The basic principle of the Laffer curve is easily demonstrable:

• First, imagine a situation where the income tax rate is 0%. Income tax revenues are $0.
• Second, imagine a situation where income tax rate is 100%. Everyone opts out of earning an income, and revenues are again $0.
• Third, we know empirically that when the income tax rate is between these extremes, revenue is more than 0.
• Fourth, assume that the distribution is continuous. I don't think it has to be, but it's easier to picture.

From these, we know that there must be at least one maximum between 0% and 100%. Furthermore, we know that there must also be some subset of the range from 0%-100% tax rate where increasing taxes decreases revenue. We can't say anthing about where that subset is, nor that there is only one such subset, just that there must be at least one.

“The Laffer Curve” refers specifically to a specific curve that Mr. Laffer drew, and his (studied, but not necessarily correct) opinions on where the US tax rate in the 1980s. That specific curve is now outdated at best, and entirely irrelevent to any discussion not focusing on economics policy of the 1980s. When I refer to the Laffer Curve on this page, I am referring only to the general principle.

Sowell makes the specific claim (Here, et al.) that for high income earners, at points in the past, tax revenue has risen after tax rates fell. He uses this economics arguement as the basis of several political arguements which are not testable, so I am looking at the only at the truth of the specific economics claim.

I pulled income tax data from the IRS website, looking at aggregate US data. The raw excel files which I downloaded are available hereus_taxes_by_income.zip. Admittedly, it's a bit of a misnomer to call them 'raw', as they are aggregate data. 1997 was the earliest that these were available in this format, and 2009 the most recent.

I pulled tax rate data from http://www.taxpolicycenter.org/taxfacts/displayafact.cfm?Docid=213. It may not be an unbiased source (I didn't bother to check) but this is factual information, and not subject to skewing. Also, various other sources agree; this one just had an option to download in excel. From the attached caveats: “Perhaps most importantly, it ignores the large increase in percentage of returns that were subject to this top rate.” I'll try to account for that, but no guarantees.

Constant dollar data is from http://oregonstate.edu/cla/polisci/individual-year-conversion-factor-tables. Again, I used it because excel download was an option. Pick the 1997 tables.

analysis.xls