In a previous post, we analyzed how changes in turnout might help Democrats win in various battleground states. Meanwhile, Trump’s campaign is doing the same math. As Trump has fallen in polls of battleground states, it’s becoming increasingly clear that his re-elections strategy is to “rally the base.” But can that strategy actually work?
Short answer: no. Trump might win a few battleground states by boosting turnout among his most die-hard supporters, but he’d still lose the rest, as well as the election overall.
According to the 270toWin polling average, as of 8/7/2020, Biden leads in the polls in several battleground states. Since we’ve spoken so much about demographically specific turnout boosts , e.g., GOTV work among college-educated or young voters, we thought it would be interesting to look at that strategy from the GOP side. Whose turnout helps Trump? The only group that was strongly pro-Trump in 2016 was White, non-Latinx voters without a 4-year college-degree (and not in-college), i.e., the so-called WWC. Polling so far this cycle shows their support for Trump is waning, both among female WWC voters and younger WWC voters, but overall, WWC voters are still, on average, Trump voters. This leads to two possible options for the Republicans:
Focus on WWC: The WWC is a big fraction of the US population, and is particularly concentrated in a few key states. So one plausible Trump campaign strategy in the battleground states might be to work at boosting turnout among all WWC voters.
Focus on the “Base”: Another strategy, which seems closer to what is actually happening right now, is to try to boost turnout only among “the base”–the portion of the WWC that voted for Trump in 2016, and has stuck by Trump since then.
Could either of these approaches lead Trump to victory in 2020?
We’ll look at the results shortly, but first a few quick notes about how we analyzed the problem:
In each battleground state where Biden currently has a lead, we show the turnout boost that Trump would need in either the WWC overall or his “base” to close the current statewide gap. The boost he’d need from his base is always smaller because, as we discussed earlier, that group has a higher preference for Trump than the WWC overall.
| State | Projected current D lead (% of Total) | Trump boost needed to tie (%WWC) | Trump boost needed to tie (% of "base") | Electoral Votes |
|---|---|---|---|---|
| AZ | 5.0 | 17.0 | 6.7 | 11 |
| FL | 4.0 | 20.2 | 7.9 | 29 |
| MI | 8.0 | 51.1 | 13.2 | 16 |
| NC | 1.0 | 3.3 | 1.7 | 15 |
| NV | 4.0 | 19.6 | 6.7 | 6 |
| PA | 6.0 | 23.0 | 9.1 | 20 |
| WI | 8.0 | 40.0 | 12.4 | 10 |
For context, typical election-to-election swings in WWC turnout are well under 5%. 10% turnout surges are virtually unheard of in recent electoral history. This paper breaks this down for the four presidential elections between 2004 and 2016, showing almost no shifts in battleground state voting for White non-Latinx voters with a high school diploma or less and 5% or less variability in the “Some college” group.
Thus, we make three observations from our analysis:
If Trump focuses on the WWC broadly, NC is winnable, but nothing else.
If Trump focuses on his hard-core “base,” none of these states is winnable under realistic assumptions based on historical turnout variability.
If Trump focuses on his “base” and somehow manages an unprecedented 10% turnout boost there, he could plausibly also win AZ, FL, NV and PA. But, as noted, the odds of this happening from a change in turnout are negligible.
If we do the electoral college math, Biden wins the election unless the Trump campaign pulls off that unprecedented 10% turnout boost in the base and everything else breaks his way. First, let’s make a conservative assumption that Biden loses all the currently close states where he’s trailing or tied right now (AK, AR, GA, IA, OH, TX). (In reality, we believe Biden could win most of these.) Under that assumption, not counting the states on the chart above, Biden would have 226 Electoral votes (EVs) and Trump 204 of the 270 EVs needed to win. If Biden wins all the states in the chart, he would end up winning the election with 334 EVs. If Trump gets a 5% turnout boost in the WWC and/or his “base” in all of those states, he wins NC, but Biden still wins the election with 319 EVs. In the black swan scenario in which Trump increases base turnout by 10%, he also takes AZ, FL, NV, and PA, and wins, though it’s important to note how unprecedented such a turnout shift would be.
So, our analysis of this table is that Trump can’t plausibly win the election just by focusing on turnout of either the WWC or his most die-hard base. However, there are some important caveats:
These calculations assume that the rest of the electorate behaves as expected. If Republicans simultaneously engage in voter suppression and other shenanigans to reduce Dem turnout at the same time that they focus on their core voters, then that could deliver the election for Trump. Thus it is especially important that we fight to preserve voters’ rights in the run-up to November.
These polls can and will shift between now and election day, so no one ought to be complacent. Progressives and Dems should work as hard as they can, to insure winning the presidential election and as many down-ticket races as possible. Fighting hard in places like GA, which may not be necessary to take the White House but where 2 senate seats are up for grabs, could well lead to Dems taking control of the Senate.
The polls themselves are weighted by some reasonable calculation of who is likely to vote. So whatever change in turnout the GOP needs, it must happen on top of whatever those voters are saying now. That makes the big shifts discussed above even less likely.
One response to this is to argue that Trump voters are “shy” when responding to polls and thus polls under count Trump voters. There is no evidence of this. The shock of 2016 was produced by last-minute changes which polling averages were slow to pick up and some issues around weighting the electorate for education, the latter of which have been addressed by the better pollsters since.
Dems could get complacent but we shouldn’t. Just as with recent modeling based on fundamentals and polls, you could make a case that Dems should be confident of winning the presidential election. We hope that’s not what you take away from this. Trump’s campaign sees the same data and will look to rally the base, the WWC at large, change minds and suppress Dem votes. And there are Senate and House races to win as well, many of them being fought in the same places that matter for the presidential election.
This analysis can explain Trump’s current campaign strategy: overt racist appeals to his base, using various attacks as wedge issues to enhance WWC support and win over suburban white college graduates, and fighting to suppress Vote-At-Home/Vote-By-Mail, which Trump perceives, incorrectly, to favor Dem voters.
So, we think Dems need be especially focused on turning out their base and combating voter suppression, particularly in FL, PA, and MI. Trump must hold his slim leads in OH, TX, and GA, win all the closer states in the table above and win 1 out of 3 of FL, PA or MI. If he loses any of the other states in our list, MI would no longer be enough for Trump and he’d have to win FL or PA.
But FL, PA and MI currently look out of reach via traditional tactics. If these polling margins hold up as the election gets closer, the primary remaining strategy the Trump campaign will have is voter suppression. COVID-19 creates various problems for local election officials, and the Republicans at the national and local level will not hesitate to use them to try and make voting more difficult in Democrat supporting regions, particularly cities. Expect demographically targeted social media, playing up the unreliability of Vote-By-Mail and the danger of voting in-person. Expect various targeted efforts to close polling places and/or make them operate more slowly in Dem heavy areas, forcing voters to wait in long lines.
With these things in mind, we recommend giving your time and/or money to the following organizations:
Support All Voting is Local, a non-profit focused on voting-rights and enfranchisement in many of the same states we are highlighting above, as well as GA, which is important this cycle as well.
Support the Brennan Center, working to improve voting systems, fight voter suppression and restore voting rights to formerly incarcerated citizens.