You don't really know the divorce rate for a particular group of people until they're all dead. Or widowed. Most of what is offered as "divorce rates" are single-year per capita totals, pure "projections" predicting from recent trends, or the ratio of weddings to divorces in a single year and jurisdiction. As the UK's Marriage Foundation notes,"This 'year of divorce' method of calculating divorce rates makes analysis of trends all but impossible to assess because it mixes marriages of different durations." But once a cohort gets past the point where their chance of divorce becomes infinitesimal anyway, you can get a handle on it.
Of course, that assumes no great increase in the numbers of people divorcing after their silver anniversary. But it's still about the best method out there, and was undertaken recently in a study by the Marriage Foundation. And in fact the gray-divorce factor mentioned above as a caveat apparently does not apply in the UK: the researchers say divorce rates after the 10th anniversary have always been stable, with all the ups and downs in rates coming in the first 10 years of marriage. The study does seem to incorporate some element of lifetime projection and prediction, especially for newer marriages, but they do have those cohorts' records up until now to look at, as well as the stability in divorce rates for long marriages.
There are a few cohort-based studies in the U.S., which are some of the best explorations of divorce rates out there, but they are slightly less reliably because of middle-aged and elderly divorce, and because the collection and analysis of divorce statistics in the U.S. is relatively primitive, barely funded, and does not even cover several states, including California. Most studies tracking individuals and couples are done through the "long-form census" sent to a small but representative sample of the population, and even that has been threatened by budget cuts and by disdain from both the left and the libertarian right for government studies and records of marriage and family life.
This article about the study in The Daily Mail blames the 1969 no-fault laws for the increase, and seems to credit premarital cohabitation and rising marriage ages for the later improvements:
"Anyone who wed in 1986 has the highest chance in modern times of ending up divorced. In fact, researchers say almost half of those who married then will eventually split up. ... It was the first of six bleak years for marriage, during which the easy availability of divorce and the erosion of traditional family values meant that more couples parted than ever before or since. After 1991 the figure waned, until by 2008 – the year that produced the most stable marriages in recent times – the proportion of couples who will see their union end in divorce is put at well under the four out of ten mark.
"The study by the Marriage Foundation think tank gives a year-by-year breakdown of a couple’s chance of divorce, from the liberal 1960s and the explosion in numbers of marriage break-ups that followed.
Divorce law reforms in 1969 removed the idea that a husband or wife had to have been at fault from many cases, and introduced the ‘quickie’ divorce for those who admitted adultery or other faults.
Since the 1990s marriages have become more stable, with younger couples often choosing to live together before marriage, and those who do marry waiting until they are older, and then enjoying more long-lived unions.
Harry Benson of the Marriage Foundation said the average age for brides and grooms at their first wedding was 30 and 32 in 2011 – up from 23 and 25 in the early 1980s.
‘The entire rise and fall in divorce rates since the 1960s has taken place within the first decade of married life,’ he said.
‘The worst two years to get married were 1986 and 1991. Some 44 per cent of couples who married in these years will end up divorced.’
The figures, based on Office for National Statistics marriage and divorce returns, revealed that in 1963 fewer than a third of weddings were due to lead to divorce – 28.2 per cent. But by 1977, following the 1969 reforms, the predicted risk of divorce for newly-marrieds had topped 40 per cent.
The 1986 peak was matched with 44.4 per cent predicted divorce levels for those marrying in 1988 and 1991.
Then, gradually, divorce began to tail off and in 2008 bottomed out at 38.3 per cent. The lifetime likelihood of divorce for those married in 2012, the latest year for which figures are available, was 38.4 per cent."
Why 1986 was the worst year to get married: Nearly half of unions have ended in failure
- Those who wed in 1986 have highest chance in modern times of divorce
- Study by a think tank found 1991 was the second worst year to get married
- Since 1990, marriages have been more stable with couples moving in first