As Krugman says, this doesn't prove increased government spending has been better for countries over the last few years, but it certainly doesn't suggest the opposite.
Since everybody thinks they're really smart when they say "correlation does not imply causation," usually without quite knowing what it means, people tend to shy away from just doing simple regressions and instead do something worse... pick an example and tell a story. That isn't even data, that's just anecdote!
