With the midterm election less than two weeks away, polling has turned bleak for the Democrats, not only increasing the likelihood that the party will lose control of the House, but also dimming the prospects that it will hold the Senate.The key question is whether Republicans will wipe out Democratic incumbents in a wave election.In a 2021 article, “The presidential and congressional elections of 2020: A national referendum on the Trump presidency,” Gary Jacobson, a political scientist at the University of California San Diego, described how the Trump administration and its 2020 campaign set the stage for the 2022 midterms:Reacting to the [Black Lives Matter] protests, Trump doubled down on race‐baiting rhetoric, posing as defender of the confederate flag and the statues of rebel generals erected as markers of white dominance in the post‐Reconstruction South, retweeting a video of a supporter shouting “white power” at demonstrators in Florida, and vowing to protect suburbanites from low-income housing that could attract minorities to their neighborhoods.The headline and display copy on my news-side colleague Jonathan Weisman’s Oct. 25 story about the campaign sums up the party’s current strategy:With Ads, Imagery and Words, Republicans Inject Race Into Campaigns: Running ads portraying Black candidates as soft on crime — or as “different” or “dangerous” — Republicans have shed quiet defenses of such tactics for unabashed defiance.Republican strategies that emphasize racially freighted issues are certainly not the only factor moving the electorate. Republican skill in weaponizing inflation is crucial, as is inflation itself. Polarization and the nationalization of elections also matter, particularly in states and districts with otherwise weak Republican candidates.Jacobson is one of a number of political analysts who argue that the calcification of the electorate into two mutually adversarial blocs limits the potential for significant gains for either party. In a recent essay, “The 2022 U.S. Midterm Election: A Conventional Referendum or Something Different?” Jacobson writes:Statistical models using as predictors the president’s most recent job approval ratings and real income growth during the election year, along with the president’s party’s current strength in Congress, can account for midterm seat swings with considerable accuracy. For example, applying such a model to 2018, when President Donald Trump’s approval stood at 40 percent and real income growth at 2.1 percent, Republicans should have ended up with 41 fewer House seats than they held after the 2016 election — improbably, the precise outcome.Applying those same models to the current contests, Jacobson continued,the Democrats stand to lose about 45 House seats, giving the Republicans a 258-177 majority, their largest since the 1920s. For multiple reasons (e.g., inflation, the broken immigration system, the humiliating exit from Afghanistan) Biden’s approval ratings have been in the low 40s for the entire year. High inflation has led to negative real income growth.No wonder then, Jacobson writes, that “the consensus expectation at the beginning of the year was an electoral tsunami that would put Republicans in solid control of both chambers.” Now, however, “this consensus no longer prevails.”Why?Partisans of both parties report extremely high levels of party loyalty in recent surveys, with more than 96 percent opting for their own party’s candidate. Most self-identified independents also lean toward one of the parties, and those who do are just as loyal as self-identified partisans. Party line voting has been increasing for several decades, reaching the 96 percent mark in 2020. This upward trend reflects a rise in negative partisanship — growing dislike for the other party — rather than increasing regard for the voter’s own side. Partisan antipathies keep the vast majority of Democrats and Democratic-leaning independents from voting for Republican candidates regardless of their opinions of Biden and the economy.Jacobson noted in an email that over the past weekthe numbers have moved against the Democrats, and they should definitely be worried. The latest inflation figures were very bad news for them. But I still doubt that their House losses will approach the 45 predicted by the models and I think they still have some hope of retaining the Senate — or at least, their tie.Jacobson points out that in the current lead-up to the midterms, there is an exceptionally “wide gap between presidential approval and voting intentions, with the Democrats’ support on average 9.2 percentage points higher than Biden’s approval ratings.” He also notes that in previous wave elections, the spread between presidential approval and vote intention was much closer, 5 points in 1994, 4.9 in 2006, 0.3 in 2010 and 4.1 in 2018.Julie Wronski, a political scientist at the University of Mississippi, argued in an email that polarization has in very recent years changed the way voters evaluate presidents and, in turn, how they cast their ballots in midterm contests. “There is a higher floor and lower ceiling in presidential approval,” she said:If anything, approval is fairly resistant to external shocks in ways that look very different from either George W. Bush or Obama. An approval rating below 50 percent seems to be the new norm. But if we think about this from a partisan lens, an overwhelming percent of Democrats will always support the Democratic president, while an overwhelming percent of Republicans will oppose him.Put another way, Wronski said, “it wouldn’t matter what Biden does or doesn’t do to curb inflation, Democrats will largely support, and Republicans will largely oppose.”In this context, “partisanship serves as lens through which economic conditions are evaluated. The stronger partisanship exists as a social identity, the more likely it will be used as the motivation to view and accept information about economic conditions, like inflation.”Negative partisanship, Wronski wrote, “has emerged in recent elections as a driver of voting turnout and vote choice,” with the resultthat partisan antipathies keep Democrats from voting for Republican candidates. No matter how bad economic conditions may be under Biden, the alternative is seen as much worse. The threat to abortion rights and democracy should Republicans take control of Congress may be a more powerful driver of voting behavior.While polls show growing public fear that adherence to the principles of democracy have declined, Wronski pointed out thatthose concerns do not trump more immediate needs like being able to afford food, housing, and gas. To be fair, people cannot fight for lofty ideals like democracy when their basic needs are not being met. People need to be secure in their food and housing situation before they can advocate for bigger ideas.There is another factor limiting the number of House seats that the Republican Party is likely to gain: gerrymandering.Sean Trende, senior elections analyst at RealClearPolitics, makes the case that in state legislatures both parties “hoped to avoid creating districts that were uncertain for their party and/or winnable for the other party. One upshot of this is that in a neutral or close-to-neutral environment, there aren’t many winnable seats for either party.”Trende elaborates: “In the swingiest of swing seats where Biden won between 51 percent and 53 percent, there are just 19 seats. Of those seats, 10 are held by Democrats, seven are held by Republicans, and one is a newly created district.” In a neutral year when neither party has an advantage in the congressional vote, Trende writes, if “Republicans won all the districts where Joe Biden received 52 percent of the vote or less and lost all of the districts where he did better, they would win 224 seats.Gerrymandering has created what Trende calls “levees” — bulwarks — that limit gains and losses for both parties. The danger for Democrats is the possibility that these levees may be breached, which then turns 2022 into a Republican wave election, as was the case in 1994 and 2010: “In a universe where Republicans win the popular vote by four points, sweeping all of the districts that Biden won with 54 percent of the vote or less, the levee would break and the Republican majority would jump from 232 seats to 245 seats.”When Trende published his analysis on Sept. 29, the generic congressional vote was almost tied, 45.9 Republican to 44.9 Democratic, close to a “neutral” election. Since then, however, Republicans have pulled ahead to a 47.8 to 44.8 advantage on Oct. 22, according to RealClearPolitics. FiveThirtyEight’s measure of the generic vote shows a much closer contest as of Oct. 25, with Republicans ahead 45.2 to 44.7 percent.In 2010, the Republican Party’s generic advantage in late October was 9.4 points, a clear signal that a wave election was building.Educational polarization — with college-educated voters shifting decisively to the Democratic Party and non-college voters, mostly white, shifting to the Republican Party — in recent elections has worked to the advantage of the right because there are substantially more non-college voters than those with degrees.This year, the education divide may work to some extent to the benefit of Democrats.James L. Wilson, a political scientist at the University of Chicago, pointed out in an email that not only do “polarization and party loyalty make the election outcomes less likely to depend on immediate economic circumstances,” but also “educational polarization, combined with the fact that better-educated voters tend to turn out at higher rates in midterm elections than do less-educated voters, may help the Democrats despite voter concerns about Biden or the economy.”Even with inflation as one of the Democratic Party’s major liabilities, the intensification of polarization appears to be muting its adverse impact.In their 2019 paper, “Motivated Reasoning, Public Opinion, and Presidential Approval,” Kathleen Donovan, Paul M. Kellstedt, Ellen M. Key and Matthew J. Lebo, of St. John Fisher University, Texas A&M University, Appalachian State University and Western University, wrote that “Polarization has increased partisan motivated reasoning when it comes to evaluations of the president,” as the choices made by voters are “increasingly detached from economic assessments.”As partisanship intensifies, voters are less likely to punish incumbents of the same party for failures to improve standards of living or to live up to other campaign promises.Yphtach Lelkes, a professor of communication and a co-director of the polarization lab at the University of Pennsylvania, wrote by email that “people (particularly partisans) are far less likely to, for instance, rely on retrospective voting — that is, they won’t throw the bums out for poor economic conditions or problematic policies.”In the early 1970s, Lelkes wrote, “partisanship explained less than 30 percent of the variance in vote choice. Today, partisanship explains more than 70 percent of the variance in vote choice.”This trend grows out of both identity-based partisanship and closely related patterns of media and information usage.As Lelkes put it:There are various explanations for this. There is an identity/motivated reasoning perspective, where people think better us than them and would prefer a lampshade to an out partisan. Another possibility is that people get skewed information. If I watch lots of Fox News or pay even marginal attention to Republican candidates, I’ll hear lots about the economy. If I watch MSNBC and pay attention to Democratic candidates, I’ll hear a lot about abortion, but less about the economy.Not everyone agrees that polarization will limit Democratic losses this year.John Sides, a political scientist at Vanderbilt, wrote by email that “it is absolutely true that party loyalty in congressional elections has increased. But this does not stop large seat swings from occurring.”There is, Sides continued, “some evidence that midterm seat swings can be driven by people actually switching their votes from the previous presidential election,” suggesting that “clearly not every voter is a die-hard partisan.”Sides remained cautious, however, about his expectations for the results on Nov. 8: “The recent poll trends are pushing toward larger G.O.P. gains but I am not sure those trends suggest the 40+ House seat gains that the national environment would forecast.” A narrow win, he wrote, would mean that Republican leaders in the House will face “a very delicate task. On the one hand, they have to appease Freedom Caucus types. But they also have to protect potentially vulnerable G.O.P. members in swing districts. I do not know how you manage that task, and so I do not envy Kevin McCarthy.”Dritan Nesho, a co-director of the Harvard CAPS/Harris Poll, was distinctly pessimistic concerning Democratic prospects:An empirical analysis of the 2022 midterm polls in the final stretch suggests that this election will tip both the House and the Senate toward Republicans, and it’s no exception to historical trends suggesting the incumbent party tends to lose an average of 28 seats in the House and 3 or so seats in the Senate. Key numbers around lack of confidence in the economy, the pervasive impact of inflation, and a worsening personal financial situation among a majority of voters today, actually suggest a stronger loss than the average.The two best predictive variables for election outcomes, Nesho writes,are presidential approval and the direction of personal finances. Both are severely underwater for Democrats. In our October Harvard CAPS/Harris Poll, Biden has plateaued at 42 percent job approval and 54 percent of voters report their personal financial situation as getting worse. 55 percent blame the Biden administration for inflation rather than other factors (including 42 percent of Democratic respondents), and 73 percent expect prices to further increase rather than come down. 84 percent of voters think the U.S. is in a recession now or will be in one by next year.If that were not enough, Nesho continued,at the same time Democrats are seen as disconnected from the key issues of concern for the median voter. Republicans are connecting better with general voters on inflation and the economy, crime, and immigration; Democrats are seen as preoccupied with Jan. 6, women’s rights/abortion, and the environment, which are further down the list of concerns.Republicans, in turn, have pulled out all the stops in activating racially divisive wedge issues, relentlessly pressing immigration, crime and the specter of generalized disorder.In Missouri, for example, Brian Seitz, a state representative, is determined to “shut down” critical race theory, declaring, “There is a huge red wave coming.” Elise Stefanik, chair of the House Republican Conference, ran a Facebook ad that read: “Radical Democrats are planning their most aggressive move yet: a PERMANENT ELECTION INSURRECTION. Their plan to grant amnesty to 11 MILLION illegal immigrants will overthrow our current electorate and create a permanent liberal majority in Washington.” In Ohio, J.D. Vance, the Republican Senate candidate, contends that Democrats are recruiting immigrants and “have decided that they can’t win re-election in 2022 unless they bring in a large number of new voters to replace the voters that are already here.” Blake Masters, the Republican Senate nominee in Arizona, warns that Democrats want to increase immigration “to change the demographics of our country.”Robert Y. Shapiro, a political scientist at Columbia, observed in an email: “By all rights this should be a debacle for the incumbent party based on the fundamentals — the relative bad news about the economy — inflation — crime, the southern border, and the lingering Afghanistan fiasco.”But, Shapiro added:There are mitigating factors: a very important one is that the Republicans picked up many seats in the House in 2020 so those seats are not at risk now for the Democrats, thanks to around 11 million more Republican voters in 2020 than in 2016. The other factor is the Dobbs abortion decision that led to a surge in Democratic voter registration, very likely significantly women and younger voters. This at best has just helped the Democrats to catch up to Republicans.The crucial question in these circumstances, in Shapiro’s view, “will be relative partisan turnout — will this be more like 2010 or 2018? I sense the enthusiasm and anger here is at least a bit greater among Republicans than Democrats for House voting.”Bruce Cain, a political scientist at Stanford, emailed me to say that he agrees “with those who think the Democrats will lose the House,” but with Republicans seeing “a below historical average seat gain, i.e. under the 40-45 seats that some models are predicting.”Cain argued that a Democratic setback will not be as consequential as many on both the left and right argue: “It’s not like either party needs to worry about being locked out of power for very long. The electoral winds will shift, and the window to power and policy will open again soon enough.” Polarization, Cain noted, “has made it clear to both parties that you have to grab the policy prizes while you have trifecta control” — as both Trump and Biden have done during their first two years in office.One difference between the current election and the wave election of 1994 is that this time around Republicans have no attention-getting, mobilizing agenda comparable to Newt Gingrich’s Contract With America. They have contented themselves with hammering away on the economy, race and immigration.Republicans are fixated on an ethnically and racially freighted agenda of gridlock and revenge. They propose to reduce immigration and to roll back as much as they can of the civil rights revolution, the women’s rights revolution and the gay rights revolution. They threaten to hound Biden appointees, not to mention the president’s son Hunter, with endless hearings and inquiries. The party has also signaled its refusal to raise the debt ceiling and promised to shut down the government in order to force major concessions on spending.While this agenda may win Republicans the House and perhaps the Senate this year, it contains too many contradictions to achieve a durable Republican realignment.The Times is committed to publishing a diversity of letters to the editor. We’d like to hear what you think about this or any of our articles. Here are some tips. And here’s our email: letters@nytimes.com.Follow The New York Times Opinion section on Facebook, Twitter (@NYTopinion) and Instagram. More