The Corner

The one and only.

About That CRS Paper, Conservatives, and Data


Yesterday, the New York Times reported that, “after Senate Republicans raised concerns” about a paper by the Congressional Research Service on tax rates’ economic effects, the non-partisan research body withdrew the paper.

George Gilder wrote about the conclusions of the paper for a recent issue of National Review, and pointed out some of the argument’s follies. The paper is short enough and simple enough that most readers could breeze through it quite easily: All the author, Thomas Hungerford, did is linear regressions of various economic data against the top marginal income-tax rates and capital-gains rates in each year going back to the 1940s. In doing so, he found little or no evidence of positive economic effects: Most of the results were quite weak, though he did find some evidence, for instance, that higher tax rates reduce saving but increase investment. He finds no statistically significant connection between GDP growth per capita and the top marginal income-tax rate or the capital-gains rate. He did control for a small number of other factors, including stock-market returns and monetary policy (sort of — he used the change in yields on triple-A bonds over the course of a year. As just one example, there are a lot of ways monetary policy could shift without it being well reflected in that number.)

Essentially, lower tax rates don’t clearly produce better economic outcomes in a given year. It doesn’t take an economist (if you trust this non-economist) to realize that this is not a very insightful conclusion, and isn’t based on very impressive analysis. It is, simply, useless — really not much better than pointing out that we had higher tax rates and stronger economic growth in the 1990s than in the 2000s.

The paper doesn’t begin to prove whether or how lower marginal tax rates do or don’t lead to higher rates of investment or stronger economic growth, as classical models suggest they might. Answering that question, or related questions, in a way useful to policy makers and Congress requires a much deeper and more nuanced study than Hungerford has performed. For that reason, with which I think even economists who are quite skeptical of the positive effects of tax cuts would agree, CRS should probably never have bothered producing or publishing it in the first place. The service exists to provide Congress with reliable, insightful research that’s useful in forming legislative policy — something this paper does not do.

And, indeed, conservative critics were absolutely justified in pointing that out. Liberals have gleefully seized on the backlash to this paper as another example of Republicans’ being detached from reality and detesting data. That’s simply not what happened here. While the details aren’t perfectly clear, there’s nothing wrong with the CRS realizing the legitimate, serious criticisms of what the paper purported to be, and withdrawing it to preserve their reputation. Congressional Republicans raised some of these criticisms, but it seems that the problems with the study are certainly sufficient for this to have been a decision to which the CRS could acquiesce without it being an example of over-the-top political interference.

There are illegitimate criticisms, though, of work like this — some conservative outlets derided this study as worthless, or partisan, because its author is a Democratic donor. The political views of an author don’t blunt the study’s point; the study itself does a good job of that. Or to consider a more frequent issue, the fact that the Congressional Budget Office doesn’t dynamically score legislation only weakens the reliability of its reports inasmuch as one can assemble a coherent, respectable case for certain feedback effects; it doesn’t delegitimize or bias everything, or even most, of what the group does.

The Times compares the controversy to criticisms of the Tax Policy Center, a nonpartisan group that produced a controversial study about Mitt Romney’s tax plan, which did attract much hysterical political criticism, too. A couple weeks ago, the Times had a somewhat amusing report on their response (my emphasis):

At the Tax Policy Center itself, responses ranged from irritation at the partisan nature of some attacks to incredulity over the political hysteria. “There was this résumé-hunting, White-House-visitor-log” searching feel to the response, said the center’s director, Donald Marron, a former Bush administration economist. “That was unanticipated,” he added dryly.

Like the CRS report, some conservative objections to the TPC were well-founded, and relied on the facts of the situation, but some were not, and relied on completely unfounded accusations of liberal bias. Especially since the TPC relied on defensible and clear premises, the episode could provide some evidence for the Left’s oft-repeated accusation that conservatives or the House Republican caucus simply ignore data or believe it biased when it serves them poorly. In other circumstances, conservatives have definitely been guilty of this: Accusations that the Bureau of Labor Statistics, for instance, had manipulated the September jobs numbers, that official inflation statistics are inaccurate, are without any basis whatsoever, and, I believe, do reflect a problematic current in today’s conservatism, though far from a a mainstream one.

But there are plenty of legitimate criticisms, as in this case, of government data or scholarly output — and especially liberals who are the strongest advocates of government funding of researchers must accept that, sometimes, they will come under political attack for researching or publishing certain things. And since that might happen, they’re probably best off making sure that the papers are worth publishing in the first place.


Sign up for free NRO e-mails today:

Subscribe to National Review