Rawls, the budget and the case for punitive taxation

Rupert Read has instigated a fascinating – if slightly surreal – debate about what John Rawls would have made of the British government’s decision to cut the rate of income tax on the highest earners from 50% to 45%. Read observes that many of the arguments made in favour of the cut are Rawlsian in tenor, and suggests that this undermines Rawls’ philosophy. However, I think that Read’s framing of the debate as being about Rawlsianism and levelling down is unfortunate, as it obscures the important objections he has to the tax cut, and divides him off from potential allies who are likely to be sympathetic to his view. The real argument at the heart of Read’s post – that there is good reason to tax the rich, even if it has limited fiscal impact, and even if it does some harm to the economy – is one that deserves greater emphasis, especially given the fact that a similar debate looms imminent in France.

Rawls’ ‘difference principle’ insists that inequalities can be just if they maximise the position of the worst off in society. Certain Conservatives and businessmen have argued that the worst off in Britain are better off without the 50% tax. Therefore, Read argues that Rawlsians ought to defend the tax cut.

Read’s argument depends on two empirical assumptions:

a) that the 50% tax rate does not benefit the worst-off, and

b) the 50% tax rate reduces inequality.

Read doesn’t actually claim that a) is true – he only asks what would be the case IF it were true. As he observes, many Conservatives and businessmen have argued that the 50% tax rate doesn’t raise a significant amount of money for the government, and harms the poor by acting as a drag on the economy. However, these claims have been rejected by others. Rawlsians can therefore reject the tax cut on empirical grounds.

Even if a) is true, Read’s argument in favour of keeping the 50% tax rate also depends on assumption b) – that it promotes equality. Now if the 50% tax doesn’t reduce inequality by making the poor richer (since we’ve granted assumption a)), then it has to make the rich poorer. It is far from obvious that the 50% tax does have this effect – if the rich mostly avoid it, then they will not be any the worse off. However, the 50% tax may have incentive effects which do not help the poor – the rich might choose to work less, and therefore earn less. It could also be the case that the economic loss resulting from the tax hits the rich disproportionately. Assumption b) is therefore plausible, but does not obviously follow if a) is true.

Read’s argument is that even if cutting the 50% tax leaves the worst-off financially better off, it ought to be scrapped “Because we would still be creating a more unequal society”. In other words, Read looks to be arguing that equality is valuable even if it doesn’t benefit anybody. This makes it appear that the argument is basically about the levelling down objection: Rawlsians don’t believe in levelling down, Read does.

This is an old and entrenched debate in political philosophy. Read’s apparent position – that equality can be good, even when there is nobody benefiting from it has been defended, most famously by Larry Temkin. But looking at the arguments Read provides in favour of equality, this framing of the debate is entirely unnecessary.

The levelling down objection suggests that equality makes nobody better off. But Read invokes the book The Spirit Level in defence of equality. The whole point the arguments in The Spirit Level is that whether or not equality is good in itself, it has all manner of beneficial consequences – it leads to a happier, healthier and less violent society. So Read actually rejects assumption a) – he thinks that the 50% tax does improve the lot of the worst-off, just not financially. Indeed, given that Read’s fundamental point is that the effect of tax changes ought not to be evaluated in purely economic terms, it is bizarre that he makes this error, assuming that making someone richer is the only way to make them better off. (Rawls’ notion of primary goods is less crude, but has still been criticised as too narrow).

Read offers a second argument in favour of keeping the 50% tax that doesn’t even have anything to do with equality: “high tax rates are a good thing inasmuch as they discourage the culture of overwork which grips societies like ours”. This could be framed in a couple of different ways. First, it could be justified as a paternalist measure, similar to taxes on cigarettes and alcohol, which seek to discourage harmful behaviour. Second, taxing overwork might be seen as adjusting for a negative externality, or solving a collective action problem. Your decision to work long hours imposes a cost on me, since it increases the pressure on me to work similar hours. Perhaps we could all benefit from cutting our hours, but nobody can do it unilaterally – this provides the classic conditions for government intervention.

What I’ve tried to suggest here is that Read’s target is mistaken. Rawlsians, or others who accept the levelling down objection, are not his real enemies. Rather, the significance of Read’s argument is that it provides a corrective to the materialism of so much of the budget analysis. His insight is that taxes are not just about raising money, and it is only a narrow Rawlsianism that would object to that.