ANALYSIS — Writing in The Washington Post recently, columnist and veteran political analyst Henry Olsen suggested that we may all be missing something important: President Donald Trump’s showing in head-to-head ballot tests against expected Democratic nominee Joe Biden is lagging his job approval numbers in recent RealClearPolitics polling averages.
After noting the strong relationship between presidential job approval and election outcomes, Olsen wrote: “Biden leads Trump by nearly six points, 48.3 to 42.4 percent, in the most recent RCP average. Trump’s approval rating in the RCP average was 46.0 percent on Wednesday morning. If Trump’s true vote share approximates that, he only trails Biden by about 2 points. If that happens on Election Day, Trump could once again win the Electoral College while losing the popular vote.”
That’s the secret stash of voters that could help the president win a second term, apparently — voters who now say (in the middle of a health care crisis) they approve of the president’s job performance but don’t also say they will vote for him in the fall.
All we have to do is manipulate the ballot test by inserting Trump’s job approval number where his percentage of the vote is.
I immediately looked at four random dates from this month and found Olsen was correct. On April 1, 6, 13 and 20, Trump’s job approval was between 45.2 percent and 47.4 percent, while his share of the ballot test ranged from 41.7 percent to 44.3 percent. On the four dates I checked, Trump’s job approval was higher than his ballot average by 3.1 points, 2.5 points, 3.5 points and 3.3 points.
If job approval is a strong predictor of the presidential election, then maybe Olsen is right.
But when I looked back into the recent past, the data showed something different — something that undercuts Olsen’s hypothesis.
I looked back at four random days in October 2019 (Oct. 4, 9, 16 and 27), November 2019 (Nov. 1, 10, 20, 28) and January/February 2020 (Jan. 5, 13, 23 and Feb. 3) at Trump’s job approval and ballot test aggregations and found nothing like the April results.
The Trump job approval and ballot test numbers were remarkably similar. Sometimes one was a fraction of a point higher, and sometimes the other was a point or a half a point higher. This trend held for all three months, making it completely different from the April data.
This brings us to an obvious question: Given that we are in the middle of a national health care crisis, wouldn’t it make sense not to draw sweeping conclusions about 2020 until things settle down?
Handicappers of old may not use all the snazzy models of today, but we always had one rule of thumb on which many of us relied: For incumbents, what you see is what you get.
I’ve watched thousands of races, and in all but a few, an incumbent trailing in a race can’t count on getting most of the undecided voters. If, after years in office, he or she hasn’t sold himself or herself to voters, then that incumbent is unlikely to win late-deciders or reelection.
With an extremely negative race expected, one more development should be noted. Among voters who don’t like either presidential candidate, Biden currently has a strong advantage over Trump. That’s a marked difference from 2016, when the 18 percent of the electorate that disliked both Trump and Hillary Clinton went for the GOP nominee by 17 points.
Obviously, Trump has scrambled our politics. It’s possible (even likely) that his supporters don’t participate in polls, which could be a factor if Trump overperforms on Election Day. Maybe Trump will turn out an army of previously silent whites without a college degree. Or maybe handicappers have overestimated Democratic enthusiasm, and voters of color don’t turn out for Biden.
So yes, there are scenarios under which Trump loses the popular vote again but squeezes out the narrowest of Electoral College wins. We all know it is possible, although unlikely.
I expect Trump’s job approval and ballot test numbers will come together again, if only when November nears. But I am more sure of this: Merely replacing Trump’s ballot test number with his job approval number in a presidential matchup in the middle of a national crisis strikes me as more of an effort to prove something rather than an effort to understand what is happening in our politics.