How happy couples got happier 💍

The slow revolution in marriage since the 1970s.

Here’s a fun fact that I’ve been forcing on passersby lately:

In England and Wales today, most 38-year-old men and 35-year-old women have never married.

Let me quickly explain how we can be almost certain of that. According to ONS data on opposite-sex marriages in 2023, the median first-time groom was 32.7, and the median first-time bride was 31.2. Those are fairly well known numbers, as numbers go. They’re also quite intuitive, because we can all call to mind the sort of age that people getting married tend to be.

But these numbers tell us nothing about the people who aren’t marrying. And as it happens, people these days are not marrying a lot.

We can measure this with some of the less famous numbers at the back of the ONS collection, which show the marriage data by year of birth, covering both straight and gay marriages this time. With this data we can track the people born in 1987, say, who would all now be at least 38, and count how many of them had married at least once with each passing year. By 2023, the men had got to 43.6%.

We don’t yet have data for 2024 and 2025, but the additional percentage who marry each year gets smaller quite predictably at this point, making it very unlikely that most of today’s 38-year-old men will have married already. And the same is true of women who were born in 1990 and are now 35. (The 39-year-old men and 36-year-old women, I think, will have just scraped over the 50% line.)

Lots of unmarried people will be in relationships, of course. But if you’re an unmarried 38-year-old man, or an unmarried 35-year-old woman, and you’re starting to feel like a straggler, you’re really not. It’s the married people your age who are slightly unusual in fact.

Marriage grew up so fast

The gradualness of the change in our marrying habits hides from people how profound it’s been, I think. To make my point, here are some more fun facts that the people close to me have recently enjoyed being told:

😳 The overall rate at which people married in 2023 was about a quarter of what it was in 1971.

👶 From 1971 to 2023, the number of teenage brides fell from 107,982 to 856, and the number of teenage grooms from 34,702 to 347. (These include a few dozen same-sex teenage marriages.)

🏠 In 1994, about 60% of marrying couples lived together before their wedding day. In 2023 it was 89%.

😱 In 1972, 70% of weddings were religious ceremonies. In 2023, it was 15%.

🤯 As recently as 1985, most grooms were under 25 and most brides were under 23.

Marriage used to be a sanctified leap into a whole new life at a very young age. In half a lifetime it’s become the rubber-stamping of a middle-aged relationship.

Change for the better?

If you love the romance of weddings, this sensible revolution may be sad news. If you have a horror of divorce, it’s not. Because these same years also transformed the way that marriages ended.

The data on divorce over time paints a messy picture. Legal changes making divorces easier probably harvested some overdue ones, and maybe accelerated others that were on the cards. So we shouldn’t read the changing divorce rate as a close barometer on the happiness of marriages.

But you know, they’re not completely unconnected, so let’s run with it a bit. Can we use the divorce data to get at least a vague sense of how successful each year’s marriages have been? In particular, do the numbers suggest that the generations who waited longer to get married were also less likely to get divorced?

This introduces another layer of mess, because the median period between marriage and divorce has changed over time, from a low of 8.9 years in 1985, to an all-time high of 12.7 years most recently. But just to keep things casual, and ignoring all the better but more complicated ways to do this, let’s say there is a divorce-delay of 10 years. This doesn’t actually show us the ultimate fate of each year’s marriages, but it ties the two trends together into a vague story that makes a kind of sense.

If it’s not clear at first how to look at these charts, start with the dot in the bottom left. Its position shows the median age at which people married in 1960 and, on the vertical axis, what the divorce rate turned out to be 10 years later, in 1970. So it tells us that people married very young back then, and that very few couples were getting divorced 10 years later, which would have been a fairly common time for them to do so.

Knowing what we do about changing attitudes and laws, you’d guess that the rarity of divorce in 1970 said more about the obstacles to ending marriages than the harmony inside them. It would also explain why, when those obstacles fell away, the divorce rate rocketed as the 1970s wore on.

Although interestingly, many of the marriages that became those divorces were still formed very young. Signs of people waiting longer to marry first show up in the 1980s. And for a while, there’s not much change in the subsequent divorce rate.

But from the mid-1990s onwards, by which point couples were generally in their late twenties, the subsequent divorce rate starts to fall quite quickly. Couples just get older and older, and divorce gets rarer and rarer, until it is now back to the level of the early 1970s.

Then progress stutters in the 2010s. I’m probably reading too much into these charts already, but I wonder whether this indicates that the advantages of delaying marriage are starting to run out. Some people are always going to make mistakes in choosing who they marry, and some will just be unlucky and face unforeseen problems afterwards, so you’d never expect the divorce rate to close in on zero.

And while it seems reasonable to guess that people who choose their spouse at 30 are more likely to find the right one than people who jump into marriage at 25, the extra benefits of waiting until 35 probably aren’t quite so great.

Does marriage still mean anything?

It’s now completely normal for people to settle down without marrying—indeed births out of wedlock are more common than in it—so perhaps I’m missing the point here. Perhaps a lot of couples who would once have married and had kids in their twenties are still having kids and not marrying, but otherwise living very similar lives? And perhaps they have similar rates of separation later?

Perhaps. Although my guess is that won’t be the whole story. For one thing, people are having children older too, so there is a general drift towards taking big decisions later in life. For another, I think it’s worth asking why people used to get married when they were so young.

Reports that reach me from those who remember the 1960s and the 1970s suggest an obvious answer: that young people who wanted to live together back then didn’t have much choice If you cared about society’s good opinion, you were all but forced to marry your first really serious boyfriend or girlfriend. It must have been intoxicatingly romantic. And since everyone else was doing it, it probably didn’t seem completely mad.

In the end, I find it hard to escape the feeling that this is why marriage used to be four times more popular than it is today. Because back then it was a magic wand that transformed people’s lives in an instant. Now, by comparison, it promises very little that people don’t already have. As result, marriage has become something they get round to. If they get round to it at all.