The dramatic erosion of marriage in the UK is one of the key social changes of the late 20th and early 21st centuries.
Last week, the Office for National Statistics (ONS) published data showing that for the first time the proportion of the population aged over 16 who were married had fallen to below 50 per cent. Just two generations ago, in 1960, the figure was around 90 per cent.
The idea that all family structures are equally valid is one that has become an article of faith amongst the liberal left. It has been a key concept underpinning the collapse of marriage.
But in terms of the outcomes for children, for example, the evidence shows that this is simply not true.
In most families, most children grow up to be reasonably well-adjusted and integrated into the wider economy and society. This is the case regardless of the particular structure.
But this is just one of many examples of situations in which although most outcomes are benign, different types of behaviour do lead on average to different outcomes.
For example, driving with a safety belt does not guarantee protection in the event of an accident and in some cases may even be a disadvantage because it may trap the driver inside the car. But, on average, it is far safer to drive with a safety belt than without. Most car journeys made when the driver is over the legal alcohol limit end without accidents. But a larger proportion do not when the driver has been drinking than when he or she has not.
In the same way, averaging across large numbers of individual examples, the evidence on family structure is clear.
Children brought up in single parent families, for example, on average end up with lower educational qualifications and are more likely to get caught up in crime. This is the case even when taking into account factors such as different levels of income and wealth.
Marriage remains a valuable institution, for the individuals concerned, for their children and for society. The evidence is overwhelmingly in support of all three of these propositions. Indeed, there are few hypotheses in the social sciences which receive such confirmation from serious research.
The idea that economics can say useful things about social issues such as family structure and crime was the innovation of the Chicago economist Gary Becker. He received the Nobel Prize for his work in 1992.
Becker argued that through the institution of marriage, individuals realise the gains of what economists call comparative advantage. Each partner specialises in the tasks at which he or she is more efficient with a net benefit to both. Of course, this argument could be applied to co-habitation rather than formal marriage as such.
A different and more powerful argument is put forward by economists such as Bob Rowthorn, former head of department at Cambridge. Marriage is seen as an institution for providing couples with the confidence to make long-term investments in their relationship. It is a mechanism for building trust and helping to cope with uncertainty and the adverse shocks which families can receive.
Trends in family life are not merely personal indulgences. Private decisions have public consequences. Decisions made by individuals impose costs on the rest of us.
The disturbing aspect of the release of the new ONS data on marriage was not so much the data itself but the lack of reaction – it drew very few comments about the huge social costs created.
Many will argue that the trend is too firmly established for it to be reversed. But it is high time for governments to begin the process of trying to reverse it. Always remember that the longest march begins with a single step.