David and Christy Romer have a new paper (via Tyler Cowen) about the effects of the marginal tax rates on reported income of the very rich using data from the interwar period. They find an elasticity of around 0.2 (meaning a 1% change in after tax income share increases reported income by 0.2%).
There findings support those summarized in Saez, Slemrod, and Giertz that find an elasticity between 0.1 and 0.4. Low elasticities generally mean that the top of the Laffer Curve is pretty far to the right. The Saez paper estimates it to be around 70%. That doesn’t mean that raising taxes doesn’t decrease the utility of those who have to pay it or that it doesn’t reduce reported income, just that tax revenue is maximized at a rate similar to what we had in the 1960s.