Welfare and the Tinker Bell policy
The UK benefits cap is a stunt policy designed to win attention, writes Tim Harford
‘The benefit cap starts being introduced across the country today, restoring fairness to the welfare state . . . with the amount of benefits working-age households can claim limited to the average working wage – £500 a week.’ Press release from the Department for Work and Pensions, July 15
“Restoring fairness to the welfare state.” I like that.
Indeed. The government spends about £2,500 per citizen a year on benefits – £160bn in total – and this policy is projected to save about £3 per citizen a year in the long run. That’s an awful lot of “restoring fairness” resting on a very small sum of money.
Surely it’s the symbolism of the thing.
Less “restoring fairness” than “creating the superficial appearance of fairness”, then? This is a stunt policy, designed to win attention, and I am depressed that you insist on talking about it. I think we should treat Iain Duncan Smith, the work and pensions secretary, the way we treat a three-year-old boy who blows raspberries in public: don’t point and make a fuss, because you’ll only encourage him to do something similar again.
All very well for you to suggest that the policy’s impact is tiny and we should ignore it – households affected by the benefits cap can’t just pretend it isn’t happening.
That is true. The DWP believes that 40,000 households will see their benefits reduced – and 20,000 households will see benefits reduced by £62 a week or more. So with some misgivings, let’s talk about the merits of the policy. We can think about it morally, practically, politically or not think about it at all. Mr Duncan Smith is firmly in the “don’t think about it at all” school of thought. He has made the statistically unsupportable claim that the threat of the benefits cap has pushed people back into work. He has been censured for this by Sir Andrew Dilnot, who chairs the UK Statistics Authority, and when challenged about it he responded that “I have a belief that I’m right”. It’s the Tinker Bell school of public policy: if you believe in fairies, the fairies will live.
If we vetoed any policy just because a foolish or duplicitous minister was attached to it, nothing would ever happen. Ignore Mr Duncan Smith: what about the policy itself?
There’s obvious moral appeal to the idea that people should get more money if they work than if they don’t. But this policy appeals to that moral intuition only in a crude way. What, for instance, is the moral significance of average wages? By definition – since the DWP tells me this is the median average – half of working households earn less than that. Then there’s the regional issue: the cap is national, but whether you look at the cost of living or the average wage, £26,000 looks a lot less generous in London than it does on Tyneside. In any case, working households earning £26,000 after tax enjoy plenty of top-up benefits. The figure of £26,000 sounds plausible but it could scarcely be more arbitrary.
But even if the ethical case is a bit vague, perhaps there is a practical case. Mr Duncan Smith believes that the cap will encourage people to find work.
That is plausible, although if Mr Duncan Smith had convincing evidence to support his belief he wouldn’t need to abuse the data he does have. But “will the benefit cap encourage some people to find work?” is not a very good question to ask. A well-designed benefits system looks at incentives in more detail. The issue isn’t whether a family on benefits is making more money than some entirely irrelevant benchmark. It’s what the effective marginal tax rate is as benefits are withdrawn when a household member gets a full or part-time job at a low wage. Mr Duncan Smith also seems uninterested in the cost of withdrawing benefits – both for the innocent children whose parents are claimants, and for the administrators who must figure out the details. Thanks to a leak, we know that Eric Pickles, the communities secretary, thinks the policy may leave the taxpayer worse off because of the administrative fiddliness and the cost of temporary housing for the suddenly homeless.
But it may still work.
Whether it works or not seems to be irrelevant. If the DWP cared about that, it would have run a randomised controlled trial of the policy. Instead it ran a pilot with the aim of ironing out the kinks. Why bother finding out whether the policy will work? As a DWP spokesperson informed me, “the policy is already decided”.
Also published at ft.com.