The neo-Laffer curve is a satirical construct created by Martin Gardner to establish the fallacy of one of many "laissez-faire" ideas that became collectively known as Reaganomics. It demonstrates a basic error of mathematical confusion that held in its sway, among other things, the executive office of the United States of America.
Arthur Laffer, a supply-side economist, postulated a simple graph of taxation percentage vs. government revenue, with the former on the y-axis and the latter on the x-axis. He argued that this graph would be represented best by a simple curve, roughly:
|******
| *****
| *****
| ****
| *
| ****
| *****
| *****
|******
----------------------
This Laffer curve was based on the obvious assumption that if the tax rate was 100%, there would be no incentive to work and thus no government revenue; and if there was a 0% tax rate, there would similarly be no government revenue. It is not hard to extrapolate that a 99% tax rate would be almost as burdensome, just as a 1% tax rate would be almost as pointless. As one approaches 90% and 10% on either extreme of the curve, government revenue increases, by extrapolation, and grows larger at 80% and 20%.
It is no exaggeration that this simple graph was used as the intellectual foundation of Reaganomics. Following Laffer's conclusions, the economic planners in the executive branch concluded that since the richest tax bracket had a rate on the upper end of the curve, lowering this rate would, as the Laffer curve predicted, actually increase government income until it reached a perfect "point" at the highest x-coordinate of the curve. This mathematical explanation seemed so simple and easy to follow that it was presented even by learned academics as uncontroversial proof of supply-side economics and the need to lower taxes immediately for the richest Americans.
Martin Gardner, writing in Scientific American, offered an alternative and more realistic view of government revenue, which he bitingly called "the neo-Laffer curve". It looked approximately like this:
|****
| * **********
| * ****** *
| ******** * * *
| **** ** *****
| ***** ****** *
| ** *** *
| **** *
|****
----------------------
Which is to say that past a certain point of common-sense extrapolation (the 100-99% and 0-1% extrapolations mentioned above), the curve was a tangled mass of gobbledy-gook with seemingly random and potentially infinite "high points" on the x-axis, and thus lowering the tax rate on the rich was as guaranteed to raise government revenue, as, say, President Ronald Reagan changing the color of his socks. Why? Because plotting government revenue and tax rates on a simple curve was frankly just plain dumb. Dumb beyond belief.
The truth of the matter is that the optimal tax rate for government revenue is guided by intensely immediate circumstances and numerous, nigh-uncountable variables, and all the wicked webs that make up history. Laffer's fallacy, whether conscious or unconscious, occurred when he assumed that simply because two ends of his graph could be easily examined and quantified, that the rest of the graph would follow the same rules and thus arrive at a "point". In fact, not only did the graph not follow those rules, it could not by definition of the system it represented.
Cognitive philosopher Daniel Dennett, in his works on human consciousness, has used Laffer's fallacy and Gardner's correction to demonstrate the fallacy written in an unspoken assumption of many of his colleagues: that because there is an easy recognition of the period of time when stimulus is applied and the period of time when the subject responds, that there must by definition be a theoretical "point" wherein "conscious" recognition of the stimulus occurs, between the stimulus and the response. There need be, in fact, no such point - and there also may be many such points - because the brain is far too complex in its workings for anyone to assume that the "line" of stimulus going in and the "line" of response going out leads to anywhere or anytime in particular.