There has been some recent discussion of the Laffer curve on the blog as a result of reaction to a blog by Jolyon Maugham. As is, I think, quite well known, I am dubious about the existence of a Laffer effect at any of the current headline rates of tax in the UK, and the vast majority of those proposed. I think it may just appear if the Greens have a top rate of income tax of 60% with uncapped national insurance contributions.
In response to these comments Howard Reed, an occasional co-author of mine and a man whose work I much admire, posted some comments I thought worth sharing more widely. He said:
Most of the Laffer effect (I feel bad calling it the “Laffer effect” as Laffer is a total far-right crank and charlatan… for example, do a search on “Krugman New York Times Laffer” on Google and you’ll find several examples of why Laffer is a nutjob…) is NOT due to reduced hours or effort by high earners because the tax rate is so high. It’s due to increased avoidance activity and (e.g.) shifting remuneration from income to capital gains. A General Anti Avoidance Rule (as proposed by Richard) would put paid to most of the Laffer effect. Personally I think we would be fine up to at least 65-70% top rate but I appreciate this is a matter about which there can be intelligent disagreement. Some commentators (e.g. Saez/Piketty) have suggested that 80% is the revenue maximising top rate.
More fundamentally, it’s not clear to me that Laffer is right even on its own terms. Would the revenue yield from a 100% tax rate be zero? If we had (e.g.) a very high Citizens Income, so that no-one *needed* to work, I’m sure some people would go on working because they enjoyed their work even if tax rates were 100%. This might well be a minority of people and so I wouldn’t suggest 100% tax rates as a serious proposal, but Laffer’s argument (that the yield of a 100% marginal rate is zero) seems to me to be flawed if work conveys any positive utility, at least for some of the population. And there are certainly people around who like their jobs – for example there is evidence of people who win the Lottery and carry on in their job.
I well recall several of my staff in some of the companies I have run in my time for whom pay (and by implication, therefore, tax) was not the key or even, in all likelihood, a significant motivation for their coming to work. Various other factors were at least as big a motivational effect. In other words, I think Howard may be right with his second suggestion.
That tax avoidance is the main reason for the Laffer effect is beyond dispute. And as Howard notes, we could beat that if we wanted to.
It’s Diamond and Saez who calculated the peak of the Laffer Curve and found it to be 76%, not 80. And that’s in a hypothetically pure tax system, with no allowances. And with allowances that falls to 54%. And that’s in the US tax system, based as it is on a passport tax, not residency basis. Here in Europe we have a residency based system which is an obvious allowance: think tax is too high you can just bugger off.
There is, of course, no doubt at all that avoidance creates some Laffer effects. But Diamond and Saez reach their conclusions without considering them at all: they use the mixed influences of the income and substitution effects, the classical underpinnings of the theory. So, you can’t go around referring to their research and then conclude that it’s all about avoidance. Because their research shows it ain’t all about avoidance.
And I’m afraid that this is important, not just pendantry over the expressed opinions of idiots. Ritchie one year wrote the TUC’s Budget Submission. In which he claimed that a rise in the tax rate would lead to more labour being offered in the market. He was, obviously, assuming that the income effect prevailed over the substitution effect. Which for at least some people, some of the time, at certain income levels, it does, that’s why we’ve got a curve here. Digging deeper I finally managed to get him to explain his assumption. Which was that women married to high earners would be especially subject to that income effect and very little to the substitution effect. Much more so than the general population and that’s where the increase of labour on offer would come from.
Which is fascinating because one of the things we know we know about the income effect is that married women are less subject to the income effect than the substitution effect, much less so than the general population.
This is why these morons fuck up so often. They’ve not the slightest clue of the things they try to pontificate upon.