Very interesting.

Finally, as economist Greg Mankiw points out in his blog, reacting to a similar calculation by Alan Blinder (both of them former chairs of the president's Council of Economic Advisers), correlation is not causation. Maybe economic statistics are better when the president is a Democrat for reasons having nothing to do with the president's skill in handling the economy. My own feeling about that is that as long as the pattern continues, who cares why? Correlation will do just fine.

As cool as that sounds, correlation isn't the best thing to base voting on. Don't get me wrong, Obama still has my vote but not because of statistics. You can't forget that during certain republican presidential terms, there were things that happened that were outside of their control. It's not just black and white.

However, it would be an interesting experiment to see whether those things matter in general. It's possible that the way Democrats do things is a step up from the way Republicans do, but I think each election (and the runners therein) is its own animal. That is to say, unless they honestly promise nothing significantly different from the last president.
All in all, good find. :]