Significance

Several posts lately, notably from Kevin Drum, have purported to show that the American economy prospers more under Democratic Presidents. This conclusion is based on about 50 data points, so we should not expect much significance, but it has been eagerly reported.

It occurred to me that, since the US economy does not exist in a vacuum, we could take a different look at GDP growth numbers by comparing US growth to that in other countries.

For the purpose, I used this data: it reaches back only to 1960, so my results are still less significant than Mr. Drum's. I use growth in per-capita GDP, rather than absolute GDP, which should make very little difference in the variation.

Here is the subset of Mr. Drum's data, showing total GDP growth over 1948-2001 as a function of the party of the president a few years before:

GDP Growth	3 Yrs	4 Yrs	5 Yrs
Democrats Republicans	3.56% 3.35%	3.78% 3.16%	3.71% 3.21%

And here are the numbers I obtained, using GDP per capita over 1960-2003:

GDP Growth	3 Yrs	4 Yrs	5 Yrs
Democrats Republicans	2.06% 2.42%	2.43% 2.11%	2.46% 2.10%

The advantage is still generally to Democrats, though it is somewhat weaker than Mr. Drum's showing. Next, I subtracted per-capita GDP growth in France. Thus, in the following chart, positive numbers show a tendency for the US economy to outperform the French:

Growth Diff	3 Yrs	4 Yrs	5 Yrs
Democrats Republicans	-0.64% 0.08%	-0.28% -0.23%	-0.37% -0.17%

It could be argued that France differs from the US in important ways, such as its heavy use of nuclear power and its presently unfavorable demographics. I repeated the computation, using Sweden instead of France:

Growth Diff	3 Yrs	4 Yrs	5 Yrs
Democrats Republicans	-0.40% 0.47%	-0.03% 0.14%	-0.19% 0.25%

These were the only countries for which I made this computation; readers are invited to repeat or expand it.

Let me make it clear that this is more gamesmanship than serious statistical analysis, due to the comical paucity of the data set. I do not think these data provide meaningful support for the hypothesis that Republican presidents help the US economy outperform the rest of the world. However, if Democratic partisans want to use data like these to show that Democratic presidents are somehow better for the economy, they must explain how it is that a Democratic president in the US can help the French and Swedish economies (at least) even more than our own.

[Added to Beltway Traffic Jam. Also at Outside the Beltway, James Joyner looks at causes, while Steve Verdon has a somewhat complex argument about the impact of recessions.]