Babies should be savers, too

Sometimes you have to feel sorry for Gordon Brown. His advisers scour the globe for the next big thing. There are seminars, regional pilots, elaborate costings and appraisals. Finally, he stands up on budget day to announce the national roll-out. What happens? The commentariat takes one look, yawns and dubs it a transparent effort to distract from higher borrowing or another example of micro-meddling by Number 11.

That's on a good day. On a bad day, they don't pay any attention at all. Such was the fate of the Child Trust Fund. This is free money, for everyone born after September 2002, and it's scarcely raised an eyebrow since the Chancellor announced it in 2003. Now the plan is up and running, the Mirror has started a 'Don't be a Dummy Mummy' campaign to encourage foot-dragging parents to set up their children's accounts (1.2 million have yet to do so). But from the pundits, not a peep.

It was easy to miss the significance of the idea. For one thing, like many a Brown scheme, you could have sworn he'd announced it several times before. And naturally, it wasn't the only new wheeze announced in that budget. But probably the biggest reason for the collective sneer was the amount: £250 for every child born after September 2002, with another £250 for children born in the poorest third of households. To most in the Westminster bubble, this was such a laughably small amount for a trust fund that it wasn't worth the wasted breath.

In the long run, they may be right. And it will take 18 years for the policy to show its first effects. But the initial reaction speaks volumes about the very gap between the 'asset rich' and the 'asset poor' that the Child Trust purports to be trying to address.

If poorer parents put the £500 voucher in an account and don't contribute another penny, with 5 per cent real returns and modest fees, the child could expect to have upwards of £900 waiting for them at the age of 18. That sounds like not very much at all. But according to the Institute for Fiscal Studies, half of the UK population had net savings of less than £600 in 2000. That doesn't include housing, but the vast majority of people with little or no financial savings have no housing wealth either.

So £900 is at least an amount worth talking about. Will it make a big difference to the life chances of those 18-year-olds? It depends on which piece of government-funded research you believe. The Treasury based the 'baby bond' idea on research by the Institute for Public Policy Research, using a long-term database of 12,000 people. It suggested that even very small amounts of positive savings early in life could make a big difference to people's lives. Other things being equal, it found that anyone with £300 or more in savings at the age of 23 was more likely to be employed 10 years later and more likely to be leading a healthy, happy life.

For a government that wants to do something about the life chances of the poor, without scaring the markets or the rich, this magical 'asset effect' had to sound like the ultimate free lunch. The latest evidence suggests that social mobility is declining, with parental income and wealth playing a larger role in children's future than in the 1950s.

In fact, one academic says the postcode a baby is born in can now tell you more about their future life than can their DNA - their chances of going to university, their future income, even how long they're going to live. But the amounts involved are enormous. By one estimate, a child born in the top 10 per cent of postcodes in the UK will now have an average net worth, at birth, of £80,000. The asset effect says you can make a dent on this generational postcode lottery for a measly few hundred million a year.

That is why governments from around the world - right and left - have shown a lot more interest in the Child Trust Fund than the British media have. People like David Willets are interested in the idea as well; declining social mobility is a problem for one-nation Conservatives as well as for Gordon Brown. But there's a catch. In 2003, the Treasury and the Department of Work and Pensions commissioned more research into the asset effect. The academics went back over the IPPR findings.

They found other factors - whether people had children in early life, for example - could entirely account for both the savings in early life and the outcomes later on. Taking these other things into account, the asset effect disappeared.

I put this to David Blunkett, now Secretary of State of State for Work and Pensions. Interestingly, he blamed it on senior officials at the department who had long been 'highly sceptical of the asset agenda'. The research was by independent academics, not civil servants. But you can see why he would prefer to dismiss it, for it suggests that you can't have your New Labour cake and eat it.

A fund of £600-£1,000 at the age of 18 won't do anyone any harm. Given the average wealth figures I mentioned, it will probably do quite a lot of good. But to make a big difference to social mobility in the UK, the new evidence suggests you would have to spend a lot, lot more.

Bruce Ackerman, an American academic who is one of the spiritual godfathers of the Child Trust Fund, thinks that to make a real difference with asset-based welfare, you need £50,000 trust funds for everyone when they turn 18, paid for by raising inheritance taxes to penalty rates. Now that we would notice on budget day. It would also cost about £35 billion.

Something tells me Gordon Brown will be sticking with his £250 cheques for a while yet.

· Stephanie Flanders's Analysis programme, The Asset Effect will be broadcast on Radio 4 on Thursday at 8.30pm. Will Hutton is away