[My Preferred Candidate] Would Have Won

It might be 2017 but we still hear a lot of speculating about the Presidential Election last fall, especially who could have won were the circumstances different. I remain unconvinced that anyone other than Trump or Clinton would have–they won their primaries by double digits, after all.

Those who argue that candidate X would have won instead should elicit great skepticism. It’s an enticing idea, but a dangerously reductive one. In any election, especially one as high profile as a Presidential one, it’s unrealistic to think that swapping candidates would exclusively gain votes. Even if a different candidate would have won more votes in some place or with a particular demographic group, they can just as easily lose votes somewhere else.

I’ll give the specific examples of Florida & Pennsylvania, two states that Hillary Clinton lost by 1% or less, and would have changed the outcome of the election had she carried them. In Florida, there’s a swing part of the state known as the I4 corridor. Typically we’d expect the person who wins this part of the state to win overall, as it has correlated with the winner in the last five elections. Clinton won it by a solid 6%, but it didn’t put her over the top because of a larger than expected number of Republican votes in north Florida (a Republican stronghold). Now let’s say we swap in Biden, who is the Democratic nominee in an alternate situation. Biden keeps the Republican margins down slightly in north Florida, but fails to turn out enough Democratic-leaning voters in the central Florida/I4 area and loses the state by 1%. The same thing could have occurred in Pennsylvania. Hillary Clinton won the Philadelphia suburbs, a key swing area, but lost statewide due to large Republican margins in rural Pennsylvania. Swap in Biden, and maybe he does slightly better in the rural areas, but doesn’t attract moderate voters in the suburbs, and loses just the same. You can repeat this to infinity with any candidate, state, or geographic region, but the reality is that changing candidates can, and probably will, lose almost as many votes as they might gain.

And it’s not just about geography or demographics, there’s a large number of other key variables at play too. Fundraising, campaign infrastructure, and experienced staff all play a crucial role, and there’s no way to prove that swapping candidates doesn’t increase weakness in any of those factors, some or all of which could be decisive in a close race.

Ultimately, this speculation is irrelevant. The election is over, and there will never be another one like it (I hope, for the sake of our sanity). The hardest thing to accept about Presidential elections is that the sample size is very small, and each is subject to a unique set of macro and micro variables, some of which have never appeared before and will never appear again. Trying to re-litigate the last one distracts from the present, and if you haven’t heard, 2020 has already begun.