It is still early to gather much public opinion that will help us understand next year’s midterms. In many races, candidates have not yet announced (or even decided) and at least one Senate incumbent has not even made clear whether he is retiring or not (Johnson R-WI). However, there are some indicators in public opinion that can give us a sense of what the political environment might look like in the 2022 midterm election. We’ll look at a couple of those indicators today.
Democratic Enthusiasm
Fewer Democrats are enthusiastic for the midterm election in 2022 in July than were in an earlier poll from April according to Morning Consult (B). The enthusiasm gap between Democrats and Republicans is narrowing. While more Democrats say they are generally enthusiastic than Republicans (and more Republicans say they are not enthusiastic) there are some troubling signs for the Democratic Party. A third of Republican voters say they are very enthusiastic compared to 24% of Democrats, a gap that is nearly the same when comparing Biden voters (23%) to Trump voters (31%).
Enthusiasm matters in midterms. Conventional wisdom says Republicans have held the edge in voter enthusiasm in midterms, but it is more accurate to say that the party out of power usually hold that edge. A very enthusiastic Republican voter base resulted in landslide defeats for Congressional Democrats in 2010 and 2014, which not only thwarted Obama’s ability to govern it is the reason that Merrick Garland never got a hearing for the Supreme Court seat the president nominated him for in early 2016. Democratic enthusiasm returned the favor in 2018, electing a Democratic House that finally held Trump accountable for at least one crime he committed as president.
One problem for Republicans concerning enthusiasm is the damage that has been done to its own base’s confidence in the electoral system. We saw this help elect two Democrats to the Senate from Georgia in January. Since then, attacks on American democracy and voting rights by Republican officials and pundits have only increased that lack of confidence. Republicans seem to be putting a lot of faith in their state-level voter suppression laws – which is one reason why Democrats should at the very least make a voting rights exception to the filibuster and pass both the For the People Act and the John Lewis Voting Rights Advancement Act. The most insidious parts of the GOP voter suppression strategy is putting discretion over voting decisions in the hands of partisan election officials and letting courts use a much lower bar to throwing out elections due to perceived voter fraud. That aside – and both of those things might not survive judicial review – the typical voter suppression tactics of making it harder to vote by mail or early are just as likely, perhaps more likely, to suppress Republican voters. Add this to another in a long line of “believing your own bullshit.”
Generic Ballot Polling
Last update was the Quinnipiac (A-) poll from the end of May. At that time, Democrats had a +9 point advantage (50-41). Since then, we’ve seen several B and C rated pollsters release findings on the generic ballot with results ranging from a +1 point advantage for Republicans on June 24th (McLaughlin, C/D rating) to a +7 advantage for Democrats on July 8th (Ipsos, B- rating). Three other pollsters released results in between with advantages for Democrats ranging from +2 to +4 points.
We are not seeing a lot of generic ballot polling right now, but that is not unexpected. The generic ballot question is not one that would interest many voters this far out from an election. We may see more pollsters ask this question in the fall, but it will become an important polling question next year. The generic ballot typically gives us some insight into which party is likely to control the House of Representatives, but as we saw in 2020 it is a mistake to think that the margin in the polling will match the margin in the actual vote. A better system would be to regularly poll all 435 Congressional districts, but the cost of that is prohibitive for any one pollster.
Senate Races
Data for Progress (B) released a poll this week that should concern Sen. Kyrsten Sinema (D-AZ). Two-thirds of Democratic primary voters in the Grand Canyon State say they would vote for another candidate if she continues to support the filibuster. Another 13% said they were not sure. That means just 21% of Democratic primary voters say they will vote for Sinema if she faces an opponent that opposes the filibuster. The good news for Sinema is that she is not up for reelection until 2024. It’s hard to see what her strategy is here, and her flirting with Republican donors will not inspire confidence among a Democratic base in Arizona that is already angry with her obstructing Biden’s agenda. She should note that she won a very thin victory in 2018 on the backs of hundreds of progressive activists working in that state (full disclosure: I was one of those activists, but only for the final GOTV push; the real hard work was done by local activists for over a year before the election).
In Florida, Rep. Val Demings (D), who was on Biden’s short list for vice president, is challenging Sen. Marco Rubio (R). An early poll has Rubio with a big lead (60-40). I have some big problems with this poll, which I’ll lay out below. Demings raised over $4.6 million for the race during the last quarterly reporting period, with an average donation of $26. In the same period just 14 incumbent Senators running in 2022 raised as much money as Demings did.
The Florida poll referenced above was conducted by a political consulting firm called Political Matrix and a marketing firm called The Listener Group. While some political consulting groups can do good polling, I am always skeptical when a marketing firm is involved. The combination of the two suggests that they might be engaging in push-polling, which is not actually polling (which seeks to measure public opinion) but an attempt to frame public opinion. Push-polling usually uses loaded questions like “if I told you that Candidate X supports amnesty for all illegal immigrants, including violent criminals, would you be more likely to vote for her or less likely?” This Florida poll does not use such questions. In fact, the questionnaire is pretty straightforward and legitimate. But there are important problems with this poll. First, the head-to-head question (Rubio vs. Demings) does not give the respondents a “not sure/other” response. While both of these candidates are their party’s likely nominees, each has actual and potential primary opponents. Even if they did not, it is poor practice to not give voters a chance to say “neither” – and we know that especially this far out from the election that some number of respondents would choose a third option, so the results are not as reliable as they could be.
The second problem is more serious. The pollsters only used landlines to contact voters. What this means is that they have had to significantly weight voters under 50. Polling only landlines will give us better information about older voters (since almost all landline calls these days go to older voters and thus, less weighting if any is needed). Since we know that older voters support Republicans in higher numbers than younger voters, it seems unlikely that Rubio has a 20 point lead with the general electorate when that is all he could get in a poll that appears to be designed to lean towards his base. The pollsters used IVR – robocalling – which cannot legally be used to reach mobile phones. This is an inexpensive method for pollsters and there is a good chance that the poll’s use of IVR-only calls was a financial one. However, that does not cure the underlying methodological challenge.
There will be more Senate race news coming up as the Democrats are doing a good job recruiting candidates and at the moment Republican recruits are demurring or staying quiet about their intentions.
Just for fun: Former Presidents who have run again
The Washington Post today made a reference to Trump “stockpile[ing] case for another potential White House run, an unprecedented maneuver for a former president.” One might think this means it is unprecedented for a former president to run for the office again. However, what’s unprecedented is the “stockpiling” of cash for another run. This is true, but not particularly interesting because the role of cash in candidate preparation is a relatively new phenomenon in American politics. Until the 1960s – and an argument can be made that it was not really until the 1980s – political parties were largely responsible for raising the cash for presidential (and Congressional) runs. In that time, there have only been three presidents before Trump who were eligible to run for another term after leaving office: Johnson, Ford, and Carter. Johnson, who died in 1973, had only one chance to run again and his ill health prevented him from even considering it. To my knowledge, Carter never considered running again. Ford, however, did consider running in 1980, forming an exploratory committee. Even though he never entered a primary, Ford was considered both as a replacement for Reagan and as his vice president at the Republican National Convention. Here is a list of former presidents who have run for president after leaving office.
Martin Van Buren (Free Soil Party nominee, 1848)
Millard Fillmore (American “Know Nothing” Party nominee, 1856)
Ulysses S. Grant (Republican candidate, 1880)
Grover Cleveland (Democratic nominee and only former president to win back the office, 1892)
Herbert Hoover (Republican candidate, 1940)
Theodore Roosevelt (Progressive Party nominee, 1912 - Roosevelt finished second behind Democrat Woodrow Wilson and ahead of incumbent Republican President William Howard Taft. The best showing ever for a minor party candidate in US presidential history.)
Gerald Ford (Republican Party candidate, 1980 – Ford came close to being Reagan’s VP and some convention delegates considered replacing Reagan with Ford as the nominee)
I wanted to include Andrew Johnson on this list because he tried unsuccessfully to get both the Republican and Democratic nominations for president in 1868. However, he was not a former president at the time. An accidental and impeached - and a widely disliked - president, but not yet a former president.
AAPOR Report
The polling post-mortem on the 2020 election from the American Association of Political Opinion Research has been released. I have an idea of what’s in it, but I plan to read the report before commenting on it. Hopefully, I’ll have the time to post about its findings next week before I disappear for a few weeks on vacation.