This is a bleg. I'm looking for someone who: understands optimal tax theory better than me (shouldn't be too hard); can explain it simply (may be harder).
Here's the question: can it ever be part of an optimal tax system to have 100% marginal tax rates on some part of the income distribution?
Leave aside the case where some people enjoy working, and so will work for free.
Leave aside the case where there are gaps in the distribution of attributes over the population, so that nobody would actually be facing those 100% marginal tax rates.
My intuition says there's a Laffer Curve out there somewhere, and so 100% marginal tax rates can never make sense. We could cut tax rates, make the people taxed better off, and collect more revenue at the same time.
But I also recognise the argument behind marginal tax rates falling as income rises in optimal tax theory. (Lowering the marginal tax rate on the poor reduces the disincentives on the poor, but also reduces average tax rates on anybody richer than them.)
Can the second argument ever outweigh the first, so that it might be optimal to have 100% marginal tax rates on the poor? My gut feeling is it can't. But I can't think of how to prove it.
The policy relevance is that, as Stephen's previous post reminds us, in many countries the poor do pay 100% effective marginal tax rates, once clawback of benefits is taken into account.
This has always struck me as some sort of travesty. I can't understand why the "left" (who worry about the poor), and the "right" (who worry about disincentives and Laffer Curves), don't join forces to change this stupid policy. But I can't quite prove it's a stupid policy.
(And before anybody piles on and starts wailing about regressive tax systems, make sure you understand the distinction between marginal and average tax rates, especially when we are talking about "effective" tax rates, net of benefits. It is perfectly possible to have falling marginal tax rates with rising average tax rates, if people with low income pay negative (net) taxes.)