Wednesday, 9 January 2013

Living Beyond Our Means-Testing

The political fiasco that is the impending withdrawal of child benefit for those earning over £50-60k has resurfaced in the news cycle over the past week. Several different objections to the policy have been raised, and I find some of them far more compelling than others.

Do the changes unfairly penalise the ‘squeezed middle’?
One of the biggest objections to the change is simply that, for people earning around the threshold, losing child benefit is going to hurt them a lot. Politically, this charge has quite a lot of weight, and it resonates with the indignant middle-classes who follow politics closely. It’s likely an example of salience bias. The costs of the policy are concentrated on a small, vocal group – you’ve probably seen some of them talking up how they’re not poor, but not really rich either, and the money really helps (which of course it does). The benefits of reducing the deficit, or spending the money on something more socially useful, are widely spread across people who can’t fully perceive or appreciate them.

The reality is that, as the graph below shows, the ‘squeezed middle’ are anything but. Austerity measures have hit the very richest hard, because of higher capital gains taxes, cuts to pension relief, attempts to clamp down on various forms of tax avoidance, and a host of other measures. They’ve also hurt the poorest in society, who rely heavily on the tax credits and benefits that were cut so controversially in the Autumn Statement. The middle classes, on the other hand – most especially the upper-middle class households who are looking at losing their benefits – have largely escaped scott-free.

What about single and multiple earners?
One of the other criticisms of the child benefit changes is that, because they assess your income through the tax system, single-earner families are penalised. If a household with one earner breadwinner is taking home £60000, they will lose all of their child benefit, while households with two earners on £50000 each – a much higher combined income – will lose nothing.

One of the main problems with assessing income by household is that it would be much more complicated to do. It’s (relatively) easy to take child benefit away from people who earn over £50000, because they tell the government how much they earn when they pay taxes on that money. The government, however, can’t find out as easily what the combined income of a two-earner household is. Mandating that it do so would introduce more unnecessary and costly complexity to an already byzantine regime.

Moreover, though, the whole debate seems absurd to me. One of the most infuriating things about the US tax system, according to Americanwonks, is the so-called ‘marriage penalty,’ where a couple who file their taxes jointly end up paying a lot more in tax than they would if they were single and filing separately. To criticise child benefit for hurting single-earner families is to criticise child benefit for not having a marriage penalty. If child benefit was assessed on joint incomes, we might end up in the absurd scenario where people’s child benefit would rise if they divorced. Further incentives for fraudulent living arrangements in an already-flawed system seem like a pretty bad idea.

There’s another fundamental economic point to be made here about the distortionary nature of taxation. In this case, the distortion comes about because taxes value wage income, but they don’t value work in the home. Imagine the scenario with one person earning £60000, and their wife – sadly, but realistically, it’s going to be the wife – staying at home, doing domestic labour, looking after the kids etc. Imagine that the wife’s earning potential is, say, £40000, if she went back out to work full time. This suggests that the time she spends with her kids, and the value of her domestic labour, is worth more to her than £40000 a year.

Imagine, however, that that domestic time is worth only £30000 a year. If there were no taxes, the woman would probably go out to work. Because a £40000 / year job would result in less take-home pay than the £30000 ‘family time’ is worth, the woman won’t go out to work, even though her preference is to do so. (Now that I think about it, I’m surprised that I don’t see libertarians advancing these arguments more often).

Of course, these distortions are necessary if you’re going to collect taxes – and collecting taxes is obviously necessary for society. But that doesn’t mean it’s a good idea to make the tax code even more distortionary. If child benefit was allocated by household earnings, then the £60000 breadwinner’s wife is going to face even more of a disincentive to enter the workforce, because her hypothetical income would put the couple above the withdrawal threshold for child benefit.

Raising Marginal Tax Rates
Perhaps a better criticism of the child benefit changes is that they will lead to unacceptably high marginal tax rates on many individuals. As the chart below shows, as you have more children, your marginal tax rate in the £50-60k region gets very, very high. With seven kids, you’d face a marginal tax rate of over 100% - meaning you’d earn more with a pretax income of £50k than with a pretax income of £60k – but even with only two or three, you’re looking at serious work disincentives, especially if you were to add on the effects of, say, a student loan. That’s not good for the economy.

One of the problems with the tax system, though, is that for a given level of (negative) tax on the very poor and on the very rich, someone is going to have to pay big marginal rates. I haven’t seen much empirical data on this, but it seems plausible that very high marginal tax rates on the middle class might be quite economically effective. The very rich have low marginal utility of income, and so they work much less if you tax them more. The very poor have low social incentives to work, so high marginal tax rates have serious effects on them. The same is not true for the middle classes. Their marginal utility of income is still quite high – extra income is worth a lot to them even if they don’t take it all home – and there are strong social pressures counteracting work disincentives.

There’s actually a surprisingly strong case that, given marginal rates are going to be very high somewhere in the code, it would be most economically efficient to have them hit middle earners. The middle aren’t being squeezed, but maybe they should be.

Complexity in the System
Ultimately, the strongest argument against the child benefit changes is that they don’t raise a whole lot of money, and they do make the system a lot more ridiculous and complicated. Anyone who didn’t opt out in time – and only about 250,000 of the million or so affected households did so – will have to fill out a tax return. This is a horrible punishment to inflict on people. Further increasing the confusion, they will still be paid child benefit; the state will then claw the money back through the self-assessment process. Given the other costs of the cut, and the relatively meagre £1.5bn it would save, you do wonder if there isn’t a way to take money from the welfare budget that doesn’t involve horribly expensive benefits.

More broadly, the complexity of the welfare system is one of its biggest problems. There are a huge number of different benefits and credits that all phase out at different levels. Many people don’t know what they’re entitled to, and what they aren’t. Ian Duncan-Smith’s Universal Credit is likely to simplify the process, but marginal tax rates on the poor will remain high. There really isn’t an easy solution to these problems, and it’s not clear that the benefits of means-testing child benefit do in fact outweigh the costs.