I don't have Blinder's book handy--I'm on a flight--but here are some excerpts from Krugman's 1994 book, Peddling Prosperity.
If you try to tax people, they will try to find ways to avoid paying. Some people will resort to simple fraud--hiding their income and cooking their books. Some will resort to elaborate legal evasions, arranging for paper losses to offset real gains. But in the United States the main way that people try to avoid taxes is by avoiding doing things that are taxed. Unfortunately, these things include working and investing.
Economists, left and right, have always agreed that the tendency of people to change their behavior to avoid taxes--in the jargon of economists, the distortion of incentives associated with taxation--represents a hidden, extra cost of government. This is a basic truth. (p. 66)
[S]uppose that the government decided to spend an extra $10 billion--1 percent of national income. [He's dealing with a hypothetical economy whose national income is $1 trillion.] You might think that it could pay for this spending by raising the tax rate 1 point, from 30 to 31 percent. But in fact, this would certainly not be enough, because a rise in the tax rate would induce at least some people to work less hard or work shorter hours. (p. 67)
Furthermore, the reduction in the incentive to work is somewhat more than this number would suggest, because the U.S. tax system is designed to be somewhat progressive, that is, to make high-income people pay a higher share of their income in taxes than lower-income people. This seems reasonable on social grounds, but has the side consequence that the "marginal" tax rate--the rate you pay on the last dollar of your income--is higher than the average tax rate. If the economic program proposed by Bill Clinton in early 1993 passes, very high income individuals will pay a marginal federal tax rate of 46 percent, plus a few extra points for state and local taxes. This is not trivial: without any question, the negative effects of taxes on incentives are significant. (pp. 68-69)
So why did I think that Paul Krugman denied that the incentive effects of high marginal tax rates on high-income people are potentially large? Because when he talks about the supply-siders, the people who essentially brought this insight about incentive effects to the forefront in the economic debates of the late 1970s, he is so hostile to them.
You can't, for example, agree with the quotes above and, at the same time, deny the Laffer Curve. As I've written earlier, the Laffer Curve is necessarily correct. But in this same 1994 book, Krugman writes:
[T]he sophisticated public finance arguments of Martin Feldstein were eventually to give way to the crude and silly Laffer curve. (p. 56)
The Laffer Curve is crude: so are demand and supply curves. But they're all awfully useful. So it's not silly.
Indeed, later in the book, Krugman recognizes that the Laffer Curve is not silly. He writes:
Nobody questions that something like the Laffer curve exists; but even the supply-siders are skeptical about whether the U.S. economy is really in the "backward-sloping" section. (p. 95)
This is an interesting admission on Krugman's part for two reasons. First, he recognizes, as one must, that the Laffer curve exists. Second, he admits that supply-siders don't kid themselves that we are in the backward-bending portion, the part where an increase in tax rates reduces government revenues and a decrease in tax rates increases government revenues. I so miss the Paul Krugman of the 1990s.