Krugman seemed to be selling this as the researchers deliberately avoiding/ignoring data that didn't fit their hypothesis, and politicians taking the study seriously only because it suggests policies they approve of.

Why it's being sold with a "LOL math error" soundbite, I'm not sure.