IF WE all received social grants, would some of us stop complaining about them? A little-noticed passage in last week’s budget speech opens an intriguing possibility — that we might find a way of reducing or ending the constant carping that undermines our most effective weapon against poverty. Evidence shows that social grants are an affordable and effective way of addressing poverty. Not only do they offer millions of people a way to deal with poverty, they have boosted local economies and created opportunities for buying and selling.

Across our racial and political divides, commentators denounce grants as a cause of the dependence to which, in reality, they are an antidote and insist they are unsustainable when they are the antipoverty measure that gives most value for money. It has become fashionable to insist people should receive them only if they send their children to school or visit public health facilities, even though no one shows that people keeping their offspring out of school or refusing to visit hospitals is a serious problem.

The attack on grants is not especially South African: it happens in any society that tries to ensure that social spending reaches only the "truly poor". While it seems to make sense to aim spending only at those most desperately in need, this almost inevitably ensures that the antipoverty programme will become the target of ill-deserved attack.

"Targeting" means that ways (such as means tests) have to be found to keep the less poor and the middle class away from benefits. This sets up a rivalry between the poor and those immediately above them. Since those above are usually better organised, this forces the grants programme onto the back foot.

Giving only to the very poor means also that a slice of the budget is very visibly allocated to the weakest members of society only. This usually stigmatises the grants. The recipients are made to seem like freeloaders taking advantage of their hard-working fellow citizens: think here of the frequent claims, never backed by credible evidence, that thousands of teenagers fall pregnant to get grants. And again, because the poorest are usually the least organised, grants tend to have many opponents and few defenders.

The more grants are targeted at the weakest citizens, the more pressure there is likely to be to keep them as low as possible — or scrap them. Our grants seem to be in no danger of being scrapped, but there are few voices pushing for increases and many parroting the usual cliches, unsupported by evidence, that grants "create dependency".

This is why the antipoverty programmes around the world that work best are those available to everyone — this ensures that the better organised have a stake in them and will defend them. Britain’s National Health Service was not demolished under Margaret Thatcher because too many organised people used it and would mobilise to defend it. The best way to defend programmes is to make them available to all.

This is exactly what, according to Finance Minister Pravin Gordhan, the government plans to do with social pensions. He said the means test would be phased out by 2016 and everyone "over a designated age" would receive the pension. This, he said, would make administering pensions easier and ensure that people with meagre incomes would no longer be discouraged from saving. In theory, this change should give everyone a stake in pensions and ensure that they enjoy widespread support. In practice, this is not necessarily so.

The middle class here doesn’t need free public schools or hospitals — this is why there is little middle-class enthusiasm for the National Health Insurance. And it certainly doesn’t need state pensions. This suggests that scrapping the pensions means test may achieve easier administration and encourage savings but will not build more support for grants — middle-class people won’t defend a benefit they don’t really need.

But, despite this, there is a chance that ending the means test could reduce pressure on grants. People are not driven by their economic interests alone — sometimes a symbolic change might alter attitudes even if it makes no difference to people’s incomes. The middle class may not need state pensions but getting rid of the means test would at least mean that pensions are no longer something paid to "them" with "our" money. And the result may well be that, over time, state pensions may come to be seen as a benefit we all enjoy rather than a scheme that gives goodies only to other people.

This is, surely, at least worth a try. Social grants are one of the few antipoverty programmes that work — for the economy as well as for the poor. We need, therefore, to ensure that they survive and continue to play their role. And this means that grants need to be seen by as many of us as possible as an asset, not a drain on resources. Ensuring everyone above an agreed age receives a pension might do something to make that more likely.

Friedman directs the Centre for the Study of Democracy.