In an election that Newt Gingrich understandably says produced “a surprisingly confusing outcome” and was “the least predictable election in my lifetime,” voters actually conveyed a lot of important information to both major political parties. For starters, they made it clear that they aren’t in love with either of them. According to exit polling done by Edison Research for ABC, CBS, CNN, and NBC, voters gave Republicans a net favorability rating of -8 percentage points (44 percent favorable, 52 percent unfavorable) and Democrats an almost identical net favorability rating of -9 percentage points (44 percent favorable, 53 percent unfavorable).

Midterm elections normally serve as referendums on incumbent presidents—to whom voters seldom give the benefit of the doubt, and whose parties almost always lose substantial numbers of seats in the House of Representatives and often the Senate, as well. To the surprise of most commentators (including yours truly), that didn’t happen this time. As of a week after Election Day, the number of seats that the GOP will gain in the House looks likely to be in the single digits—a far cry from the 54 seats that Gingrich and company gained in 1994 or the 64 seats that a Tea Party-fueled GOP picked up in 2010.

So, what happened? A pair of numbers leaps out of the exit polling: 32 percent of voters said that they cast their House vote to “oppose” President Joe Biden, while 28 percent said they cast their House vote to “oppose” former President Donald Trump. In other words, for every eight votes cast against Biden, all but one was negated by a vote cast against Trump. This is surely unprecedented in a midterm election. It’s nearly impossible to imagine a previous midterm in which almost as many people voted against the loser of the previous presidential contest as voted against the winner. How many people, for example, bothered to vote against Richard Nixon in 1962, Jimmy Carter in 1982, George H. W. Bush in 1994, or even Hillary Clinton in 2018?

Of course, it didn’t help Republicans that the leading establishment faces of their party are even less popular with voters than Trump. The former president’s favorability rating in exit polling was -19 percentage points (39 percent favorable, 58 percent unfavorable), worse than Biden’s -15 points (41 percent favorable, 56 percent unfavorable). But Republican House leader Kevin McCarthy’s favorability rating (-26 points, with 27 percent favorable and 53 percent unfavorable) was not only lower than Trump’s but also lower than House Speaker Nancy Pelosi’s (-24 points, with 36 percent favorable to 60 percent favorable). RealClearPolitics lists Republican Senate leader Mitch McConnell’s unfavorability rating as of Election Day as being nearly triple his favorability rating (59 percent vs. 21 percent). Per RCP, McConnell’s net favorability rating of -38 percentage points is 24 points worse than that of his Democratic counterpart, Chuck Schumer (-14 points, with 33 percent favorable and 47 percent unfavorable).

Many observers have blamed Republicans’ lackluster showing in the midterms on their positions on abortion, but exit polling suggests a more nuanced picture. On the one hand, Fox News exit polling indicates that few voters (10 percent) considered abortion to be “the most important issue facing the country.” On the other, 25 percent regarded the Supreme Court’s overturning of Roe v. Wade as the “single most important factor” to them personally when “thinking about voting in this election.”

The network consortium’s exit polling found that only slightly more voters think abortion should be “legal” (30 percent) as think it should be “illegal” (26 percent) “in most cases.” But those in the latter camp appear to hold their positions with more conviction, as they were far more likely to support Republicans (90 percent to 9 percent) than those in the former camp were to support Democrats (60 percent to 38 percent). Indeed, the Republicans’ 81-point margin among voters who think abortion should generally be illegal swamped the Democrats’ 22-point edge among those who think it should generally be legal.

Moreover, Republicans won a majority of the vote among the 58 percent of voters who, in response to the overturning of Roe, felt “enthusiastic” (16 percent), “satisfied” (21 percent), or even “dissatisfied” (21 percent). Combining those three groups, Republicans won by a margin of 50 points (74 percent to 24 percent). Only the 39 percent who felt “angry” in response to Dobbs v. Jackson Women’s Health Organization supported Democrats (85 percent to 14 percent).

Thus, more moderate voters on this contentious issue—those who were neither enthusiastic nor angry, and those who think abortion should be neither legal nor illegal in every case—were more apt to favor Republicans. Among those with more intense views either way—those who were enthusiastic or angry, and those who think abortion should either always be legal or always be illegal—Democrats prevailed.

Many have cited increased mail-in voting as a potential cause of Democrats’ relative success in this election. Leaving aside questions about its advisability, mail-in voting doesn’t seem to have led to a huge increase in voting among the relatively disengaged. Indeed, 114 million ballots were cast in the pre-Covid midterm election of 2018, whereas 104 million votes have been counted in this election, with about 5 percent to 10 percent remaining to be tallied. By the time the counting is done, the number of votes in 2022 will likely be about the same as in 2018. Nor can mail-in voting explain the surprising 9 percentage-point win by Democratic senator Maggie Hassan over Republican challenger Don Bolduc in New Hampshire, after polling had showed them separated by just 1.4 points. New Hampshire runs basically an old-school election, in which most voters cast their votes in person on Election Day.

When it came to mobilizing its own voters and appealing to independents, each party had something to be happy about in this election. Republicans could be pleased that they outnumbered Democrats, 36 percent to 33 percent. Democrats could be pleased that they won by 49 percent to 47 percent among independents, an unusual result in a midterm election in which a Democratic president has an approval rating in the low 40s.

Democrats did even better among independents in the eight Senate races decided by single-digit margins (in Arizona, Georgia, Nevada, New Hampshire, North Carolina, Ohio, Pennsylvania, and Wisconsin). In these races, Democrats beat Republicans among independents in all but Wisconsin (where Ron Johnson won among independents by 50 percent to 48 percent). Democratic Senate candidates won among independents by 11 points in Georgia (53–42), 16 points in Arizona (55–39), and a whopping 20 points in Pennsylvania (58–38).

Republicans’ poor performance among independents is part of the GOP’s larger problem of losing close races in this election. Republicans appear on course to win the nationwide House vote by 4 percentage points, 51-47, yet seem poised to pick up a single-digit tally of House seats. Contrast this with 2020, when Republicans lost the nationwide House vote by 3 percentage points yet gained 13 seats; or with 2010, when they won the nationwide House vote by a little under 7 percentage points and gained 64 seats. It seems safe to say that Republicans were outmaneuvered in the redistricting battles. However, the GOP’s poor performance in close races extends to statewide contests as well. In Senate seats decided by single digits, Republicans posted a win-loss record of 3–4, with Georgia still to be decided. In gubernatorial races decided by single digits, Republicans went 2–7. Not since Barack Obama swept the nine key swing states versus Mitt Romney in 2012 have Democrats won so many close, high-profile races in one election.

There has been much talk about “candidate quality” on the GOP side, with Republicans allegedly having run “extremist” candidates. Exit polling, however, doesn’t suggest that voters were generally turned off by Republicans as extremists. In Senate races decided by single digits, only Blake Masters in Arizona was regarded by a clear majority of voters (54 percent) as holding “views” they thought were “too extreme”—and even that perception may have been heavily influenced by his having been outspent nearly eight to one by Democratic incumbent Mark Kelly. In House races, meantime, when asked whether either party was “too extreme,” 39 percent thought only the Republicans were extreme, 38 percent thought only the Democrats were extreme, and 13 percent thought both were extreme.

Candidate quality may have helped Democrats and hurt Republicans in one way: candidate résumés. Trump and Virginia governor Glenn Youngkin notwithstanding, voters generally expect candidates to have held political offices widely viewed as stepping-stones for the one for which they are currently running. Before Trump, for example, all but one candidate elected as president had been either a vice president, governor, senator, cabinet secretary, or commanding general. The singular exception is Abraham Lincoln, the foremost spokesman on the great issue of his day, and even he had been a congressman. In 2022, all 17 Democratic candidates in Senate or gubernatorial elections decided by single digits had either previously won statewide elective office, been a leader in the state legislature, or been elected to Congress. Only seven of 17 Republicans met these criteria. Democrats did a far better job than Republicans of heeding voters’ expectations for experience.

But to the extent that commentators mentioning candidate quality really mean that voters crave more milquetoast establishment candidates, neither the election results nor exit polling provide much support for that. While John Fetterman did fit the traditional résumé requirements (as Pennsylvania’s sitting lieutenant governor), he was hardly an establishment candidate.

Likewise, when voters were asked which officeholders they “strongly approve” of (a question apparently asked only about chief executives in a half-dozen states or nationally), the officeholder who fared best was Florida governor Ron DeSantis. His performance as governor garnered strong approval from 46 percent (versus only 29 percent who strongly disapproved). Having defied public-health officials on masks, vaccines, and lockdowns; the academic elite on critical race theory; and a powerful lobby on sexualizing K–3 education, DeSantis is anything but a milquetoast establishment candidate. The candidate with the second-highest “strongly approve” percentage (40 percent, versus 36 percent strongly disapproving) was Texas governor Greg Abbott, who imposed a mask mandate in Texas but has been more anti-establishment on issues like immigration. Meantime, only 18 percent “strongly approve” of President Biden (versus 45 percent who “strongly disapprove”), a long-time establishment politician. That Republicans couldn’t take advantage of Biden’s unpopularity is of course the great surprise of this election.

Perhaps another reason for Republicans’ failure to capitalize on the political moment was that very few Republicans ran on their opposition to Covid mandates and lockdowns. (DeSantis was an exception.) The exit polling can’t really answer this question because it didn’t ask about issues on which people weren’t campaigning.

The exit polling does tell us about the divide in this country. For example, about two-thirds of the 31 percent of voters who attend religious services on a weekly basis voted Republican, while about two-thirds of the 30 percent who never attend religious services voted Democratic. Likewise, about two-thirds of the roughly half (53 percent) of Americans whose households own guns voted Republican, while slightly more than two-thirds (68 percent) of the roughly half (47 percent) of households who don’t own guns voted Democratic.

Exit polling captured further opportunities and concerns for both parties. Almost three-quarters of voters (74 percent) are either “dissatisfied” or “angry” about “the way things are going in the U.S.” The quarter who are either “enthusiastic” (5 percent) or “satisfied” (20 percent) about America’s trajectory overwhelmingly voted Democratic (85 to 13 percent). A slight majority of voters (52 percent) aren’t “very confident” that their “state’s elections are fair/accurate,” so it’s hardly just the fringe Right that harbors these concerns. Likewise, by about a two-to-one margin, voters think “society’s values on gender identity/sexual orientation” are “changing for [the] worse” (50 percent), rather than “for [the] better” (26 percent). And by a 14-point margin, voters were more apt to think Biden’s policies are “hurting” (47 percent) America rather than “helping” (33 percent).

On the flip side, a slight majority of voters (53 percent) think racism is a “major problem” (73 percent of whom voted Democratic), rather than one that’s either “minor” (30 percent) or nonexistent (15 percent). A plurality of voters (46 percent) think “climate change” is a “very serious” problem; 83 percent of them voted Democratic. Only a slight majority of voters think the climate is either a “somewhat serious” problem (25 percent) or “not [a] serious” problem (27 percent), and 81 percent of them voted Republican.

Republicans fared much better among Hispanic voters in this election (losing by 21 points) than Romney did in 2012 (losing by 44 points), and also much better among black voters (losing by 73 points, versus Romney’s 87-point deficit in 2012), continuing trends that began with Trump. The more years voters spent in school, the more apt they were to vote Democratic; the more money they made, the more apt they were to vote Republican. Thus, for Republicans, the sweet spot would seem to be the relatively affluent tradesman who didn’t go to college (a plumber, for example); and for Democrats, the grossly underpaid adjunct professor whose degrees fill up most of the wall space in the tiny apartment he can barely afford.

Republicans won by at least 7 percentage points among married men, married women, and single men, while Democrats won by 37 points among single women. Democrats won in urban areas (by 17 points), Republicans won among the smaller collection of voters in rural areas (by 29 points), and voters in the suburbs more or less matched the national tally (favoring Republican by 6 points).

The exception to these general trends was in the Florida gubernatorial election, where Republicans not only closed the gap but also won outright among urban voters (by 13 points), Hispanic voters (by 18 points), and independents (by 8 points). They almost won even among single women, losing by just 1 point.

Overall, exit polling shows an American citizenry dissatisfied with both parties, most politicians, and the general direction of the country. It tells us that citizens are looking for genuine leaders who are up to the challenges of these times—leaders who can weather the storm, steady the ship, and steer it in the right direction. Until voters see leaders better fitting this description, neither party will hoist its flag for long.

Photo by Anna Moneymaker/Getty Images


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next