By Kent R. Kroeger (February 28, 2020)
If you are a Democrat, you may want to sit down before you read this.
According to RealClearPolitics.com’s poll average, President Donald Trump’s current job approval (46.4%) is the highest it has ever been during his presidency (see Figure 1 below).
Figure 1: President Trump’s Job Approval (RCP poll average)
And while Trump’s job approval remains under water (more U.S. adults disapprove of his job performance than approve), he has crossed a threshold (46%) that represents the minimum level needed to win re-election, according to one prominent econometric prediction model.
It should be noted, however, that no incumbent president in the modern polling era has been re-elected with approval below 50 percent on Election Day. Nonetheless, a number of other prediction models are corroborating the importance of the 46%-threshold, including the PredictIt.com futures market (see Figure 2).
Since mid-January, around the time of the U.S. Senate’s acquittal of Trump on the impeachment charges, the PredictIt.com futures market has pegged Trump’s probability of re-election to be higher than 50 percent. As of today (27 February), that probability stands at 56 percent.
Though that represents a competitive presidential election in November — the Democrats are hardly out of it — the trend has been on Trump’s side since October 2019.
It was only a few months ago that I was writing about Trump being too far under water in job approval to get re-elected. I put his re-election chances at effectively zero.
That judgement was based on Trump’s job approval having the lowest variability of any president in the polling era, according to the Gallup Poll. In other words, his job approval numbers are hard to move, and that is a problem if a president with low approval wants to get re-elected. Since Trump’s all-time low in approval in December 2017, on average, it takes him about 3 months to gain one percentage-point in approval.
In late 2019, I didn’t think there was enough time to get him near 50 percent approval and thereby ensure his re-election. It looks like I was wrong.
Figure 3: Key Movements in President Trump’s Job Approval (RCP poll average)
But something changed in the fourth quarter of 2019 and first quarter of 2020.
Up to this recent period, Trump has had low approval for his entire presidency, ranging from the current high of 46 to a low of 37 (around the time of the Charlottesville, Virginia Unite the Right Rally in August 2017).
Furthermore, when you break up Trump’s presidency into discrete time frames, you can summarize most of the approval variation in just three periods: (1) Start of the Russiagate Investigation & the Charlottesville March (Jan. to Nov. 2017), (2) North Korea Crisis/Summit (Dec. 2017 to May 2018), and (3) the Impeachment Hearings and Trial (Sept. 2019 to present). I also identify a fourth period where there was no sustained trend in Trump’s approval (“The Period of Relative Quiescence”).
My presumption is that Trump’s recent rise in job approval (from 41.6 percent in late Oct. 2019 to 46.4 percent at present) is directly attributable to the impeachment hearings and trial (a rally effect similar to what Bill Clinton experienced during his impeachment).
That is my assertion, at least. In point of fact, there are many possible explanations that could also explain Trump’s recent rise in approval. Among them are: (1) a growing economy, (2) the particularly brutal Democratic debates that began around the time Trump’s approval rise, and (3) then there is the ascension of Bernie Sanders as the Democratic Party front-runner.
I am willing to dismiss the economy as a primary explanation, given that nothing unusual distinguishes the U.S. economic growth in the last two quarters from prior quarters. That said, research has consistently shown that presidential job approval does move with changes in the economy and, at some level, Trump’s current job approval must be partially a function of the strong economy.
Yet, its the other hypotheses — the savage Democratic nomination debates and Bernie’s rise into the lead — that give me pause when asserting Trump’s impeachment may have backfired on the Democrats.
As seen in Figure 4, Bernie Sanders began to surge in popularity in early December 2019 and, by early February 2020, has put significant distance between himself and the second place candidate, former Vice President Joe Biden.
Figure 4: The Democratic Nomination Race (RCP poll averages)
In a February 2020 Vox.com article, political scientists David Broockman and Joshua Kalla make a compelling data argument that the nomination of Bernie Sanders could drive a significant percentage of swing voters to Trump. Using data from a 40,000-person survey fielded in early 2020, Broockman and Kalla show that “nominating Sanders would drive many Americans who would otherwise vote for a moderate Democrat to vote for Trump, especially otherwise Trump-skeptical Republicans.”
Could this finding also explain why a higher percentage of U.S. adults are now approving of Trump’s job performance? Are swing voters giving Trump a second look now that Sanders is the probable Democratic nominee? Without individual-level panel data at my disposal, I would be speculating if I said ‘yes,’ but I am skeptical that Bernie is driving people into Trump’s arms.
Countering my skepticism, Broockman and Kalla offer evidence that Sanders’ strong head-to-head performance versus Trump in the national polls depends heavily on a strong turnout among young voters in November. They write:
“Sanders’s electability case requires this 11 percentage point turnout increase among young voters in 2020 to occur on top of any turnout increase that would otherwise occur if another Democrat were nominated.
If the turnout of all age groups increases from 2016 to 2020 (as happened from 2014 to 2018), then the turnout among young people must increase by 11 percentage points above and beyond this broader trend, and must do so solely due to Sanders’s presence on the ticket. Finally, youth voter turnout doesn’t usually go up or down by nearly as much as 11 percentage points from election to election; the Sanders boost would have to be truly unprecedented.”
I’m almost convinced. But I hesitate, because in the aggregate, there has been no significant change in Bernie’s head-to-head advantage over Trump in the past three months, despite Trump’s significant increase in job approval (see Figure 5 below).
It is hard to explain the variation in one variable (Trump approval) using a variable showing little variation (Sanders advantage over Trump in head-to-head polls).
Figure 5: Donald Trump versus Bernie Sanders (poll averages from Dec. 2019 to present)
It is possible Broockman and Kalla are correct that head-to-head polls overstate Sanders’ advantage over Trump (due to overly optimistic youth turnout assumptions); while, at the same time, other factors are still required to explain Trump’s recent rise in approval.
Data from one cross-sectional survey (no matter how well designed by Broockman and Kalla) is not enough to convince me that the rise of Bernie Sanders is helping get Donald Trump re-elected. Changes in Trump support over time, by definition, require dynamic models in order to be used for prediction.
Until I see that (time-series panel) data, I’m not drawing any firm conclusions.
Send comments to: firstname.lastname@example.org