In a recent podcast, Weekly Standard editor Bill Kristol made the observation:

if Obamacare had worked, Hillary Clinton would be the president-elect, right?  I do think those those premiums going up two weeks before the election probably moved the necessary number of votes.

That got me to thinking how one might prove this empirically. The short answer is that I cannot prove this proposition beyond a shadow of a doubt, but the evidence is stronger than one might suppose.

To be sure, national exit polls showed that the most important issues in this campaign were the economy (52%), terrorism (18%), immigration (13%) and foreign policy (13%). However, according to David Wasserman’s definitive vote tallies at Cook Political Report, a swing of only 38,595 votes in 3 states (MI, PA, WI) would have given the election to Hillary Clinton. This represents 0.0293% of votes cast, making it the 7th closest presidential election in history in terms of this metric (p. 330). The issue for our purposes here is whether we could make a good case that Obamacare was the factor that led these crucial election-deciding voters to pull the lever for Trump rather than Clinton.

. . .