Harburger Triangles

Greg Mankiw:

A smart reader (who, because of his current job, prefers anonymity) offers this input about the recent debate over fiscal stimulus:

I've enjoyed your stirring the pot on the great stimulus discussion in recent weeks. One thing I haven't seen much discussion of is the idea that there are intertemporal tradeoffs involved. Suppose, for purposes of discussion, that stimulus does actually work, but that taxpayers do ultimately have to pay for it. In that case, there is a tradeoff between increasing economic welfare in the short-run and reducing economic welfare in the long-run (because of the distortions of raising taxes). Are there any well-established macro models that provide some guidance on how that balancing should be done? Many analysts seem to believe that the stimulus should be sized to close as much of the output gap as possible. But it isn't obvious to me that's true if, in a perfect world, we are balancing the short-term gains against long run costs.

The reader is right that this consideration has not gotten much attention of late. Why? I suspect the answer is that those who are most confident in Keynesian policy prescriptions are most skeptical about the distortionary effects of taxation. To put it perhaps a bit too bluntly, the Keynesian mutliplier is about income effects, while neoclassical tax distortions are about substitution effects. For those of us eclectic enough to see the world including a variety of effects, both Keynesian and neoclassical, policy decisions are far harder than they are for those eager to focus on one effect while setting the other close to zero. You can guess which effect those on the far left and those on the far right focus on.