Traditionally, Labor Day is considered the time when general election campaigns really kick into gear. That’s not always true these days, and a few states — Delaware, New Hampshire, Rhode Island, and Louisiana — have yet to even hold primaries. But there is some genuine suspense as we head into the fall season, as Democrats are doing far better than expected. It’s a good time to look at some historical precedents and consider where things stand, and what we can expect in the last two months of the midterm cycle.
The generic ballot is a draw.
The two parties are basically tied in the most reliable indicator of midterm voting intentions, the generic congressional ballot, which asks voters which party they want to control the U.S. House of Representatives (it’s an approximation of the national House popular vote). Presently, Republicans lead by 0.2 percent in the RealClearPolitics averages, and Democrats lead by 0.9 percent in the FiveThirtyEight averages. This situation has actually been stable for a while: the margin between the two parties has been under one percent for over a month in both data sets.
So how much could this change in the coming weeks? Bill Scher at RealClearPolitics took a long look at late trends in the generic congressional ballot in recent midterms and found a common pattern, with a couple of exceptions:
Typically, over the course of a midterm election year, “the out-party gains in the polls and ultimately at the ballot box,” according to a 2010 political science research paper from three professors at different universities, based on data from the 16 prior midterm elections.
The RealClearPolitics averages in the 2010, 2014, and 2018 midterms largely bolster that analysis.
In 2010, Republicans took a generic ballot lead in July that steadily widened until Election Day. In 2018, Democrats led post to post, and their margin began to swell steadily in mid-August. But the trend lines for 2014, a midterm that has been often compared to 2022, should create some jitters for the currently ebullient Democrats:
Democrats clung to a small lead for most of the period between early March and early September, but Republicans seized a lead by late September that was not relinquished, largely holding between two and four percentage points. On Election Day, Republicans retained the House and flipped the Senate.
Republicans flipped the Senate hard in 2014, netting nine seats. But there is no predictable relationship between the national House popular vote (and the generic ballot that predicts it) and Senate results, which depend heavily on the specific “landscape” of seats at stake; only one-third of Senate seats are up in any election. This year’s Senate landscape is pretty balanced, and at the moment FiveThirtyEight gives Democrats two-to-one odds to hold onto the chamber.
Keeping the House is still a reach for Democrats.
There were two recent midterms, in 1998 and 2002, that broke the mold, with the president’s party making gains in the late generic congressional ballot indicator, and ultimately gaining House and Senate seats. In both cases, the president’s job approval rating (Clinton in 1998 and George W. Bush in 2002) was above 60 as voters went to the polls. While Joe Biden’s numbers have been slowly improving since July, he’s a long way from that level of popularity.
Many observers believe there is a distinct Democratic bias in generic congressional ballot numbers, so a tie in the polls might be misleading, in part because of a less efficient distribution of Democratic voters and a slight disadvantage due to redistricting. Democrats might need to over-perform in the House national popular vote in order to hang onto control. Political scientist Alan Abramowitz has estimated that Democrats might need as much as an eight-point lead in the generic congressional ballot to win a House majority in November. That may be a bit pessimistic, but there is no question Democrats under-performed the polls in 2020.
Republicans can’t count on the “switchover.”
In past elections, the polls themselves tended to shift toward Republicans after pollsters moved from samples based on registered voters to likely voters. This “switchover” tends to take greater account of the superior participation rate — particularly in lower-turnout midterm elections — of voting groups that lean Republican, while also factoring in the relative enthusiasm of each party’s base voters (typically the non-presidential party in midterms).
At present, quite a few pollsters are already using likely voter screens. And the pro-Republican effect of the “switchover” has probably abated as Democrats have recently improved their performance among older and suburban voters.
The abortion factor could be huge.
The backlash to the Supreme Court’s Dobbs decision may be the most important midterms variable. This summer, a series of special elections and a Kansas abortion referendum provided anecdotal evidence of a big rise in Democratic voter enthusiasm following the loss of Roe v. Wade. It’s unwise to read too much into special elections, but it seems renewed concern about reproductive rights, coupled with recent Democratic legislative wins, could have turned the tide.
Dobbs is a reminder that external events can shift election trends, even fairly late in the cycle. Democrats are still anxiously hoping for better economic news, though perceptions of the state of the economy tend to get baked into the public consciousness well before Election Day. And arguably, the “economic issues” that matter most in shaping political allegiances are long-term trends like globalization, wage stagnation, and falling living standards rather than last quarter’s GDP or this month’s jobs report. But certainly, if the economy produces tangibly less inflation without signs of slipping into recession, it will help voters focus on other issues that benefit Democrats, including the struggle to defend abortion rights.
Poor candidate quality could dash GOP Senate hopes.
The quality of individual candidates and the skill of their campaign operatives is, of course, an important variable — though in House races they probably matter less than the more fundamental factors shaping the size of the electorate and its partisan leanings. In Senate races, with the enormous attention they get in the really competitive states, the relative competence of candidates and campaigns are more likely to produce idiosyncratic results, which is why Republicans are so worried about the quality of their Trumpy Senate candidates in Arizona, Georgia, Ohio, and Pennsylvania.
The race may be a referendum on Trump, not just Biden.
Midterms are typically referenda on the sitting president, but there is one person whose outsized presence could change the usual equation: Donald Trump. If the former president announces his 2024 candidacy, and/or is indicted before November 8, it might at least partially make this a “choice” election. That could boost base turnout in both parties, but it’s more likely to benefit Democrats; suburban swing voters who supported Biden in 2020 but flirted with the GOP in a midterm dominated by economic jitters, fears of crime, and “woke” elites may be put off by Trump’s lingering presence.
The 2022 midterms could still take some twists and turns in the fall. Control of the Senate might not be resolved until another Georgia general election runoff, this one in December. Politics could stay hot even as the weather turns cold.
More on the Midterms
- Are Democrats the Party of Low-Turnout Elections Now?
- New Midterms Data Reveals Good News for Democrats in 2024
- The Return of the Emerging Democratic Majority?