October 7, 2010

The Reversal of the College Marriage Gap

III. Marriage Delayed or Marriage Foregone?

III. Marriage Delayed or Marriage Foregone?

Whether Americans are deemed to be retrenching from marriage depends partly on the age group examined, how far back the comparison is made, the education group and some judgment as to what constitutes a large decline in marriage. Setting aside the timing of marriage momentarily, as of 2008 more than 90% of persons in their late 50s (early baby boomers) had married. On the one hand, this continues the decline in marriage from the peak levels observed in 1990 when 95% of Americans in their late 50s had married. On the other hand, it remains the case that more than nine-in-ten adults marry sometime in their life. And, as has been true for much of the country’s history, Americans marry at higher levels than do people in most other Western countries (Cherlin, 2009).

The parents of the baby boomers, or those born between 1920 and 1940, married young and at an exceptionally high level. About 95% of those born between 1920 and 1940 had married by ages 55 to 59. So, at 91%, it is true that the early baby boomers (2008’s 55- to 59-year-olds) have not married at the rate their parents did. But the early baby boomers are marrying at nearly the rates their grandparents did. In 1950, 92% of 55- to 59-year-olds had married. So whether there has been a retreat from marriage partly depends on how far back we look.

As Figure 7 shows, adults lacking a college degree married at rates in excess of 90% throughout the past 60 years, and there has been a decline in marriage among the less educated if the comparison is made to the high marriage rates among the boomers’ parents.

It is not clear that the college-educated are foregoing marriage entirely to a greater extent than earlier generations did. In 2008, about 91% of college-educated 55- to 59-year-olds (early baby boomers) had married. This is slightly lower than the marrying propensity of the college-educated among the parents of the baby-boomers, but it is much higher than among the college-educated grandparents of the baby boomers. In 1950 only 80% of college-educated 55- to 59-year-olds had ever married.

It is less disputable that Americans are delaying marriage. Commencing with the baby boom generation and continuing to more recent birth cohorts, Americans are less and less likely to have married by age 30 (Figure 8). In 1970, nearly nine-in-ten 30-year-olds had married. By 2008, that was down to only six-in-ten. The decline in marriage among the college-educated has been well-documented (for example, Isen and Stevenson (2008). Less appreciated is the fact that young adults lacking a college degree have been delaying marriage to an even greater extent than college-educated young adults. In 2008, 43% of college-educated 25- to 29-year-olds had married, a 33 percentage point decline from the 76% of 1970 college-educated 25- to 29-year-olds who had married. Among the less educated, 46% of 25- to 29-year-olds had married in 2008, a 40 percentage point decline from the 1970 level (86%).

Young adults lacking a college degree are delaying marriage to such an extent that long-standing differences in marital timing between the more and less educated no longer exist. A common benchmark of marital timing is the age at which half the population has married. As Figure 9 shows, less-educated adults have markedly delayed marriage since 1990 so that, by 2008, there is no difference in median age at first marriage between adults who are college-educated and those without college degrees. In 2008, half of college-educated 28-year-olds have married, the same age at which exactly half their counterparts lacking a college diploma have married.6 As recently as 2000, less-educated young adults crossed the 50% marriage threshold two years earlier than college-educated young adults (age 26 versus age 28). While entry into adulthood may be increasingly complex, varied and gradual (Settersen, Furstenberg, and Rumbaut, 2005), the age of a traditional marker of adulthood-marriage-displays growing uniformity among the more and less educated.

In sum, over the past several decades there has been a shift in the basic relationship between education and entry into marriage. The college-educated are just as likely as the less educated to marry. Furthermore, the differences in timing of first marriage are disappearing. At age 25, those lacking a college diploma are more likely to have married than their college-educated peers (Figure 1), but by age 28 this difference vanishes.

Fifteen years ago, sociologists examined lifetime marriage patterns and asked if the college-educated were foregoing marriage or simply delaying it. Today, the same question arises but in regard to the opposite group: young adults lacking a college degree. In 2008, only 46% of 25- to 29-year-olds who were not college-educated had married. How many of these lesser-educated young adults will ever marry over their lifetimes? We do not yet know.

Earlier cohorts of young adults without a college education tended to marry in their 20s, and there was not extensive additional marriage beyond age 30. For example, using Census Bureau data, we can track the marital behavior of late baby boomers. In 1990, about 65% of 25- to 29-year-olds who were not college-educated had married (Table 1). Looking at the same cohort nearly two decades later, in 2008, 85% of 45- to 49-year-olds who were not college-educated had married.7 If the lives of today’s 25- to 29-year-olds who are not college-educated followed a similar trajectory, then perhaps only two-thirds of them will ever marry. But the most recent cohort has already demonstrated markedly different marriage behavior than previous cohorts, so the past may not be a good guide to future marriage behavior.

The apparent delay in entry into marriage among less-educated adults may be due to a variety of factors, including the increasing prevalence of cohabitation before marriage. In 2009, there were 6.7 million unmarried opposite-sex partner households, up from 2.9 million in 1996.8 Based on 2003 data, unmarried partners tend to be much younger than married partners (U.S. Census Bureau, 2004). About four-in-five married persons were 35 or older, versus only roughly half of unmarried partners. Married adults tend to be better educated than cohabiting adults. About a third of married adults were college-educated, while only about one-fifth of cohabiting persons had a college education. And married men tend to earn more than cohabiting men. About 31% of husbands earned more than $50,000 in 2002, compared with 14% among cohabiting men. These figures are only suggestive, because the married population includes a much larger share of older persons with more labor market experience, but cohabitation is more prevalent among less-educated adults.

The delay in marriage among young adults without a college degree has also occurred during a period in which less-educated young adults have fared relatively poorly in the labor market. From 1990 to 2008, the inflation-adjusted median annual earnings of college-educated men ages 25 to 34 modestly rose from $52,700 in 1990 to $55,000 in 2008. Among men ages 25 to 34 with only a high school diploma, median annual earnings declined from $36,300 in 1990 to $32,000 in 2008 (National Center for Education Statistics, 2010). Young people are taking longer to transition from school to work, and those who are less educated are less successful at the transition than are those who are more educated (Blanchflower and Freeman, 2000). However, while most observers agree that there has been substantial divergence in labor market opportunities across education groups and that the deteriorating economic position of young men could be a plausible explanation for delayed marriage, empirical studies reach different conclusions on how important changes in men’s economic characteristics have been on marital behavior (Hill and Holzer, 2006; Oppenheimer, Kalmijn, and Lim, 1997).

  1. Looking at the lower panel of Figure 1, the crossover point occurs at age 28 and at the 50% ever-married level.
  2. This slightly overstates the subsequent marriage of less-educated 25- to 29-year-olds in 1990. In 2008, this cohort is 43 to 47 years old, not 45 to 49 years old. In 2008, 83% of less-educated 43- to 47-year-olds had married.
  3. Census Bureau historical table UC-1 on Unmarried-Couple Households, by Presence of Children: 1960 to Present, available at http://www.census.gov/population/www/socdemo/hh-fam.html.