Correlation = Causation, right?

Paul Krugman produced this chart on annual rates of job growth under each post-war President since Truman. Although he doesn't actually elaborate, it's not hard to see what he's driving at: Democrats = jobs.

The problem, of course, is that this graph is very misleading. For starters, the President is not responsible for economy (see Tyler Cowen's excellent essay here), regardless of how the electorate chooses to vote. Certainly we wouldn't expect Krugman to suggest that FDR was responsible for the high levels of unemployment during his first two terms in office.

This graph also ignores the fact that early post-war Democrats were the champions of international trade and liberal markets while the Republicans were isolationist and protectionist. Those positions have now been (largely) reversed. How might this alter the conclusions reached by looking at this graph?

Maybe a more accurate picture could be gained by including House and Senate majorities, but the larger point is that the state of the economy is not decided at the voting booth. Krugman is doing his readers a disservice by suggesting otherwise.