Should-Read: I find Jason Furman totally convincing here. Which raises the question: what are those he is criticizing thinking?
More distressing, perhaps, is Jason’s defensive “(warning: irrelevant nerdy thread)”. If economists’ models are worth anything at all, it is only because you take the time to understand them and do them right…
Jason Furman: Wage Increases Under the Unified Framework: Some talk lately of Ramsey models and their implications for wage increases under the Unified Framework (warning: irrelevant nerdy thread)…
…I love the Ramsey model as much as the next person (OK, not as much as Greg or @caseybmulligan) but it has major limitations for this q. Ramsey ignores: (1) transition (2) distribution, (3) financing, (4) Δ interest rate, (5) effective marginal rates & (6) supernormal returns.
TRANSITION: May take a long time to get to the new steady state level. Is relevant because direct changes are sooner, matter more in PV.
DISTRIBUTION. POTUS claim is middle-class better off. Stipulate Greg Mankiw’s toy example is right: $200b corp tax cut, $300b wage increase. The means wages +3% (or $2,000—already below CEA but nevermind). But implicit in model is lump sum financing of $1,600/household. A household making $50,000 and no capital ownership gets a $1,500 wage boost and a $1,600 tax increase—or $100 worse off. Broader point: in model growth comes from replacing distortionary taxes with lump sums. Can’t ignore when evaluating middle class impact. Other models CEA has been pointing to also assume lump sum/other financing—but somehow only TPC gets attacked for filling in details.
FINANCING. Of course the Unified Framework does not have the $1,600 per hh lump sum financing. So it doesn’t have the growth effect either. The Unified Framework has deficit financing. Over time this reduces National Income—through less investment or more foreign borrowing. See, eg, the Penn-Wharton Budget Model of the President’s earlier principles (they have not updated for new plan). You can get different results in different models, is sensitive to when/how financed, but generally cuts or reverses growth effect.
INTEREST RATES. The above assumed that the US is not a small open economy & interest rate not the invariant social rate of time preference. I am very comfortable with those assumptions. I suspect you are too. Enough said.
EFFECTIVE MARGINAL RATES. Much investment is already debt-financed. It faces an effective rate ~0%. Cutting corp rate from 35% to 20% would cut the Effective Marginal Tax Rate by ~5pp give or take a lot. This is one reason why full-scale dynamic analysis of rate cuts with Ramsey models are so much smaller than these toy examples would suggest. If you have expensing and no interest deduction then cutting statutory rates does not reduce the EMTR at all. I know, State taxes, possibility that avg rates affect lumpy location decisions, etc. It is complicated. Point is simple model overstates.
SUPERNORMAL RETURNS. The Ramsey model only has normal/competitive returns to capital. The real world also has supernormal returns/rents. The deleted Treasury study estimated that 63% of corp income is rents—and some evidence rents are rising as concentration grows. In theory cutting taxes on rents will lead to higher after-tax profit but not anything else (“you get what you get and you don’t get upset”). In practice, maybe some of the extra rents shared with workers in a bargaining model. But not much and likely higher-income workers.
CONCLUSION. A lot of ways to look at impact of corp cut on middle class. None of them show middle-class families WAY better off.
Part of the problem here is that Unified Framework is incomplete (legislation will also be incomplete by not specifying future financing). A full(er) specified plan will allow a more complete assessment of the growth/distn/welfare aspects of the plan together. I hope we have time for that complete assessment…