`I do' ... and this time they seem to mean it

In the US, the forecast of divorces has been cut and the proportion of people living alone has stabilised

Is the institution of marriage at last about to make its comeback? Just this week has come some American evidence that it might.

In social attitudes, the pendulum always swings. But for more than a generation, the evidence has been that it is still moving against marriage: both divorce rates and the proportion of children born out of wedlock seem to carry on climbing inexorably. Thus in the UK, some 40 per cent of marriages taking place are expected to end in divorce: already marriages contracted in 1975 and 1981 are past the 30 per cent failure rate. And births out of marriage are above 30 per cent, compared with less than 5 per cent in 1960.

Much the same pattern occurs in most other developed countries. In the United States, the divorce rate was estimated at 50 per cent. And while we are towards the top of the international league for out-of-wedlock births, we are behind France and the Nordic countries. In both Iceland and Sweden, more than half the births are to unmarried mothers.

If this pattern were to reverse itself, you would expect the first signs of change to come in the US, partly because so many social trends originate there and partly because of the power of the religious right. Now, a report from the US Census Bureau, Household and Family Characteristics, suggests that while there may not be a turn-down, at least there seems to be a plateau.

For the first time in nearly two decades, the proportion of babies born to unmarried mothers has fallen: a 4 per cent decline since 1994. The projection of divorces has been cut, from 50 per cent to 40 per cent. The proportion of people living alone, which also rose rapidly through the Seventies and Eighties, seems to have stabilised. Family size, too, seems to have stopped shrinking, and the total fertility rate (number of live births per mother) is about back to the replacement rate of 2.1.

So here are a whole set of family characteristics that seem no longer to be moving in one direction, and may be starting to swing back. But this is the US. Is there any evidence of the same thing happening elsewhere?

In Britain, it seems that marriages contracted in 1985 may be surviving a little better than those in 1975 or 1981 - the divorce rate is running just a touch lower. There may be some tendency towards slightly larger families, for while the total fertility rate remains below replacement rate, it no longer seems to be falling. Looking at the UK figures, you can perhaps discern the first signs of a plateau, but there is no strong evidence of change. In fact, the very latest UK statistics, out yesterday, in the General Household Survey, show that the number of couples choosing to cohabit rather than get married, is still rising. They also show that fewer couples are having children than in the Eighties, though this may just be a trend towards having children at an older age. Elsewhere, in continental Europe and Japan, there is even less evidence of a social change - so we really have to come back to the US figures and assume that what happens there will eventually happen here.

If that assumption does prove to be right, there are the most profound implications for the entire developed world. These fall into two groups; those related to family structure and those related to family size. So much has been written about the benefits that accrue if families do stay together - and so much about the dangers of an illiberal climate of opinion, bashing single parents struggling to bring up children under difficult circumstances - that it seems pointless to add to the pile here. What is worth repeating is that if there were to be a spontaneous decline in both divorce and single parenthood, many of our other social and economic difficulties would become more manageable. If it were a decline led by politicians - and you can catch a bit of that from both major parties, witness the change in single-parent allowances in the Budget and Tony Blair's personal views on the need to rebuild family structures - then I'm not so sure. But if we (and other Western societies) rebuild the institution of marriage because we, as individuals, want to, then it is a win, win situation.

A change in family size - a trend back up towards replacement rate, maybe past it - would also have important social and economic effects. In the short-term, it would mean more dependants, more children to feed and educate for the same number of workers. But in the longer-term, it would mean that the apparently inexorable ageing of the population would slow, and in another 30 or 40 years' time there would be more workers to support the very large army of elderly people. The two trends are probably related. If marriage comes back in vogue, I suspect larger families will, too.

But even if our societies have reached some kind of turning point, are we really heading back to the pattern of the Fifties, with families with one earner, almost invariably the chap, and the sort of social pressures to conform which was blown to bits in the Sixties? The answer to that one is surely not. Much more likely is a period a bit like the early Victorian age, when there was a slow underlying drift towards a more orderly society, but taking place in a much more fluid and experimental climate than the 1950s - if, that is, the turning point really has been reached.

We will not know about that for at least 10 years. Meanwhile, two thing are sure. One: it is utterly impossible to recognise a turning point until it is long past. Two: any institution, like marriage, which has been around for several thousand years, is not going to disappear in one generation.