I have always been fascinated by Henry Kissinger's accent. That sort of low Germanic rumble. I ponder it at the oddest of moments. So I suppose that it wasn't all that surprising yesterday, listening to Gordon Brown talk about social mobility, that I found myself thinking: “All very well, Prime Minister, but what about Henry Kissinger's accent?”

It popped into my head when, in the middle of his remarks, Gordon Brown quite casually said this about delinquent children: “We know, of course, that so much of this [antisocial behaviour] starts in the home.” I love that “of course”.

Lightly tossed off this sentence might have been, but that doesn't mean it wasn't important. The Prime Minister spoke for 40 minutes about social mobility, setting himself goals, declaring himself committed, reeling off statistics and announcing initiatives, but he never once addressed the fundamental questions - what causes children to behave as they do? What drives social mobility?

Mr Brown didn't address these questions because “of course” he knows the answers. So instead he made statements such as this - “social mobility usually starts with parents wanting children to do better than they did themselves” - without providing any evidence that this testable proposition was, in fact, true.

And even if it is true, does the parental desire cause the social mobility? Or does desire to improve and ability to achieve social mobility just happen to exist in the same families? Or perhaps most parents want their children to do better, and this is a general characteristic of families, but only some children turn out to be successful at social mobility. Who knows?

Mr Brown's confidence that he knows where behaviour comes from, however, is not as great as that of the Conservatives. The Prime Minister thinks that social mobility starts in the home, the Tories fear that social breakdown does.

A couple of weeks ago Chris Grayling, the Shadow Work and Pensions Secretary, wrote an essay in The Sunday Times entitled “Tackling the bad parent society”. “In many places,” he proclaimed, while providing only anecdotes as evidence, “parents no longer know how to bring up children.” He went on to argue that putting this right was essential to overcoming social disorder. Unfortunately, he didn't say what the correct way to bring up children was. Perhaps he is planning a second article.

There is a phrase to describe this blithe belief that behaviour is taught in the home. It is known as “the nurture assumption”, a term coined by Judith Rich Harris for use as the title for a ground-breaking 1998 book.

The debate about the origins of behaviour used to be known as the “genes versus environment” debate. But, with our addiction to catchy titles, it is now commonly given a different title - the nature versus nuture debate.

Harris makes a simple point. In making the switch, we have made an assumption - the nurture assumption. We assume that the only important environmental factor, the only one that really matters in the debate, is nurture - the way in which you are brought up by your parents. This assumption is pervasive and, Harris argues, wrong.

Which brings us to Henry Kissinger's accent. It is not especially odd that a child who came to America at age 15, should retain the accent of his childhood. What is a little strange is that his younger brother, Walter, does not. Strange though this may seem, it isn't uncommon. The children of immigrants usually speak the language of their birthplace without an accent, even if their parents speak with a heavy accent.

Why should this be? After all, children are taught to speak by their parents, and do much of their talking in a family environment. Why wouldn't all family members sound alike? The reason, Harris argues, is that we aren't really raised by our parents - we are raised by our peer group.

The biggest effort we expend as children is fitting in with our peers and finding a place for ourselves among them. It is while rubbing up against our contemporaries that our personalities are formed. There is stunningly little evidence (actually Harris finds none) to suggest that differences in parenting makes any difference to the way that children turn out.

Indeed, while there is a vast literature on parenting skills, and some of it may help with relationships inside families, there is no evidence that any of it works to change the way that children behave outside the home. Gross abuse aside, teaching parents to change their behaviour (to what? no one knows) to change children is liable to be a gigantic wild goose chase.

One reason why this is strongly counterintuitive is that intuition doesn't take enough account of genetics. Parents often appear to make an impact on their children's behaviour because their children share some of their genes. To give one example, the children of divorced people tend to divorce more often than the average. The reason for this could be that children of divorced people have been set a bad example, or it could be that, being the biological offspring of their parents, they behave in similar ways. Studies of twins raised apart suggest the biological over the environmental explanation.

Recognising the existence of the nurture assumption does not mean living with social immobility or social breakdown. It just means being more clear-eyed about how to overcome the problems. Recently, for instance, Professor Rona Campbell, of Bristol University, presented the results of work that she had done successfully to discourage teen smoking by using peer group pressure.

Parents may matter less than we think, but peer groups matter more. Let's hope that we've heard the last speech on social mobility that rests on an unexplained “of course”.

daniel.finkelstein@thetimes.co.uk

Times Recommends

• You created the job. You should do it
• A swindler with a ready maid plan
• Charming to pillory disabled Brown