50th anniversary; and 53 of every 100 marriages that last until the 40th anniversary survive to the 50th.

Given current divorce and death rates, we can estimate the chances of attaining other anniversaries. For example, 63 percent of new marriages survive to the 10th anniversary, 41 percent to the 25th, and 25 percent to the 40th. Among marriages that reach the 10th anniversary, 65 percent also achieve the 25th, and 40 percent the 40th. Of couples celebrating their 25th anniversary, 61 percent will be together long enough to attain their 40th.

The probability of divorce decreases with the duration of marriage. At today's divorce rates, 39 percent of marriages reaching the 5th anniversary and 27 percent of those reaching the 10th will eventually end in divorce. But only 11 percent of those marriages that attain the 20th, and 7 percent of those achieving the 25th anniversary, will subsequently end in divorce.

The risk of divorce declines rapidly in the early years of marriage. Compare the figures in the previous paragraph: Between the 5th and 10th anniversaries the probability of a future divorce declines by 12 percentage points (from 39 to 27), but between the 20th and 25th anniversaries the probability declines by only 4 points (from 11 to 7).

Figuring the Odds

Another way to portray the recent American divorce pattern is to calculate the percentage of marriages that will end in divorce before achieving a given anniversary. At current divorce rates, 19 percent of a given cohort of new marriages will end in divorce before the 5th anniversary,

33 percent before the 10th, 40 percent before the 15th. 47 percent before the 25th, and 50 percent before the 50th. (These percentages are somewhat higher than those presented at the beginning of this article for the actual marriage cohorts of 1952, 1957, 1962, and 1967 because divorce rates have since risen.)

Some marriages end with the death of a spouse and, of course, are no longer subject to the risk of divorce. As a result, the cumulative percentage of couples divorced at any given anniversary is not as large as it would be if some marriages were not disrupted by death. If we adjust the figures so that death disrupts no marriages at all for a fixed period after marriage, we can determine statistically the potential percent divorced by a given anniversary.

If we assume for example, that only divorce can disrupt marriages before the 30th anniversary (the risk of disruption by death is zero), 49 percent of marriages would end in divorce by the 25th anniversary. using the 1976-77 rates. This is only two points higher than the 47 percent figure derived from the duration-of-marriage tables, in which both divorce and death dissolved marriages. This comparison demonstrates how minimal the effect of mortality is during the early decades of the average marriage Expectation of Marriage

In a life table the effects of death rates at specific ages are summarized in a statistic known as the "expectation of life at birth." For example, the current expectation of life at birth is about 74 for Americans. Applying the lite table approach to marriage duration vields a similar measure, the

"expected duration of a marriage just initiated." which reports the number of years an average marriage can be expected to last after the wedding.

Today, a new marriage will last an average of 23.2 years, based on the divorce and death rates of 1976-77. For first marriages, the expected duration is two or three years more than this, and for remarriages some five to ten years less.

The expected Juration of new

marriages has declined. Between 1948 and 1965 the expected duration of marriage was 30 to 32 years. By 1969 the figure had dropped to 26.2 years, and the 1976-77 figure is a full three years below that. Since the early 1960s, the expected duration of new marriages has declined about eight years, or almost 25 percent.

This decline reflects the increase in divorces, especially considering the moderate decline in mortality

22-192 0-83-10

age 23.0 years remain, and at the 9th anniversary the figure is 24.5. Even at the 15th anniversary the expected duration of marriage is 23.4 yearsslightly more than the expected duration among newlyweds. For a marriage that reaches the 15th anniversary, the first 15 years cost nothing in expected duration.

This measure illustrates clearly that the risk of divorce is high in the early years of marriage and declines rapidly thereafter. Of course, as a marriage "ages" it faces an increased risk of disruption due to the death of a spouse, and the expected duration eventually begins to decline as a result. At the 25th anniversary the expected number of years of marriage remaining is 19.1. Those marriages which attain the 40th anniversary last on the average an additional 11.3 years. And if your marriage lasts 50 years, the odds are it will last another 7. Divorce Trends

Underlying all these probability calculations is the fundamental assumption that the current divorce and death rates will continue unchanged for several decades.

Twenty years ago, the general divorce rate was about the same as in the years just before World War II. About eight or nine divorces were granted each year per thousand married women aged 15 years or older. In 1967, the rate was only slightly higher-11.2 divorces granted annually per thousand married women. But the 1979 rate of 22.8 was more than double the 1967 rate.

The annual divorce rate now appears to be leveling off. Between 1976 and 1977, the rate remained unchanged at 22.1 divorces per

thousand married women. After increasing in 1978 and 1979, the rate appears to be stable for 1980, according to provisional figures.

Although the divorce rate may be leveling off, or increasing at a much slower pace than a few years ago, there is little evidence suggesting when the rate will begin to decline, or even that such a decline will occur. Since recent divorce rates (197677) imply that half of all marriages will end in divorce, some argue that society cannot long sustain such a high level of marital disruption. Of course, 30 years ago few would have predicted that society would be able to tolerate even one-third of married couples divorcing, but that level has already been reached by some marriage cohorts.

The large number of divorces are themselves a force that could keep the divorce rate high. With marriages so often ending in divorce, many previously married people are joining the pool of available marriage partners. Since the large majority of divorced people remarry, and remarriages have a higher overall risk of divorce, the general divorce rate is likely to rise as a result.

Pushing in the other direction is the tendency of young adults to postpone their first marriage. In 1970, only 11 percent of the women and 19 percent of the men between the ages of 25 and 29 had never been married. In 1981, 22 percent of the women and 34 percent of the men in that age group had never been married. Because women who marry in their twenties are much less likely to divorce than those who marry in their teens, the trend toward later marriage could reduce divorce rates.

No one can predict with certainty

future divorce rates because they are influenced by so many different factors. However, divorce has become an important ingredient in the marital composition of the American population. For example, in 1970 only 5 percent of the women aged 25 to 54 who had ever been married were currently divorced, and only 3 percent were separated. In 1981, 12 percent were divorced and another 5 percent separated.

Divorce seems increasingly accepted today, as divorces have become more common. That fact could push the divorce rate higher still. With more women entering the work force, new strains in family life may lead to divorce, and economic hard times could create further stress. While the future of divorce rates remains uncertain, divorce-like marriage has become an American

institution.