Welcome to FiveThirtyEights politics chat. The transcript below has been lightly edited.
sarah (Sarah Frostenson, politics editor): Were still more than a year away from the 2022 midterm elections, which means it will be a while before we should take those general election polls too seriously. But with a number of elections underway in 2021, not to mention a number of special elections, its worth kicking off the conversation around what we do and dont know about Republicans and Democrats odds headed into the midterms.
Lets start big picture. The longstanding conventional wisdom is that midterm elections generally go well for the party thats not in the White House. Case in point: Since 1946, the presidents party has lost, on average, 27 House seats.
What are our initial thoughts? Is the starting assumption that Republicans should have a good year in 2022?
alex (Alex Samuels, politics reporter): Yes, and heres why: 2022 will be the first federal election after the House map(s) are redrawn. And because Democrats fell short of their 2020 expectations in state legislative races, Republicans have the opportunity to redraw congressional maps that are much more clearly in their favor. On top of that, Republicans are already campaigning on the cost and magnitude of President Bidens policy plans to inspire a backlash from voters.
geoffrey.skelley (Geoffrey Skelley, elections analyst): Simply put, as that chart above shows, the expectation is that Democrats, as the party in the White House, will lose seats in the House.
And because Democrats have such a narrow lead in that chamber, that would mean the GOP is favored to take it. Now, because only one-third of the Senate is up every two years, its not a truly national election the way the House is, so the story there is more complicated.
nrakich (Nathaniel Rakich, elections analyst): What they said!
alex: lol
That said, I think Democrats might find some success campaigning on Bidens accomplishments from his first 100 days: the vaccine rollout and the coronavirus stimulus funding, specifically. But that might not be enough to save Democrats from a midterm shellacking.
nrakich: Yeah, Democrats are obviously hoping they can buck the trend and point to exceptions like Republicans gaining seats in 2002 as evidence that its possible. But George W. Bush was very popular in 2002 in the aftermath of 9/11: According to a retrospective FiveThirtyEight average of polls at the time, he had a 62 percent approval rating and 29 percent disapproval rating on Election Day 2002.
And in this era of polarization where presidential approval ratings are stuck in a very narrow band its hard to imagine Biden ever reaching that level of popularity.
sarah: Were getting ahead of ourselves with the Senate, Geoffrey! Well talk about that more in a minute. But it sounds like our starting point is that 2022 should, in theory, favor Republicans?
That makes sense given the historical record, but to push back on that just a little there are instances when the presidents party didnt do that poorly. For instance, in 1990, when George H.W. Bush was president, Republicans lost eight seats in the House and one seat in the Senate which, as Politifact wrote, was a setback but not exactly a shellacking. Similarly, in 1998, when Bill Clinton was president, Democrats actually picked up five seats in the House and broke even in the Senate. Republicans still controlled both chambers, but the fact that Democrats didnt really lose ground was notable nonetheless.
So is it possible that Democrats wont have that bad of a year?
geoffrey.skelley: As the COVID-19 pandemic recedes, Biden and his team are clearly banking on an economic revival that will buoy his ratings and Democratic fortunes in the midterm.
After all, the times when the presidents party avoided losing much ground in the House or even gained a little came when the incumbent president was very popular. Its going to be hard, as Nathaniel said, for Biden to be at or above 60 percent approval when things are so polarized hes at about 54 percent right now, according to FiveThirtyEights tracker but if he can hang out above 50 percent, that could help Democrats minimize their losses in the House.
alex: And at this point, Democrats seem to be more excited than Republicans about voting in the midterms, per Morning Consult. Their April poll found that 81 percent of Democrats were at least somewhat enthusiastic about voting in the 2022 midterms, compared to 72 percent of Republicans. So not a huge difference, but still interesting.
sarah: What about midterm turnout more broadly? Is there any reason to think that Republicans or Democrats hold an advantage here?
geoffrey.skelley: As a general rule, midterm elections are influenced a lot by what political scientists call differential turnout; that is, your average member of the party thats not in the White House is more likely to turn out than the average member of the presidents party.
In other words, we can probably expect Republican turnout to be up compared to 2018, and Democratic turnout to be down. However, how much more or less is the real question.
nrakich: Some analysts point to the fact that college-educated white voters, who are pretty reliable midterm voters, used to vote Republican but now vote Democratic. And while its true that Democrats have made gains with these voters in recent elections, I think its overstating things to say that will turn midterms into Democratic-friendly environments.
According to a recent report by the Democratic data firm Catalist, college-educated white voters only voted for Biden 54 percent to 46 percent (based on the two-party vote). In other words, theyre still a swing demographic, not part of the Democratic base (yet).
sarah: That certainly seems to be the big question heading into 2022, Nathaniel.
geoffrey.skelley: We shouldnt discount the role persuasion plays in midterm elections, either. Catalists 2018 midterm analysis speaking of them noted that Democrats won over some voters in 2018 who leaned Republican in 2016.
So its possible that if conditions are relatively favorable for Democrats, that might persuade some voters to stick with them and turn out. Alternatively, the GOP might be able to win over some Biden voters if they dont feel good about the status quo.
sarah: What else should we be factoring in to understand the national environment?
nrakich: The two indicators I always look at are polls of the generic congressional ballot (in other words, polls that ask, Would you vote for a Democrat or Republican for Congress?) and special election results (specifically, how much they deviate from a districts base partisanship).
And right now, those indicators point to a neutral or slightly Democratic-leaning environment. We dont have a generic-ballot polling average yet, but the few polls we do have tend to put Democrats up by single digits.
My informal tracking of special election results so far about two dozen mostly legislative elections, so not a huge sample size shows that neither party is significantly overperforming its 2016 presidential performance.
But there is still plenty of time for the national environment to change. In the 2010 election cycle (which, of course, was a great one for Republicans), Republicans didnt take the lead in generic-ballot polling until December 2009.
sarah: But uh the generic ballot polls were really off in 2020. How should we factor that in when thinking about 2022?
As former FiveThirtyEighter Harry Enten argued back in April, we might read Democrats current lead in the generic ballot as a slight GOP advantage, given how much polls underestimated Republicans in 2020.
nrakich: Yeah, Sarah, I think its fair to mentally subtract a few points from Democrats on the generic ballot polls. That would put them right in line with the special-election results so far, which show a more neutral environment.
That said, the generic-ballot polls were spot on in 2018. Our average gave Democrats an 8.7-point lead on Election Day, and they won the national House popular vote by 8.6 points. And 2022, as a midterm year, has more in common with 2018 than 2020.
geoffrey.skelley: Thats right, generic ballot polls tend to be more accurate in midterm elections than in presidential ones.
FiveThirtyEights historical generic ballot polling average on Election Day vs. the actual national popular vote for the U.S. House of Representatives, 1996 to 2020
FiveThirtyEights polling averages are calculated retroactively for years prior to 2018.
Sources: The Cook Political Report, Polls, U.S. House of Representatives
That said, even if its a somewhat neutral environment in 2022 perhaps a best-case scenario for Democrats an evenly divided national popular vote would likely produce a GOP House majority. Why? Even though Biden won the national popular vote by about 4.5 points in 2020, the median House seat only went for him by 2.4 points. Heading into 2022, that bias may only grow, considering that Republicans will draw new congressional lines in a lot more states than Democrats.
alex: Nathaniel has done a great job writing about this, but I think the new voting laws (both the restrictive ones and the expansive ones) should be a factor we examine, too.
While its hard to determine whether these bills will have some sort of partisan impact, its also very possible that young voters and voters of color have a bigger incentive to turn out to vote because of these bills.
nrakich: Yeah, Alex, itll be really interesting to see whether those laws do what they are ostensibly intended to do and depress Democratic turnout or whether they make Democrats more fired up to vote.
sarah: What about the Senate? Republicans must defend more seats than Democrats in 2022, but the Senate is often a more complicated story. Remember, the House experienced a blue wave in 2018, but the Senate actually got redder. Any sense of what to expect this year?
alex: Im less clear on Republicans prospects for taking back the Senate, but I am more inclined to say Democrats can hold onto their narrow majority there.
A piece from The Washington Post in March made the case that less than a dozen seats are really in play, and of those, there are more opportunities for Democratic pickups. For example, the North Carolina and Pennsylvania seats (both previously held by Republicans) might be easier grabs next year since their 2020 margins were so close.
And it could be hard for Republicans to flip the four Democratic seats that are considered competitive Arizona, Georgia, New Hampshire and Nevada. Biden won New Hampshire by 7 points last year! Nevada is also becoming more reliably blue (although it didnt move that much in 2020).
All thats to say, Democrats may really need to go all-in defending just two seats: Georgia (Sen. Raphael Warnocks seat) and Arizona (Sen. Mark Kellys seat).
nrakich: Yeah, this is a big caveat to all the doom and gloom for Democrats. While there is a clear trend of the presidents party losing seats in the House, the pattern isnt as consistent for the Senate.
The presidential partys performance in midterm elections, 1946-2018
2018 change includes the special election result for North Carolinas 9th Congressional District
Seat change calculated by how many seats the presidential party gained or lost based on the number of seats it held on Election Day. Seat vacancies were assigned to the previous party. Party switches after an election were not included in the calculations.
Source: BROOKINGS, GREG GIROUX, MICHAEL DUBIN, U.S. HOUSE OF REPRESENTATIVES, U.S. SENATE, UVA CENTER FOR POLITICS, VOTEVIEW.ORG
And the Class III Senate map (the class of senators who will be up for election in 2022) is arguably the most favorable one for Democrats, in terms of presenting opportunities to flip Republican-held seats. There are two Republican-held seats on the ballot in states that Biden carried (Pennsylvania and Wisconsin), and no Democratic-held seats in states that Trump carried.
geoffrey.skelley: Yeah, the makeup of the Senate classes matters a great deal and which party is defending which seats. One reason the GOP gained seats in 2018 was because it was able to pick off Democrats in red states like Indiana, Missouri and North Dakota. So its possible Democrats could find gains in Pennsylvania and Wisconsin, or, if things really go their way, perhaps a state that Biden only lost narrowly like Florida or North Carolina.
But one problem for Democrats is that they dont have the same set of juicy targets the GOP did in 2018 with states like Indiana, Missouri and North Dakota clearly red states with Democratic senators. By contrast, Pennsylvania and Wisconsin only went for Biden by around 1 point, and there arent really any clearly blue states with Republican senators up in 2022 (in fact, Sen. Susan Collinss seat in Maine is the only other seat the GOP holds in a Democratic-leaning state, and she won reelection in 2020).
Democrats also hold an array of seats that wont be easy to defend, such as Arizona, Georgia, Nevada and New Hampshire. Its definitely plausible that Democrats successfully defend some of them, but defending all of them, or defending most of them while picking up Pennsylvania or Wisconsin, is a tall order.
sarah: Yeah, Democrats might not have their worst Senate map in 2022, but it will by no means be easy, and how they fare will have a lot to do with the national environment. And as we touched on earlier, Bidens overall approval rating will also make a big difference in Democrats midterm chances.
nrakich: Yeah, if the national environment is even a bit Republican-leaning, that could be enough to allow solid Republican recruits to flip even Nevada and New Hampshire. And then it wouldnt even matter if Democrats win Pennsylvania and Wisconsin.
One thing is for sure, though whichever party wins the Senate will have only a narrow majority, so I think were stuck in this era of moderates like Sens. Joe Manchin and Lisa Murkowski controlling every bills fate for at least a while longer.
sarah: Lets talk about big picture strategy, then, and where that leaves us moving forward. Its still early and far too easy to prescribe election narratives that arent grounded in anything, but one gambit the Republican Party seems to be making at this point is that attacking the Democratic Party for being too progressive or woke will help them win.
What do we make of that playbook headed into 2022? Likewise, as the party in charge, what are Democrats planning for?
alex: Im not sure if itll work, but there is a debate in political science right now about the extent to which race-based messaging reduces support for certain policy ideas. In fact, a recent study from Yale political scientists Micah English and Josh Kalla found that highlighting the benefits of progressive policies for racial minorities actually decreases support for them overall, and this was especially true for white respondents. To prove this, the pair conducted an online survey of six progressive policy ideas increasing the minimum wage to $15, forgiving $50,000 in student loan debt, affordable housing, the Green New Deal, Medicare for All, decriminalizing marijuana and erasing prior convictions and asked randomly assigned participants to read about them in either a neutral, race-based, class-based and race-plus-class frame. They found that the class framing was most successful in increasing support for policies across racial and political groups. So Im sure Republicans will try and use this to their advantage given their current emphasis on tackling cancel culture.
Read more:
Will 2022 Be A Good Year For Republicans? - FiveThirtyEight