The evidence sat in two large federal surveys taken twenty years apart, and lined up side by side it undercuts one of the more comforting things people assume about a long marriage — that a union which makes it through the hard early decades has, in effect, been made safe.
Between 1990 and 2010, the divorce rate among Americans aged fifty and older doubled, from about five divorces a year for every thousand married people to just over ten. Over the same two decades the divorce rate for the country as a whole barely moved. By 2010, roughly one in four people getting divorced was at least fifty — a group that, twenty years earlier, had made up fewer than one in ten.
There is a limit to what a number like this can say. It comes from national surveys, not from inside any home, and reading the research is a different thing from counseling the people living it. What the figure can do is correct a picture most of us carry — and the correction is not the one the phrase “gray divorce” tends to summon.
The safest marriages are the long first ones
When the sociologists Susan Brown and I-Fen Lin looked at who was actually divorcing after fifty, the pattern ran against the stereotype. The couples least likely to split were those in long first marriages. Among people who had been married forty years or more, the rate was about three divorces a year per thousand, among the lowest figures anywhere in the study. The longer a marriage had lasted, the lower the risk fell.
What pushed the numbers up was something else: remarriages, and marriages that were younger than the graying heads at the table made them look. People in a second or later marriage divorced at about two and a half times the rate of those in first marriages. And more than half of the people over fifty who divorced in a single year had been in a remarriage — even though most of those who stayed married were still in their first.
So the picture the phrase conjures, a thirty- or forty-year first marriage coming undone, does happen. It is simply the rarer case. The larger story the data tells is one of second marriages, and shorter marriages, coming apart. The researchers’ best explanation is not a change in anyone’s heart but a change in who is now old: the generation that divorced and remarried in large numbers in the 1970s and 1980s has aged into its fifties and sixties, bringing more remarriages with it, and remarriages have always been the more fragile arrangement. The share of married people over fifty who were on a second or later marriage rose from about a fifth in 1980 to nearly a third by 2010.
The paper does gesture at why this happens, though carefully, and mostly by pointing to explanations others have proposed. Middle and later life brings its own turning points — the empty nest, retirement, a diagnosis — and each can prompt a couple to take the measure of a marriage that had been organized for years around raising children or leaving for work. One that worked well when both spouses were busy can founder once the house is quiet and the calendar is suddenly open. Longer lives stretch the exposure, too: a marriage that no longer fits now has more decades left in which to end. These are suppositions, the authors are quick to note, not things this study measured.
One thread in that literature is specifically about women. As more women have built their own earnings and pensions across a working life, the reasoning goes, more have the means to leave a marriage that has stopped working rather than stay for lack of an alternative. The independence that makes leaving possible does not make it painless, though: the studies that have followed later-life divorces find the financial fallout tends to land hardest on women. Both halves of that come from the surrounding research, not from the head count this paper actually ran.
The reasons the survey never recorded
The study never asks why. It counts divorces and lines them up against age, income, race, education and marital history, but the reasons stay off the page. It can tell you the rate doubled; it cannot tell you what happened inside any of those homes. The familiar explanations — the late affair, the reinvention, the slow drift apart — may each be true in a given case, but none of them is in this data, and the authors are careful to say the pattern is correlational, not causal. Something like income can shift because of a divorce as easily as before it.
There is a generational question underneath as well. The study compares two age groups at a single moment, so it cannot fully separate the effect of simply getting older from the effect of belonging to a particular cohort. And it reaches back only to 1990, not for lack of interest but because age-specific divorce data are not available for earlier decades, leaving these two snapshots as the best available bookends.
It is worth resisting the pull of the word “doubled,” too. A rate that doubles from a low starting point is still, for any one couple, a small yearly chance. More than six hundred thousand people over fifty divorced in a single year — a figure large enough to sound like an epidemic, and, spread across every married couple in the country, small enough to miss the house next door entirely.
What the record does carry is a correction the field itself had missed. Researchers who study later life had long treated the end of an old marriage as a story about widowhood; this work argues that divorce now belongs beside it, with consequences of its own. Those consequences fall unevenly — thinner finances afterward, especially for women, and intergenerational ties that can strain when a newly single parent leans harder on adult children. The reach extends past the couple, in other words, to children and grandchildren who find the shape of the family rearranged around them late in the day. Those downstream findings come from other studies and are still being worked out, so we hold them loosely. And if the marriage in question is your own, or your parents’, that is exactly where a family therapist or counselor belongs — sitting with the particulars a national survey was never built to see.
So the next fortieth-anniversary party is, by the numbers, close to what it appears to be — one of the marriages most likely to last. The couples the record was really tracking are elsewhere in the room: the second marriages, the newer ones, the pairs who found each other later in life and are quietly learning how much harder a remarriage can be to keep.
