Skip over navigation

Journal Issue: Children and Divorce Volume 4 Number 1 Spring/Summer 1994

Epidemiology of Divorce
Patricia H. Shiono Linda Sandham Quinn

National Trends in Marriage, Divorce, and Remarriage

Information about marriages and divorces in the United States is collected on the Standard License and Certificate of Marriage and Standard Certificate of Divorce, Dissolution of Marriage, or Annulment.5 The information provided by couples on these certificates allows us to monitor the occurrences of marriages and divorces. Long-term trends in marriage, divorce, and remarriage are presented in Figure 1 and represent a cross-sectional or a snapshot look at these events at a given time.6 Because the actual number of marriages and divorces in a given year is dependent on the size of the population, marriages and divorces are measured as rates making them independent of the size of the population. The rates plotted in Figure 1 depict the number of women experiencing one of these events per year, divided by the number of women at risk for the event. Thus, marriage rates are defined as the number of women marrying in a given year divided by the number of single women, and divorce rates are the number of women divorcing in a given year divided by the number of married women. Three-year running averages are plotted to smooth out individual year fluctuations and to provide a clearer picture of long-term trends.


To be at risk for divorce, couples must first be married. Rates of first marriages have fluctuated dramatically over the past 70 years. In the 1920s, annual marriage rates were approximately 99 per 1,000 single women. Rates declined during the depression era (early 1930s) to 81 per 1,000 women. Marriage rates then rose to an all-time high of 143 per 1,000 women in the post-World War II era (mid-1940s) and declined steadily for the next 30 years. Rates declined sharply in the mid- 1970s to 85 per 1,000 women. In recent years, the rate of first marriage continues to decline and reached an all-time low of 76 per 1,000 women in 1988. Thus, during the marriage peak in the mid-1940s, approximately 14% of never-married women wed each year, compared with the current low of approximately 7%.

The declining propensity of couples to marry also is reflected in the decrease in the proportion of women who marry to avoid having a child out of wedlock. During the 1960s, approximately half (52%) of all women were pregnant when they married, whereas in the 1980s, only one quarter (27%) of women were pregnant when they married.6 Although current marriage rates are relatively low, these low rates are somewhat offset by increasing rates of cohabitation (unmarried couples living in the same household), which have risen from approximately 11% in the late 1960s to 44% in the 1980s.7 Cohabitation largely delays marriage rather than replacing it, as 60% of cohabiting couples eventually marry.7

In recent years, there has been a growing difference in marriage rates by age and by ethnic group.6 Fewer and fewer women are marrying in their twenties. Among women 20 to 24 years of age in 1975, 63% had married, while in 1990, only 39% of women in the same age group had married. For women 25 to 29 years of age, these percentages dropped from 87% in 1975 to 69% in 1990; rates dropped from 93% to 82% for women 30 to 34, 96% to 89% for women 35 to 39, and from 96% to 92% for women 40 to 44.

African-American women are becoming much less likely to marry than white women. In 1975, 94% of white women and 87% of African-American women in their early thirties had ever married. These rates dropped to 86% for white and to 61% for African-American women in 1990. Moreover, in 1990, only 75% of African-American women in their late thirties had ever married compared with 91% of white women. If this trend continues, less than 75% of African-American women will marry at some point in their lives compared with 90% of white women. This decrease in marriage has led to increased out-of-wedlock births for both African-American and white women, but especially for African-American women (see discussion below). Data about Hispanic women has been available only since 1980, and it appears that their patterns of marriage are similar to those of white women.


Since the 1860s, divorce rates have generally been increasing, rising from approximately 10 per 1,000 married women per year and reaching a peak of 24 per 1,000 after World War II. Divorce rates decreased to approximately 15 per 1,000 in the 1950s and more than doubled in the 1970s to an all-time high of 40 per 1,000 married women in the late 1970s. In recent years, the annual divorce rates appear to be decreasing slightly; in 1988, the rate was 37 per 1,000 married women. Each year since the late 1970s, more than one million couples became divorced. Currently, most people who divorce do so early in their marriage so that half of the divorces occur by the seventh year of marriage.8 One of the most important effects of the rise in divorce, especially in the early years of the marriage, has been the increase in the proportion of divorces occurring among women with young children.7

Several important factors are associated with divorce: the age when couples marry, their educational attainment, and the timing of pregnancy.6 An inverse relationship exists between age at marriage and the likelihood of divorce. Teenage marriages are much more likely to end in divorce than are all other marriages. Women who marry when they are over age 30 are the least likely to become divorced. While there is no clear relationship between a woman's level of education and divorce, women who appear to have stopped short of obtaining a high school diploma or other advanced degrees have increased rates of divorce, compared with women who attain exact diploma or degree levels. Thus, women who completed high school or a bachelor's degree were less likely to divorce than women who stopped short of attaining these degrees. Obviously, it is not clear from these statistics if the divorce caused the women to cease their education or if dropping out of school caused the divorce. Women who were pregnant or had a child prior to their first marriage were more likely to divorce than women who become pregnant after marriage.

Demographers point to several societal events that have had major impacts on divorce rates. Divorce rates increased after every major war, decreased during the Great Depression, and decreased during the post-World War II economic boom. The large increase in divorce rates in the 1970s was bolstered by the introduction of no-fault divorce laws, the reduction in fertility as a result of improved methods of contraception, and the legalization of abortion.7 However, most scholars believe that the single most important social change which made divorce possible was the increase in the employment of women and the corresponding economic independence that employment provided.7 The rise in employment has been especially pronounced for married women. In 1940, 15% of women who were currently married (with husbands present) were working outside the home or looking for work. This proportion rose to 50% by 1979; in 1989, it was 58%. The change was greatest for married women with children. Since the middle of the 1960s, the rate of increase in employment for women with children has been the highest for women with preschool children.7 The rate of employment for married women with preschool children went from 11% in 1949 to 19% in 1959; in 1979, it rose to 43%; and by 1988, it had risen to 58%. Some scholars speculate that, if the economic status of workers in the United States remains at its relatively depressed level, then it is less likely that married women will leave the labor force in large numbers in the near future.7


Remarriage is dependent upon the rates of marriage, divorce, and widowhood. Remarriage rates have fluctuated dramatically. The levels of remarriage rates have generally been more extreme, having higher highs and lower lows than first-marriage rates, but have generally tracked the rise and fall of first-marriage rates until the 1960s. As marriage rates were declining slightly in the early 1960s and as divorce rates rose sharply after the mid- 1960s, remarriage rates soared to an alltime high in the late 1960s of 166 per 1,000 divorced and widowed women per year. In the 1970s, remarriage rates continued to track the decline in first-marriage rates. The length of time divorced individuals remain unmarried is also increasing. In 1970, the median time between divorce and remarriage was about one year; in 1988, this interval increased to about 2.5 years.9

Rates of remarriage are affected by age, income, education, ethnic group, and the presence of children in the family.9 However, these predictors often differ by sex and ethnic group. Younger women are more likely than older women to remarry, but age is not associated with remarriage among men. Men with higher incomes and education are more likely to remarry than men with lower incomes or education. However, increased educational attainment increases the likelihood of remarriage among African-American women but not among white women. The presence of children lowers the likelihood that a woman will remarry but does not affect the probability of remarriage among men. Divorced African-American women are less likely to remarry than divorced white women, and they have a longer interval between marriage. Only 34% of African-American women whose marriage ended between 1965 and 1984 remarried by 1988, compared with 60% of white women.9 During the first several years of marriage, the rate of divorce for remarriages is substantially higher than for first marriages; afterward, the rates are similar.7 By one estimate, 37% of remarriages among women end in a separation or a divorce within 10 years, compared with 30% of first marriages. The expanded families of remarriage after divorce may complicate the lives of remarried adults and their children.7 For children, there may be a substantial financial gain; however, adjustment of children in their first stepfamilies appears to be worse compared with children in single-parent families.3

Generational Differences in Marriage, Divorce, and Remarriage

Changes in marriages and divorces are affected by a complex array of several interwoven external factors, including changes in social norms over time, death rates, economic conditions, historical events, and laws. In an attempt to account for these factors, researchers often observe a group of people born during a specific time period and track their lifetime experience with marriage and divorce. When several of these different groups are compared, it is sometimes possible to see the potential effects of these external forces.

The marriage, divorce, remarriage, and redivorce rates for four groups of women born 20 years apart are shown in Figure 2.7 These groups represent four generations: women born in the early century, their daughters born during the depression, their granddaughters born during the start of the baby boom, and their great-granddaughters born in the 1970s. Because women who were born in the 1970s are currently in their twenties, lifetime rates for this cohort have been estimated using the 1990 rates of marriage, divorce, and remarriage. Despite revolutionary changes in social norms, economic conditions, wars, and fertility in the past 80 years, nearly all the women in these cohorts have married; 93% of the first generation, 97% of the second, 95% of the third, and 89% of the fourth generation have or eventually will marry during their lifetime. However, the proportion of women who have married and divorced has increased sharply for each successive generation: lifetime experiences of divorce doubled from 22% in the first generation to 44% in the fourth generation. Remarriage after divorce and second divorces have also increased sharply over the years. Virtually none of the women who were born in the early part of the century experienced a second divorce, whereas 16% of the youngest cohort are expected to experience a second divorce.

Life-table or actuarial rates of divorce have also been calculated for women born in each decade from 1890 to 1950.10 Life-table analyses follow a group of women who were born in a given year and track their movements over their lifetime to and from several marital states such as never married, presently married, widowed, divorced, and deceased. Unlike simple annual rates of divorce, which represent a snapshot view of the population in a given year, life-table rates of divorce provide an estimate of the likelihood of divorce over one's lifetime.

Each line on Figure 3 represents the cumulative probability of divorce by age for women who are at risk of divorce (women who have survived to age 15 and were married by ages 20 to 55).10 For nearly all decades, the lifetime probability of divorce for women has been increasing, and this increase has occurred for women of all ages. For women born in 1920, the likelihood of divorce by age 55 was 27%; this same level of divorce was reached at a much younger age (age 30) for women born in 1950. At least 40% of young adult women today are likely to divorce, 30% are likely to remarry following a divorce, 16% are likely to divorce twice if current divorce rates continue.7