Megan McArdle reminds us of a very important point:
... I have been puzzled by the number of liberal bloggers who have been taken with Dylan Matthews' compendium of economists and politicians commenting on where the Laffer Curve might maximize.
I mean, it's an interesting side question, but it doesn't really tell us much about policy, unless you actually think that the object of policy is to maximize the share of income the government takes.
Megan is of course correct. The goal of government policy is not to maximize how much its citizens send it in tax revenue. Government taxes are an expense for citizens and as with all expenses (gasoline, food, shelter, etc.), people should seek to minimize them.
But she doesn't go far enough. The Laffer Curve - which relates tax rates to government revenue - does tell you where the government should set tax rates. But it's just not at the peak. The amount the tax rate should be set at is the amount that maximizes total earned income for its citizens. After all, if it sets it any higher, it will deprive citizens of income - i.e. it will make them poorer.
The way to find this from the Laffer Curve is to divide total government revenue (Y axis) by the tax rate (X axis). For example, here is a sample Laffer Curve (to my knowledge, no one knows the exact shape of this curve so I just made one up). As you can see, the peak of the curve (blue line) is at a tax rate of 70% and it smoothly slopes from 0 in either direction:
The dashed orange line is the total income of the citizens in that country. As you can see, in this (albeit fictitious) example, the peak of that curve is at a tax rate of 0.
Will this value - tax rate = 0 - hold for all Laffer Curve shapes? No, and it probably doesn't hold in the "real world" version. But it's worth noting that it is rather difficult to devise a shape of the Laffer Curve where the tax rate that maximizes wealth for the citizens is not quite low.