Follow by Email

Monday, October 23, 2017

Jared Bernstein: I got issues with the new CEA report [feedly]

I got issues with the new CEA report

First, full disclosure: Kevin Hassett, the chair of Trump's CEA, is an old friend, a good dude, and a guy I'm happy to see in this prominent post. But even as I endorsed him for the job, I publically worried about one area of his research where his claims go way beyond the evidence and ignore the counter-evidence. That area is the impact of tax cuts on growth, jobs, and particularly wages and incomes.

Thus, CEA's new report on the wage-boosting effects of the proposed cut in the corporate rate from 35 to 20 percent looks fatally flawed to me. It is a literature review that picks only the ripest of cherries, ignoring the large body of literature that goes hard in the other direction.

A particularly notable example of this problem is a citation of one study by tax analyst Jane Gravelle (et al) that CEA uses to support their thumb-on-the-scale results, while ignoring another Gravelle (et al) study that directly debunks CEAs main thesis that corporate rate reductions boost workers wages.

Worse, the Gravelle (et al) study that CEA ignores specifically and critically reviews the papers upon which CEAs work depends, including one by Hassett (et al) himself:

Cross-country studies to provide direct evidence showing that the burden of the corporate tax actually falls on labor yield unreasonable results and prove to suffer from econometric flaws that also lead to a disappearance of the results when corrected, in those cases where data were obtained and the results replicated.

One of those cases, as noted, was the Hassett study. As CEA notes, this is a study that tracks many countries over many years, and such studies are always sensitive to how the analysis is specified. That doesn't mean they're wrong, but it does mean you have to check as many of the model's assumptions you can to ensure that your findings are robust. When Gravelle et al do so, they quickly find the wage effect to disappear:

In every case, the coefficient estimates are close to zero and are not statistically significant at conventional confidence levels. In using annual data, we can find no evidence that changes in the top corporate tax rate affects wage rates in manufacturing.

This is also why you want to heavily discount CEA's Figure 1, which compares wage growth across very different countries with no controls (I'm not moved by Latvia's faster wage growth and lower corporate rate). CBPPs Huang and DeBot go through these issues in readable detail here.

Other problems with the CEA report:

–The analysis assumes average impacts are similar to median ones. This is a very big deal, because, as I argue here, even if the trickle-down chain they tout is operative (and the empirical data suggests it is not), there's still a big slip twixt cup and lip in the last step: the idea that faster productivity growth lifts median, vs. average, compensation. I wish it were so, but as per the figure below, it is not. Essentially, CEA assumes away wage inequality.


–To CEA's credit, though they argue that the proposal, including the shift to a territorial system that exempts foreign earnings by US companies from US taxation, reduces profit shifting (booking profits in places with lower rates), they don't directly add that effect to wages.

–They do, however, suggest this effect would be "additive" to workers' incomes. This suggests that CEA is conflating tax avoidance with actual job and wage creating economic activity, as Howard Gleckman points out here.

–This one's important too: CEA ignores the "excess profits" problem. That is, while economists argue about who benefits from a reduction in the corporate tax, we agree that it purely benefits firms with profits beyond normal returns on investment. So, for big firms that dominate in their industries–Apple, Alphabet/Google, Amazon, etc.–the corporate cut clearly boosts profits far more than wages.

–CEA is clear that their analysis only covers the corporate part of the tax change, and that's reasonable. Except for when we're talking about distributional results, we must consider the impact of the tax cuts as we understand them in their entirety. Otherwise, we risk overlooking aspects that hurt middle and low-income households. We at CBPP have stressed these impacts, particularly the losses at the low end if the tax cuts are eventually paid for by spending cuts (see figure below).

Gussy it up any way you want, this is just the latest example of non-credible, trickle-down fairy dust. Corporate tax cuts will merely boost the profits of a sector that's already highly profitable at the expense of the working class. Given its trillions in retained earnings and low capital costs, the corporate sector could already raise investments and worker pay if it wanted to do so.

Thanks again, Donald and co.

 -- via my feedly newsfeed