Response to Paul Krugman on “Millionaire” Taxes

October 11, 2011

Paul Krugman today criticizes a blog post of ours (via David Brooks’ New York Times op-ed) estimating how much revenue a “Buffett rule” millionaires’ tax might raise in one year. He calls our blog post a “deliberate fraud” because we place the number into context with other large numbers, pointing out that one year of revenue would only reduce 8% of this year’s budget deficit or reduce 1% of the current national debt.

Krugman has no quibble with our calculations and the blog’s other comparisons; he argues that everyone should use a 10-year revenue projection when comparing revenue against the national debt, although he then reverses himself and says that that approach is “flawed” as well. He’s correct on that point at least: there’s simply no “right” way to compare a stream of new revenue (which could continue well past 10 years) against a static debt figure. What one should do is be honest about what you’re comparing so that the reader understands it; we did that and Krugman is misreading or misrepresenting our post to claim otherwise.

His preferred solution is to compare a 10-year projection but doing so is more complicated and makes the number less certain. Projections get less accurate the further out one goes, and the 10-year standard is fairly arbitrary: essentially short enough to not be preposterous and yet longer than the typical business cycle. It is also very difficult to estimate the “dynamic effects” of such a tax increase: if you were a millionaire and told the government was confiscating 50% or 100% of your income this year and in perpetuity, what would you do? A 100% tax might well only raise one year of revenue, as it would greatly reduce incentives to make money and/or report it to the IRS. Economists could reasonably predict that such a policy would be a revenue loser over a 10-year timeframe. But estimating such effects is very difficult to do with precision.

Even a 10-year projection under static analysis (assuming high-income earners don’t change their behavior) doesn’t prove what Krugman thinks it does. Using CBO data on GDP and debt, we estimate that the new tax would raise $1.2 trillion over ten years, making an 8% dent in the publicly held national debt (projected to be $14.9 trillion in 2021).

(ten-year outlook)
Taxes Paid Taxes Paid at New 50% Rate Revenue Gained Reduction of Debt
Taxing the “Super-Rich” (Annual Income of $1 to $10 Million) at an Effective Rate of 50%

$1 trillion

$2.2 trillion

$1.2 trillion


So Krugman’s “seriously significant [one-year] sum” reduces the debt by 8 percent over ten years, as opposed to 1 percent over one year as we previously posted.

Related Articles