"April 3, 2012, 4:53 pm
The Simple Analytics of Soaking the Rich (Wonkish)
Update: I see that some people are saying, “But taxes discourage work!” Um, that’s what the supply curve in the figure measures; if taxes didn’t have any effect on work, it would be a vertical line.
So President Obama is going after the Ryan plan; good. And better yet, he’s taking on the underlying economic premise, which is that low taxes on the rich are the answer to, well, everything, and good for everyone. For this premise is just bad economics.
I’ve written about this before, drawing on Diamond and Saez, but I thought I’d try a different take.
The way Diamond and Saez do the analysis is to argue that because the rich are rich, their marginal utility of income is very low, which means that at the margin their income doesn’t matter for social welfare. So they should be taxed at the rate which maximizes revenue, which is 1/(1+ε) — where ε is the elasticity of labor supply from the rich. And since we have a lot of evidence suggesting that ε is quite low, the appropriate tax rate for the rich is quite high — 70 percent or more.
But what if the rich in their Galtian goodness supply something nobody else can? Call it J, for jobcreation. Doesn’t the imperative to encourage J mean that we should keep their taxes low? Actually, no.
So here’s my alternative way to think about it: we can think of society as a whole — or, if you like, society not including the top 0.1 percent — as having monopsony power over the rich. The picture looks like this:
The optimal thing, from the point of view of the non-rich, is to set a tax that makes the cost of hiring rich people to produce J equal to the true marginal cost of that J, a cost that includes the fact that buying more drives up the price of inframarginal purchases. And if you grind through, you find that the optimal tax is … 1/(1+ε). Even if the rich are uniquely able to supply the magic of jobcreation, they should face much higher taxes than they do.
And this is all perfectly standard economics — indeed, Econ 101.
So what’s the basis for claims that we must tax the rich lightly? Often, it seems as if conservatives believe that there are somehow big positive externalities to what the rich do; it’s as if they believe that industrial policy is nonsense, unless the industry in question is jobcreation by the rich, in which case loose arguments about huge spillovers are just fine.
But the simple analytics say that we should soak the rich, hard."
I’ve written about this before, drawing on Diamond and Saez, but I thought I’d try a different take.
The way Diamond and Saez do the analysis is to argue that because the rich are rich, their marginal utility of income is very low, which means that at the margin their income doesn’t matter for social welfare. So they should be taxed at the rate which maximizes revenue, which is 1/(1+ε) — where ε is the elasticity of labor supply from the rich. And since we have a lot of evidence suggesting that ε is quite low, the appropriate tax rate for the rich is quite high — 70 percent or more.
But what if the rich in their Galtian goodness supply something nobody else can? Call it J, for jobcreation. Doesn’t the imperative to encourage J mean that we should keep their taxes low? Actually, no.
So here’s my alternative way to think about it: we can think of society as a whole — or, if you like, society not including the top 0.1 percent — as having monopsony power over the rich. The picture looks like this:
And this is all perfectly standard economics — indeed, Econ 101.
So what’s the basis for claims that we must tax the rich lightly? Often, it seems as if conservatives believe that there are somehow big positive externalities to what the rich do; it’s as if they believe that industrial policy is nonsense, unless the industry in question is jobcreation by the rich, in which case loose arguments about huge spillovers are just fine.
But the simple analytics say that we should soak the rich, hard."
No comments:
Post a Comment