The Agenda

‘Where’s the Growth?’

How can we possibly believe that low marginal tax rates are good for growth if we’ve had lackluster growth and relatively low marginal tax rates! The relevant question is this: would growth performance have been even worse under higher marginal tax rates? The problem, of course, is that this question is extremely difficult if not impossible answer. The folk response to this question is, “Well, growth was spectacular in the 1990s, after an increase in MTRs!” Which is true! But one can again argue that growth would have been stronger had MTRs been lower.

Moreover, we have to tease out the short-run impacts of changes in MTRs from long-run impacts, as firms and industries respond over long periods of time to changing tax environments. Path dependence is a very powerful force, and prime-age workers vary in their elasticities, e.g., a worker with children and a mortgage isn’t in a great position to choose more leisure, but a worker approaching retirement is in a somewhat better position to do so. This is a case for varying tax rates according to demographic characteristics, e.g., age and gender. Yet I suspect that this wouldn’t go down very well with the voting public, though of course we offer innumerable tax breaks pegged to other qualities that aren’t obviously sensible for discrimination. Why people who have a strong proclivity towards homeownership should get a huge tax break remains a mystery to me, in part because I find the arguments from good citizenship to asset-building unconvincing in the extreme. (There are other, cheaper ways to pursue said goals.)  

And all kinds of other issues enter the picture. Tax policy interacts with other policies in many different, complicated ways. My basic view is that, for the reasons Alan Viard and others have carefully explained, high marginal tax rates have negative incentive effects that outweigh the potential revenue gains, while curbing tax expenditures yield revenue gains without the negative incentive effects, though of course there are distributional considerations.

I find that I feel compelled write this post at least once every two weeks. 


The Latest