Underclass is an ugly word, and we live in an age that abhors ugly words, so it is good to hear that the Blair government has devised a cheerier label: Neet, an acronym for “not in education, employment or training”.

Once a government has given a problem a name it must develop effective new strategies for dealing with it. That too is in train, The Sunday Times told us last week, replete with urgent cabinet meetings, study groups roaming about the country and even a “Neet target” to reduce the Neet population by 20% by 2010.

You may use whatever euphemism the government adopts, but it’s still the underclass. Its numbers are not going to be reduced by 20% by 2010. Its numbers will increase. The good news is that the rate of increase will probably begin to slow in a few years and in another decade or two Britain will have learnt to manage the problem — meaning you will have learnt how to keep the underclass from getting underfoot, even though its numbers are undiminished.

When The Sunday Times first asked me to look at the British underclass in 1989, the American underclass was about 15 to 20 years ahead of Britain’s. You were tracking the American experience with remarkable fidelity then and you are still tracking it.

From the beginning I have used the simple-minded assumption that Britain 16 years on would look like America did when I was writing, and that’s more or less the way things have worked out. Nothing about the underclass is rocket science. It’s all basic, the kind of thing our grandparents took for granted. It just has to be rephrased to accommodate today’s delicate sensibilities.

Our grandparents thought bastardy was a problem to be avoided at any cost. Today’s translation: children who grow up without being nurtured by two biological parents are at risk. Poverty isn’t the problem. Inadequate educational opportunities aren’t the problem. Social exclusion isn’t the problem.

Throughout history, societies around the world have been poor, with inadequate educational opportunities and with socially excluded people. Those same societies have been remarkably successful at ensuring that almost all children came into the world with two biological parents committed to their care. That’s the difference between societies with small underclasses (for every society has had an underclass) and with large ones.

Children today usually still have a mother with them. The problem is the growing number of children who have no father and who live in areas where hardly anyone has a father. Girls without fathers tend to be emotionally damaged.

Among other things, they tend to search for father substitutes among young males, which in turn increases the likelihood of repeating their mother’s experience. Boys without fathers tend to grow up unsocialised. They tend to have poor impulse control, to be sexual predators, to be unable to get up at the same time every morning and go to a job. They tend to disappear shortly after the baby is born. These are not the complaints of a conservative lamenting the lost good old days. They are social science findings that are as robust and unambiguous as social science findings get.

I use the word “tend” because none of these outcomes is carved in stone for any particular child. But we can’t deny a problem exists because some children of single women do well. Of course, there are many exceptions but the statistical tendencies are pronounced, and tendencies produce a large and problematic underclass.

Our grandparents thought you couldn’t “do” with a youngster who wasn’t brought up right. Today’s translation: social programmes for intervening with children at risk have consistently meagre results. This finding has even longer shelves of analysis than the literature on the children of single parents.

During the 1960s and 1970s, the Americans tried everything: pre-school socialisation programmes, enrichment programmes in elementary schools, programmes that provided guaranteed jobs for young people without skills, ones that provided on-the-job training, programmes that sent young people without skills to residential centres for extended skills training and psychological preparation for the world of work, programmes to prevent school dropout, and so on. These are just the efforts aimed at individuals. I won’t even try to list the varieties of programmes that went under the heading of “community development”. They were also the most notorious failures.

We know the programmes didn’t work because all of them were accompanied by evaluations. I was a programme evaluator from 1968 to 1981. The most eminent of America’s experts on programme evaluation — a liberal sociologist named Peter Rossi — distilled this vast experience into what he called the Iron Law of Evaluation: “The expected value of any net impact assessment of any large-scale social programme is zero.” The Iron Law has not been overturned by subsequent experience.