That’s the question two Princeton economics professors set out to answer in a new study. The answer: The impressively large gap seems to have little to do with the presidents and policies themselves. There happen to have been fewer oil shocks under Democratic presidents, consumer confidence has been higher, and something called “total factor productivity” has been higher (it reflects the value the economy produces on top of the labor and capital inputs).
Those things, according to Alan Blinder (a former member of President Clinton’s Council of Economic Advisers) and Mark Watson, explain the majority of the difference between the average economic performances of Republican and Democratic presidents (which is huge: economic growth under Democrats has been 1.8 percentage points higher than under Republicans) – the rest, they decide, has to be due to other, unexplored factors or just random variation.
One interesting finding: Some suggest differences between the two parties’ defense spending might explain the gap, because wars sometimes force a whole lot of fiscal stimulus, and much more of that has happened under Democratic presidents than Republicans (coincidentally and perhaps surprisingly). But that, Blinder and Watson find, doesn’t explain much of the variation at all – by corollary, fiscal stimulus is perhaps not all it’s cracked up to be.
Differences in total factor productivity and consumer confidence can obviously be attributed to other complicated factors, but it’s hard to say how any of them can be tied to a Republican or Democratic president — for one, because politicians affect these things and these things affect the economy over wildly varying time frames.
Matt Yglesias of Slate has a go at how these ambiguous conclusions can be read easily to support Democratic policies, but his reasoning is a little slipshod. He attributes better consumer expectations under Democrats to Americans’ faith that Democratic policies will support the middle class — but the confidence index the authors used takes into account not just people’s future expectations about their own finances, but also future business conditions. The more confident expectations could be entirely due to Democrats’ giving consumers good feelings about their own situation, but it also could be entirely due to people’s thinking businesses will be coddled and the stock market will go up, which Yglesias’s theory of Democratic middle-class-friendliness wouldn’t encourage. (Though the authors didn’t split up the two kinds of expectations, it’s probably both: Democrats’ economic performance has been better because of more optimistic consumer and business decisions.)
Yglesias attributes the greater number of oil shocks under Republican presidents to their “more confrontational approach to international relations” — the causality here isn’t hard to check, because there are just a few. The authors use a 2003 paper that collated the major post-war oil shocks, plus the 2007–08 oil shock, and the ones that happened under Republican presidents weren’t because Republicans are mean:
‐“the oil price spike of 2007-2008″: this was caused by a global economic boom (under a Republican president!), China’s thirst, a weak dollar, international tensions you can’t blame on the Bush administration (viz., not the Iraq War) and various concerns about tightening supply
Blinder and Watson consider the idea that some of these shocks are not unrelated to Republican presidents’ being in power, misleadingly highlighting the 2007–08 spike as related to a Republican president’s invasion of Iraq — a reason that’s not among the many argued by the paper they actually cite on the topic.
Yglesias’s explanation of total factor productivity has little logic to it either: First of all, TFP is generally believed to be affected most by technology, something the government has almost no control over in the short term. Second, he concocts the idea that Republicans regulate in a pro-incumbent way, while Democrats’ cross-cutting loyalties (labor, environmentalists, business) mean they “do better at making decisions in the national interest.” I suppose that’s possible, but it wouldn’t explain why the decisions are better in the short term for the efficiency that goes into total factor productivity. The reality is that total factor productivity is not controlled by the government in the short term — and neither are many of the key factors that affect economic growth.
Blinder and Watson no doubt took on this paper because, as they explain in the introduction, “the performance gap [between Republican and Democratic presidents] is startlingly large — so large, in fact, that it strains credulity, given how little influence over the economy most economists (or the Constitution, for that matter) assign to the President of the United States.” It’s hard to believe this is a coincidence, but looking at the performance of other nations under Right and Left parties, the authors find no similar gaps, except for a small one in Canada.
As I mentioned above, it’s hard to imagine how such a large gap would be clearly explained by differences between Republican and Democratic economic policies (assuming those have stayed the same over the past half century) — even if Democratic policies really are so much better, why would it show up under Democratic presidents? Economic performance in Obama’s first term was the 14th-worst of the 16 post-war presidential terms, but Democrats don’t think that’s due to the policies he pursued — it’s either because Republicans prevented him from properly implementing liberal policy, or, more likely, because Ronald Reagan’s and George Bush’s awful policies created a massive financial crisis that’s led to a slow recovery. Economic effects don’t just lag, too, they also lead: Economies could perk up at the end of one party’s term because of the hope that better policies are coming.
The divergence in outcomes between performance under the two parties’ presidents is really striking, but some things just are, without any clear explanation. Yglesias admits that the coincidence explanation makes sense, but laments that “they don’t mount a very strong argument for that conclusion” that the difference is due to exogenous events and coincidences — but that’s not how this works. The burden of proof is on economists to show how Democratic presidents did something to produce that great growth. The authors couldn’t find one compelling way that happened so they laid out the few explanations they did find, and said the rest is unexplained.
The paper’s analysis is thus the kind of exercise that may yield some interesting reminders (e.g., oil shocks matter), but never a satisfying answer. This is an economics paper, after all, not a Matt Yglesias column. There are rules.