Alternative post title: We shouldn't care about whether we're properly calculating a biased estimator.
Andrew Sullivan counters that actually we are better off than we were four years ago if "four years ago" is interpreted to mean "the day Obama took office". Because if you count starting there the unemployment rate has inched down ever so modestly.
Mr. Sullivan, you have quite a podium. You should not be validating this decision rule. You should be explaining that the right question is "are we better off over the course of the last four years than we would have been without Obama". "Without Obama" is vague, I know. Appropriate counterfactuals might include if McCain-Palin had been running the show or if Romney-Ryan had been running the show, or some kind of current policy baseline from 2008 to 2012.