Key Finding: Marriage trends in recent decades indicate that Americans have become less likely to marry, and the most recent data show that the marriage rate in the United States continues to decline. Of those who do marry, there has been a moderate drop since the 1970’s in the percentage of couples who consider their marriages to be “very happy,” but in the past decade this trend has flattened out.
a We have used the number of new marriages per 1,000 unmarried women age 15 and older, rather than the Crude Marriage Rate of marriages per 1,000 population to help avoid the problem of compositional changes in the population; that is, changes which stem merely from there being more or less people in the marriageable ages. Even this more refined measure is somewhat susceptible to compositional changes.
b Per 1,000 unmarried women age 14 and older.
Source: U.S. Department of the Census, Statistical Abstract of the United States, 2001, Page 87, Table 117; and Statistical Abstract of the United States, 1986, Page 79, Table 124; and Births, Marriages, Divorces, and Deaths: Provisional Data for 2007, National Vital Statistics Report 56:21, July 14, 2008, Table 2 (www.cdc.gov/nchs/data/nvsr/nvsr55/nvsr56_21.pdf ) and Current Population Surveys March 2007 data. The CPS, March Supplement, is based on a sample of the U.S. population, rather than an actual count such as those available from the decennial census. See sampling and weighting notes at www.bls.census.gov:80/cps/ads/2002/ssampwgt.htm .
a Includes races other than Black and White.
b In 2003, the U.S. Census Bureau expanded its racial categories to permit respondents to identify themselves as belonging to more than one race. This means that racial data computations beginning in 2004 may not be strictly comparable to those of prior years.
Source: U.S. Bureau of the Census, Current Population Reports, Series P20-506; America’s Families and Living Arrangements ( www.census.gov/population/socdemo/hh-fam/cps2008/tabA1-all.pdf).
Source: U.S. Bureau of the Census, Statistical Abstract of the United States, span, Page 34, Table 27; Statistical Abstract of the United States, 1971, Page 32, Table 38; Statistical Abstract of the United States, 1981, Page 38, Table 49; and U.S. Bureau of the Census, General Population Characteristics, 1990, Page 45, Table 34; and Statistical Abstract of the United States, 2001, Page 48, Table 51; Internet tables ( www.census.gov/population/socdemo/hh-fam/cps2008/tabA1-all.pdf ). Figures for 2008 was obtained using data from the Current Population Surveys rather than data from the census. The CPS, March Supplement, is based on a sample of the U.S. population, rather than an actual count such as those available from the decennial census. See sampling and weighting notes at http://www.bls.census.gov:80/cps/ads/2002/ssampwgt.htm.
Source: The General Social Survey, conducted by the National Opinion Research Center of the University of Chicago. The number of respondents for each sex for each period is about 2,000 except for 1977-81, 1998-2002, and 2004-08 with about 1,500 respondents for each sex.
Americans have become less likely to marry. This is reflected in a decline of about 50 percent, from 1970 to 2007, in the annual number of marriages per 1,000 unmarried adult women (Figure 1). Much of this decline—it is not clear just how much—results from the delaying of first marriages until older ages: the median age at first marriage went from 20 for females and 23 for males in 1960 to about 26 and 28, respectively, in 2007. Other factors accounting for the decline are the growth of unmarried cohabitation and a small decrease in the tendency of divorced persons to remarry.
The decline also reflects some increase in lifelong singlehood, though the actual amount cannot be known until current young and middle-aged adults pass through the life course.
The percentage of adults in the population who are currently married has also diminished. Since 1960, the decline of those married among all persons age 15 and older has been about 15 percentage points—and approximately 30 points among black females (Figure 2). It should be noted that these data include both people who have never married and those who have married and then divorced.
In order partially to control for a decline in married adults simply due to delayed first marriages, we have looked at changes in the percentage of persons age 35 through 44 who were married (Figure 3). Since 1960, there has been a drop of 22 percentage points for married men and 21 points for married women. (But the decline has not affected all segments of the population. See the accompanying box: The Marriage Gap.)
Marriage trends in the age range of 35 to 44 are suggestive of lifelong singlehood. In times past and still today, virtually all persons who were going to marry during their lifetimes had married by age 45. More than 90 percent of women have married eventually in every generation for which records exist, going back to the mid-1800s. By 1960, 94 percent of women then alive had been married at least once by age 45—probably an historical high point. For the generation of 1995, assuming a continuation of then current marriage rates, several demographers projected that 88 percent of women and 82 percent of men would ever marry. If and when these figures are recalculated for the early years of the 21st century, the percentage of women and men ever marrying will almost certainly be lower.
It is important to note that the decline in marriage does not mean that people are giving up on living together with a sexual partner. On the contrary, with the incidence of unmarried cohabitation increasing rapidly, marriage is giving ground to unwed unions. Most people now live together before they marry for the first time. An even higher percentage of those divorced who subsequently remarry live together first. And a growing number of persons, both young and old, are living together with no plans for eventual marriage.
There is a common belief that, although a smaller percentage of Americans are now marrying than was the case a few decades ago, those who marry have marriages of higher quality. It seems reasonable that if divorce removes poor marriages from the pool of married couples and cohabitation “trial marriages” deter some bad marriages from forming, the remaining marriages on average should be happier. The best available evidence on the topic, however, does not support these assumptions. Since 1973, the General Social Survey periodically has asked representative samples of married Americans to rate their marriages as either “very happy,” “pretty happy,” or “not too happy.” As Figure 4 indicates, the percentage of both men and women saying “very happy” has declined moderately over the past 35 years. This trend, however, has essentially flattened out over the last decade.
There is good news and bad news on the marriage front. For the college-educated segment of our population, the institution of marriage appears to have gained strength in recent years. For everyone else, however, marriage continues to weaken. Thus there is a growing “marriage gap” in America, between those who are well educated and those who are not.
Recent data indicates that, for the college educated, the institution of marriage may actually have strengthened. It once was the case that college-educated women married at a lower rate than their less educated peers. Indeed, marriage rates for college-educated women were lower well into the late twentieth century. Since around 1980, however, this situation has reversed. College-educated women are now marrying at a higher rate than their peers.[a] Not only that, but the divorce rate among these women is relatively low and has been dropping. This may be due partly to the fact that college-educated women, once the leaders of the divorce revolution, now hold a more restrictive view of divorce than less well educated women.[b] The out-of-wedlock childbearing of college-educated women has always been well below that of other segments of the population. Now, among those women who delay marriage past age 30, college-educated women are the only ones becoming more likely to have children after marriage rather than before.[c]
There is more good news. The marriages of the college educated have become better matched than ever, in the sense that husbands and wives are matched more equally in their educational and economic backgrounds.[d] As icing on the cake, all of this may add up to greater marital happiness. The percentage of spouses among this group who rate their marriage as “very happy” has held fairly steady over recent decades, whereas for other parts of the population the percentage has dropped significantly.[e]
In large numbers, therefore, the college-educated part of America is living the American dream—with happy, stable, two-parent families. There is one problem, however, and it is a serious one for the future of the nation. College-educated women aren’t having enough children to replace themselves. In 2004, for example, 24 percent of women 40 to 44 years old with a bachelor’s degree were childless, compared to only 15 percent of those without a high school degree.[f]
For the non college-educated population, unfortunately, the marriage situation remains gloomy. Marriage rates are continuing to decline, and the percentage of out-of-wedlock births is rising. According to one recent study, more than 50 percent of new mothers without college degrees are having their children outside of marriage, compared to just 7 percent of college-grad new mothers.[g] Because of the many statistically well-documented benefits of marriage in such areas as income, health, and longevity, this gap is generating a society of greater inequality. America is becoming a nation divided not only by education and income, but also by marital status.