They study the share of top incomes in five “Anglo Saxon” economies. And they show that they are quite sensitive to moves in the top tax rate. A one percentage point cut in the top marginal tax rate leads to a rise in the share in incomes of the top 1 per cent of between 0.11 and 0.18 percentage points, depending on how you control for other things.
This is consistent with the idea that the highest earners do indeed work harder when taxes are lower.
However, they don’t work so much harder as to offset the mathematical tendency for a lower tax rate to bring in lower revenue. These responses imply that the tax rate that maximizes the tax paid by the top 1% is between 55% and 90%; this is a wide margin, because it‘s hard to estimate at all precisely supply responses to tax changes.
So a 50% top rate gets us nearer to maximizing revenues.
You might object here, quite reasonably, that we should distinguish two things - the tax rate that maximizes work effort versus the rate that maximizes tax revenues. A higher tax rate can do the latter even whilst reducing the labour input of top earners.
This, though, raises a question which I think doesn’t get the attention it deserves: is it a bad thing if the very rich work less? For example, if Fred Goodwin had been lazier he mightn’t have wanted the hassle of taking over ABN Amro. If Wall Streeters hadn’t worked so hard, they’d not have invented the derivatives that brought the banking system down. A little more idleness would have benefited us all.
High tax rates don’t just deter valuable skilled work. They also deter counter-productive work or rent-seeking.
If you think all this adds up to a case for taxing the rich, there is one aspect of this paper which mightn’t be so welcome for the left. Leigh and Atkinson show that the share of top incomes across the five nations over the last 100 years has been highly correlated; the trend in the UK - for the share to fall between the 1920s and 70s but rise thereafter - is very similar to that in the US, Australia, Canada and New Zealand.
This suggests that top income shares might be driven by common, global factors. If so, there are limits to how far purely national policies can increase equality.