Last week I wrote a post that included the following taxonomy of "supply-side economics":
The "strong" version of the supply side argument is that tax cuts will generate enough growth to increase tax revenue. (Not to be confused with the general Laffer Curve proposition that tax cuts can do this, which will probably be true under some circumstances -- say, if a tax rate went from 100% to 99%.) The "semi-strong" version is that tax rates are the key factor governing economic growth. And the "weak" version is that the growth and efficiency generated by lower taxes is more important than the equality generated by redistribution. (The "weak" version pretty much just restates the big difference between the left and the right, so it's really quite general.)
I tried using this to make the point that, contra Greg Mankiw, "the general proposition that taxes affect incentives" is not unique to supply-side economics. Indeed, I wrote, "Everyone believes
that!" Dan Mitchell of the Cato Institute wrote me a pretty long email disagreeing with this. I'll quote a snippet:
You write that [Greg] Mankiw defines a supply-sider as "someone who believes in the general proposition that taxes affect incentives." You say that is a meaningless definition, though, because "Everyone believes that!"
This is not true, at least in an operational sense. The revenue-forecasting models used by the Joint Committee on Taxation and Treasury's Office of Tax Analysis are designed so that changes in tax policy - no matter how profound - have zero impact on the economy's growth rate. Double all tax rates? The economy will remain on the current path. Sweep away the internal revenue code and replace it with a low-rate consumption-based system like the flat tax? Doesn't matter, GDP will be unaffected.
So while the reasonable left is smart enough to admit that tax rates do affect incentives, they vigorously resist any effort to modernize the revenue-estimating system. Why? I suspect they understand that this would somewhat hinder the left's redistributionist agenda (at least to the extent to which it is financed by soak-the-rich tax policy).
I also would augment your characterization of strong, semi-strong, and weak versions of supply-side economics by adding the "somewhat-strong" version, which is that changes in marginal tax rates affect incentives and lead to changes in taxable income that can have a measurable impact on economic performance. I think it is fair to say that this is what most "supply siders" actually believe (which, to be sure, is different than what some overly-optimistic Republican politicians assert).
My response, which was really only a small attempt to goad Dan:
I don't know enough about revenue-forecasting to say much about the JCT. But while I'd merrily admit that tax cuts have an effect on incentives, it's not clear that the revenue estimates suffer to the detriment of your position for lack of a GDP component. (I am reminded of this old Jason Furman analysis.) But I shouldn't start an argument that I'd lose: Perhaps there is a widely accepted and more inclusive forecasting model that would be picked up by the government, if only there were less foot-dragging on the left.
On the taxonomy: Fair enough, although I think your somewhat-strong version would be pretty hard to distinguish from my weak version. Sure, taxes probably have a measurable impact on economic performance. But I imagine that someone on the intelligent left thinks the impact is smaller than you do, and someone on the intelligent left thinks the benefits of government reditribution are greater than you do. Or at least I believe that! (I described this difference, from your perspective, as "the growth and efficiency generated by lower taxes [are] more important than the equality generated by redistribution" -- and I hope that wasn't a cariacature. I really do think that's the best description of the reasonable economic policy differences between the right and the left.)
And Dan's response:
You have identified some of the key issues. Jason Furman types would agree that taxes affect incentives, but they would argue the effects are small. My counter response would be that even small changes in growth (say 1/10th of one percent annually) become very significant with the passage of time because of compounding.
But since we can't agree on the magnitude of those potential growth changes, we obviously would not have common ground on a model (assuming we could even overcome other methodological issues).
Shifting to taxonomy, your weak version is a good description of how the different sides approach the issue, but I don't think it captures my concern. The problem is not (at least in this context) that my counterparts on the left value redistribution more than growth. The problem is that the revenue-estimating system assumes that tax policy reforms have zero impact (not small impact, zero impact) on overall economic performance. So the reasonable left is allowed to sound, well, reasonable by admitting that tax policy can impact incentives, but those same people inevitably resist any reforms to the scoring process - even ones using their very modest assumptions about the magnitude of "supply side" effects.
I am especially troubled that they also seem uninterested in transparency. The JCT refuses to open its "model" to inspection (a step, I hasten to add, that in no way would compromise private taxpayer data). Why block public discussion and examination of the model, particularly by academics? Unless, of course, one has an agenda and fears that sunlight will make it impossible to defend the current system.
On a final note, Dan's first email began "Greetings from France," which struck me as a somewhat tragic opening salvo for a Cato economist. His second email shifted to the more appropriate, "I'm now in Austria."
This article available online at: