Mr Sandro Mangion (The Sunday Times, April 3) devoted much space to rubbishing the finding of the UK National Office of Statistics that the chances of children up to age five seeing their parents separate are eight per cent in married couples and 52 per cent in cohabiting couples - apparently for no other reason than that he refuses to believe it can be true.

If he had bothered to look up (on the Internet, say) even a fraction of the abundant research done in the last 15 years comparing married and cohabiting couples, he would soon have realised that the UK statistic is fully in line with the scientific evidence from all over the world.

Mr Mangion seems to think there is little difference between married and cohabiting couples except that the latter do not bother to get married. The research, however, suggests that marriage and cohabitation are actually quite dissimilar, that people enter into them with different aims and expectations, and that they are experienced differently by both adults and children.

This can be vividly illustrated by some statistics taken mainly from Marriage-Lite: The Rise of Cohabitation and Its Consequences (2000) by the sociologist Patricia Morgan. This study draws on reputable surveys, governmental or academic, from 19 countries. It has its limitations, but as regards the statistics, the only criticisms I have come across are trivial, for example that one of them is 19 years old.

The most striking fact about cohabitation is that it tends to be short-lived. Research shows that it is four or five times more unstable than marriage, and up to ten times more unstable where very young children are involved. A UK survey found that fewer than four per cent of cohabitations last ten years or more.

It is sometimes claimed that couples should cohabit before marriage to see if they will get on well together. This is not a good idea, for cohabitation before marriage almost doubles the risk of divorce. There is one significant exception. Most studies have found that those who cohabit with a firm intention to marry are no more at risk of break-up than those who marry without prior cohabitation.

After having a child, cohabitants are three times more likely to break up and their chance of marriage is reduced by 60 per cent. Cohabiting men tend to feel that the arrival of a child threatens their independence. Couples who marry after having a child are more likely to divorce than couples who have a child within marriage. Sixty per cent of divorced fathers provide income to the child's mother, but only 18 per cent of former cohabiting fathers do so.

Both cohabiting men and women are far more unfaithful than their married counterparts. Quarrels are more frequent, and violence is both more frequent and more severe. Cohabiting pregnant women are three to four times more likely to be physically abused. (So much for Mr Mangion's claim that "marriage is the site of most domestic violence".)

Levels of sexual satisfaction are significantly lower, whereas rates of stress, depression and alcoholism are three times higher. Far from freeing one from the difficulties of marriage, cohabitation appears to exacerbate them.

One of the factors contributing to this is that in cohabitation, mate selection is geared towards finding a "partner" rather than a "spouse". Since the stakes are perceived as less high, the screening mechanisms employed are less rigorous. As a result, the level of compatibility tends to be rather lower in cohabiting couples than in married couples, leading to a greater potential for conflict and stress.

At work, cohabiting men are less involved and less successful. They earn 10 to 40 per cent less than married men. When men marry they tend to work harder, but men who cohabit usually still act as single persons even when they have children.

Mr Mangion says, "The notion that marriage is the safest form of relationship for children is an illusion." He could hardly be more wrong. Cohabiting women are four times more likely than married women to have an abortion. Babies born to cohabiting couples are at greater risk of ill-health, and in fact have a death rate 35 per cent higher than that of babies born to married couples. Neglect and abuse of children are 20 times more likely with cohabiting biological parents, and 33 times more likely if the mother is cohabiting with a man who is not the biological father.

Mr Mangion also claims that "good parenting does not depend on a couple being married". Maybe not, but a married couple are rather more likely to provide good parenting than a cohabiting couple. Children thrive in a home environment that is stable, healthy, harmonious and lasting. Cohabiting parents, however, are rather less likely to provide such an environment.

A large-scale Australian study matching married and cohabiting couples for age, education, socio-economic status, personal attributes and relationship length found that children of cohabiting parents were significantly less likely to do well at school both academically and socially.

Without commitment and the long-term view that goes with it, cohabitants tend not to risk specialising their roles as married couples do. Since their splitting and duplication of roles is less efficient, their engagement with their children tends to be more discontinuous and of poorer quality. Cohabiting parents are typically less involved in their children's education, have less contact with school or teachers, and give less support to the children in matters such as homework, sports or educational and, cultural activities.

Delinquency is more frequent among children of cohabiting couples, and they are more likely to take drugs. They tend to start drinking earlier in life and to drink more. They are 50 per cent more likely to have mental problems. When they grow up, they are more likely to cohabit, and if they marry it is more likely that the marriage will fail.

Marriage and cohabitation, in short, are quite different things. Basically, people who marry do so because they want to make a commitment to each other, while people who cohabit do so because they want to keep their independence. This means that it is in the nature of cohabiting relationships that they lack the stability and permanence of marriage. They are far more fragile and, regardless of age or income, far more likely to break up.

The reasons for this difference appear to be evolutionary. (Evolution, as Pope John Paul II said in 1996, is now "more than a hypothesis".) It has long been known by anthropologists that marriage is a human universal. In none of the tens of thousands of human societies studied has it been found to be absent. To Darwinians, this suggests that marriage is a biological adaptation favoured by natural selection because it improves an individual's chances of getting his or her genes into subsequent generations.

Natural selection (or as the popular phrase has it, "survival of the fittest") simply means the better adapted organisms, since they have a better chance of survival and reproductive success, increase in frequency in a population with time. Strong selection pressure can lead to runaway effects as diverse as the giraffe's neck, the peacock's tail or the human brain (so much larger than that of any other primate). Marriage, with its overwhelming and amply documented superiority to cohabitation, bears all the hallmarks of being another example of this.

Mr Mangion may not like it, but he would do well to ponder a favourite saying of the Darwinian philosopher Daniel C. Dennett: "Evolution is cleverer than you are."