The Underpants Gnome and Maier’s Law

My series looking at a lot of different factors (from the economy to abortion) moved under Presidential administrations beginning with Ike has drawn a lot of criticism in private correspondence, comments on my posts, and even on the occasional post by other bloggers. The problem, its seems, is that on most of these factors, the American public seems to fare better, on average, during Democratic Administrations than under Republican administrations. That is not to say its been true on all series for all Democratic Administrations, but there does seem to be a difference – even now that I’ve for the most part stopped mentioning party affiliation at all (the Ex-GF suggested leaving it as the elephant in the living room, so to speak), simply putting up a graph that shows the performance of Presidents next to each other seems to be suggestive enough to be offensive or bad form.

So this post is about the criticism. And I note I’ve had a lot of time to think about it… its been going on for a while, although in fairness, the criticism has evolved. In the beginning, the criticism was that I had to be cherry-picking series (I started, as I recall, with real GDP per capita!) and that if I just went a bit wider, the discrepancy would disappear. It didn’t. Over time, the criticism evolved into a collection of excuses for the observed discrepancy. Regular readers know I went through a whole series of posts looking at each of the excuses – Congress did it, its due to lags, Republicans inherit worse economies, the Fed did it, going back further (to FDR) kills it, Republicans simply get elected when the public senses times are about to get bad, the results don’t hold up using textbook statistical methods, etc. None of the excuses held up. But I checked – every single one of them. This took a lot of time and effort. There have been a few more minor evolutions, but as this series gets more and more complete, and as I begin trying to modify them (with a co-author who currently wishes to remain nameless) in order to put them in book form, the criticism has taken on a new form. I want to focus on that new form.

So let’s begin with an example – this post by Megan McArdle is representative. In particular,

Cactus defends the notion that, once again, electing any president who is a Democrat has special, magical effects on desirable economic variables such as the amount of equity people have in their homes. The problem with this, as with most of Cactus’ other demonstrations of amazing Democratic presidential economic power, is that it gives us an Underpants Gnome theory of economic good:

1) Elect anyone who is a Democrat to the office of president

2) ???

3) Awesome economy!

(Apologies – I don’t know how to get blogger to put the gray background on the numbered items as she did it.)

Underpants Gnome theory of economic good… To quote Samuel L. Jackson’s character in Pulp Fiction, “Well, allow me to retort.”

First, there’s a problem with the premise. When did I ever pronounce the economy under Democratic administrations to be awesome? My position – that Democrats are marginally less offensive than Republicans – has never been a secret. Let me try an analogy…. almost every morning, I go for a run – well, I jog for a while, and walk for a while. I can’t jog too long or my knees will act up. Other folks in the neighborhood jog too. Now, anyone who observed the comings and goings in the neighborhood could probably figure out who they should bet on if we all participated in a 5K. Pointing out that the lanky 28 year old who runs with a dog could beat the pants out of the rest of us who amble about in the morning is not the equivalent of saying he’s an awesome runner.

To extend the analogy… I think a couple of the administrations did some serious training… JFK/LBJ brought in the technocrats, and Clinton was famous for the late night meetings in which policy details were hashed out and pizza was consumed. But GW has a well-deserved reputation for being on vacation – even when he’s not on vacation, when something happens we find out later he has to be informed later because he was out mountain biking at the time. Ike had a reputation of being out to golf, and Reagan would be the vacation leader in a world where GW didn’t exist. Contrast that with Carter, the (on most series) worst performing Democrat – he was derided as a micromanager who went so far as to determine who got to use the tennis court and when. In general, Presidents from Party seem to have taken a hands-on approach, and Presidents from the other have publicly stated that government is the problem – when they weren’t trying to dismantle it, they were leaving it alone to run itself. Returning to the running analogy – one group of weekend warriors tries to jog every morning, the other considers eating donuts Monday through Friday and some time spent on the golf course on the weekend to be adequate training for the LA Marathon. And Megan McArdle thinks there’s something wrong with anyone would bet on the first group to do better, on average, than the second group.

So much for the awesome economy, but this may go some ways toward explaining the “2) ???” as well.

Which brings up another point… I’ve tried to explain the differences before. As noted… I looked at every excuse and explanation for the difference that readers suggested, and tried to see if it fit the data. I came up with ideas of my own – I noted that the change in real M2 per capita precedes the change in real GDP per capita, and the correlation between the two series is very, very high. But even the Fed as an answer doesn’t work… the Fed’s policy toward Republicans (with the notable exception of its policy toward those named George Herbert Walker Bush) was more growth oriented than the Fed’s policy toward Democratic administrations. So what is left? Policy? Luck? Anything else?

Well, maybe I’m fudging the data. In my posts, I link directly to the data… I believe in all but one of them, my source for the data was the government agency most closely associated with that particular set of data. I make it quite easy to check, and to call me on fudged data. And every so often, a reader points out a mistake, and I put up a correction. But its happening less, fortunately – as I’ve automated the process, its harder to screw up.

I could go on, but I want to cover one more thing… even if I made no attempt to explain the difference, I think the series has value. I think there is some value in simply understanding what is going on, even if one doesn’t know why. One can’t come up with an adequate theory if one is simply way wrong about what happened. Getting the facts more or less right is important to coming up with a useful theory to explain them. Perhaps simply compiling the facts and organizing them into a visual, easy to understand way would be enough (assuming that really was all that I was trying to do).

But I want to come back to where I started, and to turn it around. In the end, I think the problem is with the results. I think the Megan McArdles of this world wouldn’t have much of a problem if all these series showed a random mix of Presidents, or showed the Republicans coming out on top. They are looking for confirmation of their beliefs, of that theory that sprang into existence and moves through the world completely untouched by facts – that government really doesn’t matter except to drag down the economy, or at least, that those who talk up small government are the ones who govern best. And there is intense frustration at seeing their views not only not confirmed, but also contradicted.

So as (apparently) the leading proponent of the “Underpants Gnome theory of economic good” – I want to at least point out that, even as Megan McArdle defines it, that theory makes more sense than following Maier’s Law.

update.. a few grammatical corrections were made.