« Dean Baker Has to Spend His Time Countering More Misinformation from the Washington Post | Main | Don Boudreaux vs. Dani Rodrik on Industrial Policy: I Call This One for Don--I Think It's a Knockout »

July 13, 2007


buck smith

I agree this curve is not valid. But the Laffer curve has to be true. If tax rates are zero government revenues are zero. WE have a lot of example of tax rates of 100%, Soviet Union, E Europe, etc. and those governments do not generate as much money per capita as the one s where tax rates are 10% to 50%. Capitalism works. Lots of other factors besides tax rates are important - regulations, labor laws, the entrepreneurial spirits in the populace, stable currencies, enforceable contracts, and efficient capital markets. These factors explain the spread in the graph.


Let's assume that the endpoints of the Laffer Curve are correct. (That is where Laffer reportedly began his curve, at the endpoints) Nobody knows what the shape of the curve between the endpoints might be. It is the shape of the curve that is important, not what the values of the curve might be at the endpoints.


Doing the highly reliable fit-by-eye method (be sure to lick your thumb first and stick it out at arm's length to sight with), I do have to admit that I can see the rudiments of a plausibly Laffer-esque curve there. I'll even admit that the WSJ isn't too far in the X axis off the peak I see, which I'd place in the high 20%'s. But the *level* of the peak I'd put around 4%, and the curve would be *much* flatter. Honestly, there really isn't enough data to confirm the peak or do more than vaguely suggest the shape of the curve, much less tell where the top endpoint is. But there is a clustering at 35% that is starting to pull the curve down, and that is suggestive of having left the peak behind. A linear fit seems rather too simplistic here, though it's preferable to the blatant absurdity of the WSJ curve.

Free Lunch

Aren't the AEI and WSJ misreporting the actual effective tax rate in Norway?

The comments to this entry are closed.


Search Brad DeLong's Website