MARRIAGE AND DIVORCE RATES

Marriage and divorce rates are measures of the propensity for the population of a given area to become married or divorced during a given year. Some of the rates are quite simple, and others are progressively more refined. The simple ones are called crude rates and are expressed in terms of the number of marriages or divorces per 1,000 persons of all ages in the area at the middle of the year. These are the only marriage and divorce rates available for every state in the United States. They have the weakness of including in the base not only young children but also elderly persons, who are unlikely to marry or become divorced. But the wide fluctuations in crude rates over time are obviously associated with changes in the economic, political, and social climate.

More refined rates will be discussed below, but the following illustrative crude rates of marriage for the United States will demonstrate the readily identifiable consequences of recent historical turning points or periods (NCHS 1990a, 1990b, 1998).

Between 1940 and 1946 the crude marriage rate for the United States went up sharply, from 12.1 per 1,000 to 16.4 per 1,000, or by 36 percent, showing the effects of depressed economic conditions before, and disarmament after, World War II.

Between 1946 and 1956 the rate went down rapidly to 9.5 per 1,000, or by 42 percent, as the baby boom peaked. The unprecedented increase in the number of young children was included in the base of the rate, and this helped to lower the rate.

Between 1956 and 1964 the rate declined farther to 9.0 per 1,000, or by 5 percent, as the baby boom ended and the Vietnam War had begun. Between 1964 and 1972 the rate went up moderately to a peak of 10.9 per 1,000, or by 21 percent, as the Vietnam War ended and many returning war veterans married.

Between 1972 and 1990 the crude marriage rate declined irregularly to 9.8 per 1,000 persons of all ages and on down to 8.9 in 1997. The factors involved are discussed below.

At the time of this writing, vital statistics annual reports present only crude marriage and device rates for the United States and individual states (NCHS 1998). More refined rates were last published for 1987 on some subjects and for 1990 on other subjects.

Refined marriage rates show the propensity to marry for adults who are eligible to marry. They exclude from the base all persons who are too young to marry and may also limit the base to an age range within which most marriages occur. The conventional practice of basing these rates on the number of women rather than all adults has the advantage of making the level of the rates correspond approximately to the number of couples who are marrying. Moreover, the patterns of changes in rates over time are generally the same for men and women.

Changes over time in the tendency for adults to marry are more meaningful and may fluctuate more widely if they are reported in refined rather than crude rates. To illustrate, the crude marriage rate declined between 1972 and 1980 from 10.9 per 1,000 population to 10.6, or by only 3 percent, while the refined rate (marriages per 1,000 women 15 to 44 years old) declined from 141.3 to 102.6, or by 27 percent. A change that appeared to be small when measured crudely turned out to be large when based on a more relevant segment of the population. Persons who want to have others believe that the change was small may cite the crude rates, and persons who want to demonstrate that the marital situation was deteriorating rapidly may cite the refined rates. But persons interested in making a balanced presentation may choose to cite both types of results and explain the differences between them.

Still greater refinement can be achieved by computing marriage rates according to such key variables as age groups and previous marital status. Examples appear in Table 1 and Figures 1 and 3 for the United States from 1971 to 1990. Table 1 and Figure 1 show first marriage rates by age for the only marital status category of eligible persons, namely, never-married adults (women) 15 years old and over. Figure 2 shows divorce rates by age for married women, and Figure 3 shows remarriage rates by age for divorced women. The low remarriage rates for widows are not shown here but are treated briefly elsewhere in the article. Rates of separation because of marital discord are also not presented here: Separated adults are still legally married and are therefore included in the base of divorce rates.

The marriage and divorce rates in Table 1 were based on data from reports published by the National Center for Health Statistics (NCHS). These reports contain information, obtained from central offices, of vital statistics in the states that are in the Marriage Registration Area (MRA) and the Divorce Registration Area (DRA). In 1990 the District of Columbia and all but eight states were in the MRA, while the District of Columbia and only thirty-one states were in the DRA. Funding for the central offices is determined by each state's legislature. But for states not in the MRA or DRA, the NCHS requests the numbers of marriages and divorces from local offices where marriage and divorce certificates are issued. The reports on divorce include the small number of annulments and dissolutions of marriage. Bases for the marriage and divorce rates in Table 1 were obtained from special tabulations made by the U.S. Bureau of the Census from Current Population Survey data. These are tabulations of adults in MRA and DRA states and classified by marital status, age, and sex. Because not all the population of the United States is included in the MRA and DRA, the detailed marriage and divorce statistics published by the NCHS constitute approximations of the marital situation in the country as a whole. This article contains much numerical information that was published in one or more of the NCHS reports listed in References.

FIRST MARRIAGE RATES BY AGE

Illustrations of first marriage rates appear in table 1. For the United States, the first marriage rates per 1,000 never-married women 15 years old and over were 93 in 1971 and 58 in 1990. In effect, 9.3 percent of the never-married women in 1971 and 5.8 percent in 1990, became married for the first time. For men the corresponding rates were 68 in 1971 and 47 in 1990.

First marriage rates tend to decline with age, and the rates for most of the age groups shown in Table 1 were declining over time. The rates for the age groups under 20 years of age were among the highest, but they dropped so sharply that they were only about one-third as high in 1990 as they had been in 1971. At the oldest ages, the change appears to have been slight. Obviously, the propensity to marry was falling far more abruptly among the young than among the older singles, probably in reaction to the suddenly changing cultural climate.

The generally downward trend in the marriage rate for each young age group was especially rapid during the early 1970s. By 1975, the veterans of the Vietnam War had already entered delayed marriages, and the upsurge in cohabitation outside marriage was only beginning to depress the first marriage rate. During the 1980s the slight upturn in the first marriage rate for women over 30 years of age probably reflected an increase in marriages among women who had delayed marrying for the purposes of obtaining a higher education and becoming established in the workplace. Research has produced evidence that women who marry for the first time after they reach their thirties are more likely to have stable marriages than those who marry in their twenties (Norton and Moorman 1987). Although first marriage rates among adults in their forties are relatively low, they are by no means negligible.

As first marriage rates declined between the mid-1960s and the late 1980s, the median age at first marriage rose at an unprecedented pace over this short period of time. According to vital statistics, the median age at first marriage for women went up from 20.3 years in 1963 to 24.0 in 1990, for men it went up from 22.5 years to 25.9 years. As age at first marriage increased, the distribution of ages at first marriage also increased (Wilson and London 1987).

One of the consequences of the great delay of first marriage has been a very sharp rise in pre-marital pregnancy. Only 5 percent of births in 1960 occurred to unmarried mothers, but this increased to 24 percent in 1987 and to 32 percent in 1996. Research by Bumpass and McLanahan (1989) showed that one-half of nonmarital births during the late 1980s were first births, and about one-third occurred to teenagers. Moreover, about one-tenth of brides were pregnant at first marriage. Thus, about one-third of the first births during the late 1980s were conceived before marriage.

As the first marriage rate declined, the proportion of all marriages that were primary marriages (first for bride and groom) also declined. In 1970, two-thirds (68 percent) were primary marriages, but by 1987 the proportion was barely over one-half (54 percent). Meantime, marriages of divorced brides to divorced grooms nearly doubled, from 11 percent to 19 percent, while marriages of widows to widowers went down from 2 percent to 1 percent.

Men and women, regardless of previous marital status, tend to marry someone whose age is similar to their own. But men who enter first marriage when they are older than the average age of men at marriage have a reasonable likelihood of marrying a woman who has been divorced.

Procedures have been developed for projecting the proportion of adults of a certain age who are likely to enter first marriage sometime during their lives. This measure is, in effect, "a lifetime first marriage rate." One of these procedures was used by Schoen and colleagues (1985) to find that about 94 percent of men who were born in the years 1948 to 1950 and who survived to age 15 were expected to marry eventually; for women, it was 95 percent. Their projections for those born in 1980 were significantly lower, 89 percent for men and 91 percent for women. Despite the implied decline, a level of nine-tenths of the young adults deciding to marry at least once is still high by world standards.

DIVORCE RATES BY AGE

As indicated above, crude divorce rates are the only rates available annually for the United States and individual states. As recently as 1965, the rate was only 2.5 divorces per 1,000 persons in the United States, but by 1979 the rate reached a peak more than twice that high, 5.3. Then the rate declined gradually until 1997, when it was down to 4.3, the lowest rate since 1973.

The most recent refined divorce rates available include those by age and sex in Table 1 for 1990. The report for that year also presented other refined divorce statistics on the number of divorces occurring among persons in their first marriage, their second marriage, and their third or subsequent marriage (NCHS 1995b). But the required bases for computing first divorce rates and redivorce rates are not available. During the 1980s, about three-fourths of the divorces were obtained by adults in their first marriage, about one-fifth by those in their second marriage, and one-twentieth by those who had been married at least three times.

The divorce rates in Table 1 provide illustrations of the magnitude of the rates for the period 1970 to 1990. The divorce rate per 1,000 married women was 14 in 1970 and 19 in 1990. Corresponding rates for men were the same. The rate reached a peak of 20 in 1980, nearly half again as high as in 1970, and declined slightly to 19 in 1987, a level still well above that in 1970.

The divorce rate for married women rose dramatically in every age group during the 1970s and changed relatively little from then through the 1980s. In January 1973 the Vietnam War ended, and at that time the norms regarding the sanctity of marriage were being revised. The advantages of a permanent marriage were being weighed against the alternatives, including freedom from a seriously unsatisfactory marital bond and the prospect of experimenting with cohabitation outside marriage or living alone without any marital entanglements.

By 1994, one-fourth of divorced persons were cohabiting outside marriage, one-third were living alone, and most of the rest were in single-parent families (U.S. Bureau of the Census 1996).

Married women under 25 years of age have consistently high divorce rates resulting largely from adjustment difficulties associated with early first marriage. Noteworthy in this context is the finding that 30 percent of the women entering a first marriage in 1980 were in their teens, while 40 percent of those obtaining a divorce had entered their marriage while teenagers.

More divorces during the 1980s were occurring to married adults 25 to 34 years of age than to those in any other ten-year age group. A related study by Norton and Moorman (1987) concluded that women in their late thirties in 1985 were likely to have higher lifetime divorce rates (55 percent) than those either ten years older or ten years younger. This cohort was born during the vanguard of the baby boom and became the trend setter for higher divorce rates. The lower rate for those ten years younger may reflect their concern caused by the adjustment problems of their older divorced siblings or friends.

Married women over 45 years of age have quite low divorce rates. Most of their marriages must still be reasonably satisfactory, or not sufficiently unsatisfactory to persuade them to face the disadvantages that are often associated with becoming divorced. Yet, the forces that raised the divorce rate for younger women greatly after 1970 also made the small rate for the older women increase by one-fourth during the 1970s and remain at about the same level through the 1980s.

The median duration of marriage before divorce has been seven years for several decades. This finding is not proof of a seven-year itch. In fact, the median varies widely according to previous marital status from eight years for first marriages to six years for second marriages and four years for third and subsequent marriages. The number of divorces reaches a peak during the third year of marriage and declines during each succeeding year of marriage.

Among separating couples, the wife usually files the petition for divorce. However, between 1975 and 1987, the proportion of husband petitioners increased from 29.4 percent to 32.7 percent, and the small proportion of divorces in which both the husband and the wife were petitioners more than doubled, from 2.8 percent to 6.5 percent. These changes occurred while the feminist movement was becoming increasingly diffused and the birthrate was declining, with the consequence that only about one-half (52 percent) of the divorces in 1987 involved children under 18 years of age and one-fourth (29 percent) involved only one child. It is not surprising that nine-tenths of children under 18 living with a divorced parent live with their mother, far more often in families with smaller average incomes than those living with divorced fathers.

Although only about one-tenth of the adults in the United States in 1988 were divorced (7.4 percent) or separated (2.4 percent), the lifetime experience of married persons with these types of marital disruption is far greater. Based on adjustments for underreporting of divorce data from the Current Population Survey and for underrepresentation of divorce from vital statistics in the MRA, Martin and Bumpass (1989) have concluded that two-thirds of current marriages are likely to end in separation or divorce.

REMARRIAGE RATES BY AGE

Remarriage rates published by NCHS include separate rates for remarriages after divorce and after widowhood as well as for all remarriages. In this article attention is concentrated on remarriages after divorce, which constitute about nine-tenths of all remarriages. As mentioned above, the remarriage rate after divorce is a measure of the number of divorced women who marry in a given year per 1,000 divorced women at the middle of the year. For example, Table 1 shows that the remarriage rate after divorce for the United States was 133 in 1970 and 76 in 1990. Therefore, about 13 percent of the divorced women in 1970 became remarried in that year, as compared with 8 percent in 1990, only three-fifths as much as in 1970.

Divorced men have far higher remarriage rates than divorced women (in 1990, 106 versus 76). This situation and women's greater longevity largely account for the number of divorced women 45 years old and over in 1994 being one and one-half times the number of divorced men of that age (U.S. Bureau of the Census 1996).

Like first marriage rates, remarriage rates tend to decline with age, and the rate for each age group declined after 1970. The especially sharp drop during the first half of the 1970s for women under 40 years of age resulted from the compounding effect of a rapid increase in divorce and a rapid decline in remarriage during that period.

A large majority of divorced persons eventually remarry. In 1980, among persons 65 to 74 years old, 84 percent of the men and 77 percent of the women who had been divorced had remarried (U.S. Bureau of the Census 1989). But because of the declining remarriage rates, a projection based on information from the National Survey of Families and Households conducted in 1987 and 1988 shows that 72 percent of the recently separated persons are likely eventually to remarry (Bumpass, et al. 1989). About 6 percent of those who become separated never become divorced and therefore are not eligible for remarriage. Two-thirds of the remarriages in recent years occurred to women who entered first marriages as teenagers, according to the same study. Moreover, the rate of remarriage declines as the number of young children increases. Among married parents of young children with one or both parents remarried, about one-third of the children were born after the remarriage, and the others are stepchildren with the usually accompanying adjustment problems (Glick and Lin 1987).

In a given year about two of every three adults who marry are marrying for the first time, but this includes those marrying after widowhood as well as divorce. Among those who remarry after divorce, about three-fourths have been married only once, one-fifth have been married twice, and one-twentieth three or more times. Some couples who remarry had been married to each other previously. According to unpublished data from the National Survey of Families and Households, this occurred in about 3.3 percent of all marriages and closer to 5 percent for those in the age range when most remarriages after divorce occur.

Men are older than women, on the average, when they marry for the first time or when they remarry. Moreover, the gap is considerably wider at remarriage than at first marriage, but it narrowed somewhat between 1970 and 1987. Thus, successive marriages have been happening at older ages but with shorter intervals between them. In this context, the wider gap at remarriage than at first marriage may be less socially significant than an identical gap would have been at first marriage.

A woman in her second marriage is likely to be married to a man who is about ten years older than her first husband when she married him. Therefore, her second husband was probably more advanced in his occupation than her first husband was when she married him. But research has established that her first husband was probably about as far advanced ten years after her first marriage as her second husband was when she married him (Jacobs and Furstenberger 1986).

OTHER MARRIAGE AND DIVORCE RATES

Another rate that differs from the rates shown in Table 1 is the total marriage rate. This rate is intended to show the number of marriages that group of 1,000 men and women would have if they experienced in their lives the age-sex marriage rates observed in a given year (NCHS 1990b). It is therefore a hypothetical rate analogous to a total fertility rate. Both first marriage rates and remarriage rates have as the base the total population of the United States without regard to previous marital status.

The total first marriage rate for the United States in 1987 implied that only 69 percent of men and 70 percent of women would eventually marry. The corresponding remarriage rates were 45 percent for men and 41 percent for women. Because both rates are based on the population regardless of marital status, they are additive. Therefore, the (combined) total marriage rate for 1987 implied that men are likely to have 1.14 marriages during their lifetime and women 1.11. These results may seem low because of the assumptions involved. For instance, if currently about 90 percent of every 100 adults marry, if one-half of the first marriages end in divorce, and if 70 percent of the divorced persons remarry, this would mean that 100 young adults in the 1990s are likely to have 90 first marriages and 32 second marriages after divorce (.90 + .90 x .50 x .70). In addition, many will redivorce and remarry again, and others will become widowed and remarry. Thus, realistically, the average young adult in the 1990s who marries is likely to have more than 1.1 marriages.

Remarriage among widowed adults is not featured in this article. However, the remarriage rate per 1,000 widows declined from 7 to 5 in 1990, while the rate for widowers declined from 32 to 24. The remarriage rate for widowers is much higher than the rate for widows because about one-half of them marry widows, who outnumber them five to one (U.S. Bureau of the Census 1989). Of all men and women who married in 1987, 68 percent had never previously married, 29 percent were divorced, and only 3 percent were widowed.

Divorce rates per 1,000 involving children under 18 published by NCHS shows that in 1987, the rate was 17, implying that 1.7 percent of the children under 18 years of age in the United States were involved in parental divorces in that year. This finding implies further that, if the same rate continued for eighteen years, 30.6 percent of the children would likely experience parental divorce before they reached 18 years of age. Data from the U.S. Bureau of Census used by Norton and Glick (1986) put the estimate at 40 percent. They also estimated that about an additional 20 percent of children become members of one-parent families because of premarital birth, parental separation that does not end in divorce, or death of a parent. Therefore, about 60 percent of the children born in the 1990's and later may expect to spend a significant amount of time in a one-parent family before they become 18 years of age.

Marriage rates vary among countries as a reflection of dissimilar social, demographic, and economic conditions. An analysis of marriage rates in fifteen developed countries revealed that between 1965 and 1980 the rate per 1,000 unmarried women declined in all but two of the countries (Glick 1989). The marriage rates in English-speaking countries and Israel were above the average for the entire group, but the rates were below the average in the Germanic, French, and Scandinavian countries. A special reason for the differences was the extent to which cohabitation had been accepted as at least a temporary alternative to marriage. The generally downward trend among the marriage rates shows that the changing social conditions related to the propensity toward marriage have become widely diffused.

Divorce rates per 1,000 married women in almost all of the fifteen countries went up between 1965 and 1980 and doubled in the majority of them. Most of the countries with divorce rates above the overall average were English speaking. High divorce rates tended to be associated with high marriage rates because remarriages after divorce make an important contribution to the level of the marriage rates per 1,000 unmarried women.

Remarriage rates have been falling in most of the fifteen countries. In Canada a part of the decline between 1965 and 1980 was attributed to a change in the divorce laws and an increase in the delay of remarriage. The remarriage rate in New Zealand actually rose slightly in the context of a baby boom and an increase in immigration of young adults with subsequent high divorce rates. Countries with the highest remarriage rates were English speaking, and those with the lowest were Scandinavian countries, France and certain other European countries where cohabitation outside marriage had risen sharply.

Variation in the level of marriage and divorce rates among the American regions and states can be documented only by the use of crude rates. The Northeast and Midwest had consistently lower marriage rates than the South and West. The general pattern is similar for divorce rates. Some of the variations in the rates by states result from differences in the strictness of residence requirements for obtaining a marriage or divorce license. About seven of every eight marriages occur in the state where both the bride and the groom have their usual residence.

SELECTED VARIABLES RELATED TO MARRIAGE AND DIVORCE

Marriage and divorce rates are not presented in the NCHS reports by education and race of those involved. However, some reports do show distributions of first marriages and remarriages by several categories of education and race. The 1987 report documents that those marrying for the first time had more education, on the average, than those who were remarrying. Information from the 1980 census showed that women 25 to 34 years of age who had exactly four years of college training had distinctly the largest proportion of intact first marriages; those with graduate school training had a somewhat smaller proportion; and those with an incomplete college education had a smaller proportion than those in any other education category (U.S. Bureau of Census 1985).

The 1990 report on marriage provided evidence that white adults tend to marry at a younger age than black adults. The difference was two years for both brides and grooms at first marriage and two years at remarriage after divorce. Information for 1988 from the U.S. Bureau of the Census (1989) indicated that 18 percent of black adults had not married by the time they were 40 years old, as compared to 8 percent for white adults of the same age. The pressure to marry and to remain married evidently tends to be less for black adults than for white adults.

Interracial marriage occurs between a small but socially significant proportion of those who marry. In 1987, 2 percent of black brides married white grooms, and 6 percent of black grooms married white brides. Also, nearly 2 percent of black women obtaining divorces were married to white men, and more than 4 percent of black men obtaining divorces were married to white women. Thus, among those who intermarried, the marriages of black men to white women tend to be more stable than marriages of white men to black women.

Cohabitation outside marriage increased dramatically from only one-half million heterosexual couples in 1970 to 3.5 million in 1993, with about one million of their households maintained by the woman. This numerical growth occurred primarily among adults below middle age and has contributed importantly to the decline in marriage rates as well as to the increase in the number of separated persons. According to the 1987–1988 National Survey of Families and Households, "almost half of the persons in their early thirties and half of the recently married have cohabited" (Bumpass and Sweet 1989, p. 615). In order to provide some balance on the issue, Thornton has concluded that "even though cohabitation will be experienced by many, most people will continue to spend substantially more time in marital unions than in cohabiting unions" (1985, p. 497).

The health of adults is related to marital selection and marital stability, but NCHS does not provide marriage and divorce rates by the health status of those involved. The center does, however, publish current information on several indicators of the health condition of adults by marital status. Nearly all the indicators confirm that people with more signs of good health are likely to marry and remain married or to remarry after marital dissolution (Wilson and Schoenborn 1989). Parental divorce tends to be related to health problems of children but largely through the custodial parents' loss of income and time to spend with the children after divorce (Mauldon 1988). And stress prior to an event such as divorce or premarital breakup may actually reduce the impact of the life transition (Wheaton 1990).

Current trends in marriage and divorce rates do not necessarily indicate whether the rates will tend to stabilize at or near their 1990 levels, to resume their movement in historical directions, or to continue fluctuating in response to future social developments. A few more decades of observing the impact of past changes in marriage and divorce rates on the persons involved may be necessary before a definitive evaluation can be made concerning the longtime effect of these changes on family and child welfare.

(see also: Divorce; Marriage; Remarriage)

references

Bumpass, Larry, and Sara McLanahan 1989 "Unmarried Motherhood: Recent Trends, Composition, and Black-White Differences." Demography 26:279–286.

Bumpass, Larry, and James A. Sweet 1989 "National Estimates of Cohabitation." Demography 26:615–625.

Bumpass, Larry, James A. Sweet, and Teresa Castro Martin 1989 "Changing Patterns of Remarriage." Working Paper 89–02. Madison: Center for Demography and Ecology, University of Wisconsin.

Glick, Paul C. 1989 "The Family of Today and Tomorrow." In K. Ishwaran, ed., Family and Marriage: Cross-Cultural Perspectives. Toronto: Wall and Thompson.

Glick, Paul C., and Sung-Ling Lin 1987 "Remarriage after Divorce: Recent Changes and Demographic Variations." Sociological Perspectives 30:162–179.

Jacobs, Jerry A., and Frank F. Furstenberg, Jr. 1986 "Changing Place: Conjugal Careers and Women's Marital Stability." Social Forces 63:714–732.

Lin, Sung-Ling 1987 "Marital Selection and Child-Bearing and Companionship Functions of Marriage and Remarriage." Ph.D. dissertation, Arizona State University, Tempe.

Martin, Teresa Castro, and Larry L. Bumpass 1989 "Recent Trends in Marital Disruption." Demography 26:37–51.

Mauldon, Jane 1988 "The Effect of Marital Disruption on Children's Health." Paper presented at the Population Association of America meeting, New Orleans. National Center for Health Statistics (NCHS) 1990a "Births, Marriages, Divorces, and Deaths for 1989." Monthly Vital Statistics Report 38, no. 12.

——1990b "Advance Report of Final Marriage Statistics, 1987." Monthly Vital Statistics Report 38, no. 12, Supplement.

——1990c "Advance Report of Final Divorce Statistics, 1987." Monthly Vital Statistics Report 38, no. 12, Supplement 2.

——1995a "Advance Report on Final Divorce Statistics, 1989 and 1990." Monthly Vital Statistics Report 43, No. 9, Supplement.

——1995b "Advance Report on Final Marriage Statistics, 1989 and 1990." Monthly Vital Statistics Report 43, No. 12, Supplement.

——1998 "Births, Marriages, Divorces, and Deaths for 1997." Monthly Vital Statistics Report 46, No. 12.

Norton, Arthur J., and Paul C. Glick 1988 "One-Parent Families: A Social and Economic Profile." Family Relations 35:9–17.

Norton, Arthur J., and Jeanne E. Moorman 1987 "Current Trends in Marriage and Divorce among American Women." Journal of Marriage and the Family 49:3–14.

Schoen, Robert, William Urton, Karen Woodrow, and John Baj 1985 "Marriage and Divorce in Twentieth-Century American Cohorts." Demography 22:101–114.

Thornton, Arland 1988 "Cohabitation and Marriage in the 1980s." Demography 25:492–508.

U.S. Bureau of the Census 1985 1980 Census of Population: Marital Characteristics. Washington: U.S. Government Printing Office.

——1996 "Marital Status and Living Arrangements: March 1994" Current Population Reports P-20, no. 484 Washington: U.S. Government Printing Office.

Wheaton, Blair 1990 "Life Transitions, Role Histories, and Mental Health." American Sociological Review 55:209–223.

Wilson, Barbara F., and Kathryn A. London 1987 "Going to the Chapel." American Demographics 9:26–31.

——, and Charlotte Schoenborn 1989 "A Healthy Marriage." American Demographics 11:40–43.

Paul C. Glick