With the great availability of horse race numbers at the Presidential level, and the proliferation of state and district polls of varying quality and partisan lean to shift through, one of the more basic polling measures – Presidential approval – is frequently ignored. In spite of that fact, it may actually be the simplest way to explain why Joe Biden is the favorite to win the election – and by a substantial margin, if you trust either LeanTossup or DDHQ/Optimus forecasting, both of which get to the same macro place even if we disagree on the margins.
On November 5th 2018, the Monday before the midterms, Donald Trump had a -9.5% net approval rating according to 538’s approval tracker amongst Registered and Likely Voters. The next day, the GOP would lose 40 House seats, 7 governorships, and lose the House popular vote by 8.6%. That loss in and of itself does not mean a lot, as Bill Clinton and Barack Obama suffered similar fates in their first midterms on route to convincing re-elections. Hell, neither Ronald Reagan nor Richard Nixon ever had Republican Houses to lose in their first midterms, but they both suffered losses two years in and won 49 states at their next trip to the polls. The problem is, their approval ratings went up after the midterms.
On August 19th, 2020, Trump’s Registered and Likely Voter net approval, per that same 538 aggregate, stands at -11.9%. The approval polling is worse now than it was the Monday of the Midterms, and this is a large amount of the reason why Donald Trump is stuck in the mud behind Joe Biden. Another way of saying that is that 42.8% of registered and likely voters will vote for Joe Biden, per 538’s polling averages, and 42.2% of them say they approve of the job he is doing. It is highly unlikely that there will be some grand recovery in his vote share numbers without recovery in his approvals, given the love him or hate him nature of the Trump Presidency.
Election forecasts are designed to do two things, fundamentally – to forecast votes in elections, and to tell you who is going to win. The evidence that national approval ratings would do much to tell me how the President is going to do in Texas or Iowa is contestable, and the DDHQ/Optimus model does a good job of showing state by state forecasts in a way that shows why the President is in trouble. Their three state level Tossups – Arizona, Ohio, and North Carolina – are all states which the President needs and the Democrats very much want. That informs the 85.5% chance they currently give Joe Biden to win, but even beyond that sophisticated analysis, it’s more simple than that. Donald Trump is much less popular than any of the Presidents who have gotten a second term in the last 30 years. The only one in that time not to win again had approvals in the 30s, which is where his vote share ended up (albeit with 18% going to Perot). Trump may pull this election back, but he will not raise his vote share significantly without a boost in approvals. How he could do that is for others to debate, but sometimes it is that simple – 42% of voters approve of his job performance, 43% of them will vote for him, and that means he has something like a 14.1% chance of winning on November 3rd.
Evan Scrimshaw (@EScrimshaw) is Managing Editor and Head Of Content at LeanTossup.ca and a contributor to Decision Desk HQ.