts. This is because those staying in school longer also delay getting married longer. But by age 40, the well educated have caught up with the less educated and even surpassed them in the percent that have married. Thus, ultimately the more educated are as likely or even more likely to marry as any other group. The education differences in whether people ever marry are small for whites, but quite large for blacks, owing partly to the very low marriage rates of the most disadvantaged blacks-those without high school or less.
If we look at who is currently married by age, educational differences get even more dramatic. Because the less educated divorce at higher rates, fewer of them are left still married as the decades go by. These differences are particularly pronounced for men.
In sum, if we focus on the early to mid-20s, a higher percent of the less educated are married. The higher educated groups catch up and pull ahead in their late 20s and 30s, possibly because more of them have the economic resources that young people now consider a prerequisite to marriage. (See Appendix A.) If we focus on the rest of life (represented in our data up to age 45), educational differences in those who are currently married are even larger once people move into their late 30s and 40s because those at lower educational levels have higher divorce rates.
APPENDIX A: WHY DO LESS EDUCATED INDIVIDUALS MARRY AT LOWER RATES?
One of the main reasons less-educated individuals have lower rates of marriage is that rising expectations of marriage since the 1970s have been colliding with declining economic prospects for less educated workers. Interviews with impoverished individuals reveal that while they have high regard for marriage, they think that couples need to have stable jobs and be able to pay the bills for marriage to be appropriate, and those with the lowest education often can't meet this bar. (See Paula England and Kathryn Edin, "Understanding Low-Income Couples with Children," Council on Contemporary Families, September 24, 2007; Stephanie Coontz, "Economic disparity takes toll on marriage," Philadelphia Inquirer, January 9, 2011. See also Gibson-Davis, Christina, Kathryn Edin, and Sara McLanahan. "High Hopes, but Even Higher Expectations: The Retreat from Marriage Among Low-Income Couples." Journal of Marriage and Family. 67(5): 1301-1312 and Cherlin, Andrew. 2004. "The Deinstitutionalization of American Marriage." Journal of Marriage and Family 66:848-861.)
Our analysis makes use of the National Longitudinal Survey of Youth that began in 1979. This study employed a national probability sample of youth age 14-21 in 1979 (born in 1957 to 1964).2 The cohort was interviewed every year or two; the study is ongoing, and we make use of the data through 2008 when the cohort was all at least 45 years of age.
African Americans were oversampled in the study, so that they would have a large enough sample size for separate analysis. We use sample weights provided by the Center for Human Resource Research at Ohio State University, which releases the data. Sample members in any group that was over-sampled are weighted down, so that the figures reflect the national population in that birth cohort. Graphs labeled "Black" pertain to non-Hispanic blacks; graphs labeled "White" pertain to non-Hispanic whites. Other races, as well as Hispanics of any race, are included in the graphs labeled "any race."
The study provides respondents' education in each year. However, for this analysis, we classified respondents by the final education they had attained by the latest wave of the survey, collected in 2008. (This procedure is most important for interpreting the figures for college graduates; the figures for college graduates when they are 15-20 years old do not refer to people who had graduated from college by then-this virtually never occurs-but rather to people of those ages who ultimately got college degrees.) We approximate what degrees individuals got by their reported years of education. We assumed that those with less then 12 did not complete high school, that those with 12 years were high school graduates with no college, that those with 13-15 year had completed some college but did not have college degrees, and that those with 16 or more years were college graduates.
We computed each respondent's marital status by age in months using his or her date of birth as well as the year and month of each marriage and divorce. All respondents who were interviewed in the 2008 wave are included, drawing from earlier waves' information on their marriages, even if they were not interviewed in some earlier wages.
Our graphs show simple descriptive statistics on the proportion of the sample (in the stated sex, race, age, and education group) that had ever been married for Figure 1 or was currently married for Figure 2. In Figures 1 and 2, the data are graphed using a loess smoother to reduce month-to-month variability. For a table of the numbers behind the graphs, contact Jonathan Bearak at email@example.com._________________________________________
1 Those who are widowed are also counted as "ever married" but not currently married, unless they have remarried. However, in these data that only look at the life cycle to age 45, very few people have been widowed.
2 See http://www.bls.gov/nls/.
** To download this information as a PDF, click here .
CCF is a volunteer organization whose busy professionals contribute their time and expertise gratis. Click here to support our mission by making a donation to CCF. Your contributions will help to fund our annual conferences and publicize our research briefings.