Trump might be smarter than you think

Philip Bump of the Washington Post wrote a piece published Saturday morning asking “Why was Donald Trump campaigning in Johnstown, Pennsylvania?” It’s an interesting question, given the obvious things pointed out in the article: Trump trails in Pennsylvania by 6 points”, and there are “currently 10 states where the contest is closer than Pennsylvania.” Swinging a state by 6 points is a big ask. Spending time in states where he is running much closer, like Ohio or North Carolina, seems like a much better strategy. (Both according to

But what if there is a scenario in which Trump isn’t wasting time in a state that seems clearly blue, certainly bluer than nearby Ohio? There is an alternative hypothesis to consider which would actually make the decision strategically quite smart.

I agree with Bump that for either Trump or Clinton to spend time and resources in the last days of the campaign in any state that is solidly in their corner would be a waste. Likewise, there would be no reason to spend time in any states that are solidly in their opponent’s corner. Where I disagree with Bump is in the assertion that it follows that Trump is wasting his time in Pennsylvania.

If the polls we are seeing right now are accurate, Trump has essentially no path to win 270 electoral college votes on November 8th. Trump must hope that the polls are wrong and not just randomly wrong but that there is a systematic polling error in Clinton’s favor. Now, it is very unlikely that such a widespread error exists, given the sheer volume of polling we have seen that has been consistently in Clinton’s favor. It would not, however, be an impossible scenario. Some, including Trump’s campaign manager, Kellyanne Conway, have suggested that polls might be underrepresenting support for Trump because some voters might be embarrassed to say they support him. Alan Dershowitz, as I mentioned in a previous blog about polling and odds-making, declared in the Boston Globe in September “that polling is incapable of accurately predicting the outcome of elections like this one, where so many voters are angry, resentful, emotional, negative, and frightened.”

So let’s just say that it is not impossible for the polls to be systematically wrong. Let’s also stipulate that the Trump campaign knows it can only win November 8th if the polls are wrong, that really all hope is lost if this polling error doesn’t exist. Now Pennsylvania looks like exactly where he ought to be campaigning. If the polls are consistently off by, say, 4% to 6%, then Pennsylvania becomes a toss-up, and Ohio, North Carolina and Florida go solidly in his corner. In the states that look close now, Stealth-Trump would actually be in the lead. And it would be a mistake for him to spend time in the states in which he’s (secretly) winning. As Bump points out, he should be spending time in states where he’s “close,” which would be exactly a state like Pennsylvania based on this scenario of a systematic polling error.

Bump expressed incredulity that Trump could outsmart the polls: “Perhaps Trump has internal polling that shows him closer in the state. Possible! …. Is it likely? Not particularly — especially given Trump’s antipathy to polling over the course of the campaign.” That might be true but you don’t need internal polling to understand that if the public polls are right Trump has no hope of winning anyway. If Trump needs an alternate reality to win — like a 4-to-6-point systematic polling error — the best choice would be to assume that error exists and then bet his resources where the polls have him behind by about that amount. That’s Pennsylvania.

Trump might just be smarter than he’s been given credit for.

Author of Thinking in Bets and How to Decide. Co-founder of The Alliance for Decision Education

Author of Thinking in Bets and How to Decide. Co-founder of The Alliance for Decision Education