Presidents and Growth

Pretty much whichever way you slice the numbers, the data indicates that postwar economic performance has been stronger under Democratic presidents than it has been under Republican presidents. This is, obviously, suggestive. And folks inclined to prefer left-of-center economic policies are naturally inclined to suggest that it’s evidence that left-of-center economic policies produce better performance.

Today, Megan McArdle offers some oblique pushback against this. Her argument has two parts. One is that correlation does not prove causation, a point she illustrates by observing that the long-run trend of lemon imports is correlated with trends in car safety. But of course everyone knows that. The other is to demand knowledge of the mechnanism: “What policies did Harry Truman and Bill Clinton have in common (but not with Richard Nixon) that caused this marvelous confluence?”

I’m not really sure why we’re supposed to deem this so mysterious. Over the decades, the parties have consistently had different interest group bases. The Republican Party is more closely aligned with the economic desires of American business and high-income individuals whereas the Democratic Party is more closely aligned with the economic desires of organized labor and low-income individuals. The basic landscape has shifted over time—Nixon courted elements of labor in a way that most Republicans don’t, and Clinton certainly courted an image as a “pro-business” Democrat. But it’s still true that at any given time throughout this period that Republicans have been more oriented toward what business leaders and wealthy individuals want, whereas Democrats have been more oriented toward what low-income individuals and labor leaders want. Generically, this might lead you to expect that under Democratic leadership people of modest means would do well relative to people of greater means, whereas under Republican leadership the reverse would be the case. And, empirically, that seems to be the result. Democratic politicians, of course, would say that their approach isn’t just “soaking the rich” but will actually produce better overall performance. Republican politicians, meanwhile, will say that they’re not just enriching fat cats, they’re offering policies that will actually produce better overall performance. But when you look at it, it seems to be the case that performance is better under Democrats.

Does this “prove” that political leaders attuned to the interests of the upper classes produces worse policy outcomes? Well, not in the sense that there are proofs in mathematics or physics. But in public policy discussions you rarely get proofs of that caliber. And surely this does, in fact, count as a evidence for the proposition. Alternatively, we could say that since empirical proof in the social sciences is difficult to come by we ought to ignore the evidence and instead base all of our policy conclusions on economic models that assume the correctness of laissez faire policy. But I’m not sure I understand what makes that a superior method to looking at empirical correlations.