The earliest 20th-century reference I have been able to find to a Laffer curve effect is in a 1901 article by London School of Economics professor Edwin Cannan. While conceding that a 100 percent tax rate on amounts earned that exceed £50,000 (£4.23 million today) would be fair, it would not raise any revenue, he said. After the first year, no one would earn more than that amount (at least where the state could see it), and the revenue yield would be zero.

This is a quote from Bruce Bartlett, “The Laffer Curve, Part 3,” Tax Notes, October 1, 2012. HT to Brad DeLong.

I was surprised when I read that quote. When I attended the first Austrian Economics Conference in South Royalton, Vermont in June 1974, I had met a sweet old man, still sharp as a tack, named William H. Hutt. Hutt liked to quote his mentor Edwin Cannan. When I got back to UCLA, I went to the stacks at the library and found a book of Cannan’s popular essays and newspaper columns titled An Economist’s Protest. Milton Friedman’s publisher later used that same title for a compendium of his Newsweek articles.

Reading that book, by the way, encouraged me to hone my writing skills so that I could do the same thing later in my career.

Back to Cannan. Bartlett’s claim that Cannan thought a 100% marginal tax rate would be fair just did not ring true. And, it turns out, it wasn’t true. I went to the 1901 Economic Journal article from which Bartlett quoted and found the relevant passage. Here’s what Cannan actually wrote:

Considered in its immediate effects, the kind of income tax described by Professor Edgeworth in one of his fascinating studies, which would simply remove the superfluity of the very richest persons, reducing say every one with more than £50,000 a year to that sum, would be productive of least aggregate suffering. It would make a given income go furthest. But the given income would not remain. The persons with more than £50,000 a year would not continue to have more than £50,000 a year, at any rate within the purview of the state, and next year the limit might have to be reduced to £40,000 a year, and so on till the interference with the productive force of the community far more than counter-balanced the economic saving secured by making the given income go furthest.

Note that, contra Bartlett, Cannan does not judge the fairness of such a tax. Rather, he, correctly, attributes to Edgeworth the claim that such a tax “would be productive of least aggregate suffering.”

Interestingly, and not surprisingly, there is a lot of good analysis in Cannan’s article.

On taxes on alcohol:

There is, too, in the chief European countries, a very large revenue derived from taxes on alcoholic liquors. Will any one pretend to believe that this taxation is maintained out of respect for the principle of payment according to ability? When did the habitual drunkard become so capable of bearing taxation? There are few forms of expenditure which form a less trustworthy criterion of ability than expenditure on alcohol. We all know very well that the drink taxes exist because, in addition to bringing in a large revenue, they are supposed to limit, to some extent, the consumption of a commodity in which people are prone to indulge to excess.

Another excerpt:

There might, of course, be taxes of such a character that no lapse of time would condone their original iniquity. A special tax, for example, on red-haired people, or people under 5 feet 6 inches high, could not be sanctified by any length of time.

If I recall correctly, the red-haired example has persisted. Harvey Rosen and Ted Gayer use it in their textbook, Public Finance.