The Polls are Still Wrong (But Getting Better)
Now that Vice President Kamala Harris is the Democratic nominee, the Democratic Party’s standing in the polls has markedly improved, but there remain flaws with the polls not being reflected in the real world that should give the Democratic party the optimism it should not be afraid of having.
Reading Time: 6 minutes
Quite a bit has transpired since I wrote “The Polling is Wrong” in Issue 13, Volume 114 of The Spectator during the spring. The Democratic presidential nominee is now Vice President Kamala Harris, after President Joe Biden unexpectedly withdrew on July 21 following a middling debate performance. Harris’s entry into the race has substantially improved from Biden’s position in the polls against Donald Trump. She holds a decent lead in the national popular vote and small, yet decent, leads in battleground states. Despite the changes in the dynamics of the presidential election, there remain substantial flaws within the polls, although to a lesser degree.
For one, there have been rather consistent flaws across various types of polls in their “crosstabs”—polling information regarding various demographic groups. Generally, bad crosstabs in one poll are ignorable assuming they cancel out so that the topline number is used, but the consistent patterns in numerous polls introduce justified skepticism. These crosstabs show plausible yet promising numbers for Harris regarding white voters and older voters—both with and without a college degree—from her efforts to persuade Republican-leaning independents and moderates to vote for her, similar to what happened during the 2022 midterm elections. Yet, these crosstabs contain highly implausible numbers with traditional Democratic constituencies of voters of color and young voters. For example, Harris is up only eight percent with voters aged 18-29 in the latest New York Times/Siena College poll, a twenty percent decrease from the far-Republican-leaning 2022 midterm elections. The Democratic Party’s popularity with these groups has weakened in recent years due to right-wing populism but nowhere to the degree suggested by the polls, which would indicate a generational realignment and the end of racial polarization in American politics as we know it. It is hard to argue that racial depolarization is happening to this extent unless one believes the polling. However, common sense and intuition suggest these polls are wrong. Additionally, with Harris as the presidential nominee, the smaller levels of racial depolarization may slow due to the well-documented phenomenon of racial solidarity in American elections. Harris’s nomination appears to have alleviated disillusionment with the status quo among more liberal to progressive younger voters that was mostly directed at Biden. These factors make it even harder to believe these polls are correct.
It is notable that the Trump campaign has been hinging its entire campaign strategy on these erroneous crosstabs, rather than an approach grounded in reality. Instead of trying to make gains with working-class voters while holding respectable numbers with college-educated voters, campaign managers Chris LaCivita and Susie Wiles have described the campaign’s strategy as “For every Karen we lose, we’re going to win a Jamal and a Enrique.” This is extremely reminiscent of then Senate Democratic Vice Chair Chuck Schumer boasting about how Hillary Clinton’s strategy was “For every blue-collar Democrat we lose in Western Pennsylvania, we will pick up two moderate Republicans in the suburbs in Philadelphia. And you can repeat that in Ohio and Illinois and Wisconsin” in 2016—a boast that would end up haunting the party as they only won Illinois, the most suburban of the four states. Minimizing the idea of winning voters that a candidate has weaknesses with has backfired before, and it is probable that it can backfire again for Trump, especially since suburban women—the Trump campaign calls them “Karens”—are high-propensity voters and often decisive in competitive elections. Additionally, the Trump campaign’s extremely risky strategy to rectify this deficit hinges on voters that just may not exist—unlike the 2016 Clinton campaign’s strategy that hinged on existing voters who did not support her enough to counteract her deficits.
Empirical real-world election results continue to contradict the polling narrative that has emerged. The Democratic Party continues to overperform in off-year, and special, elections. They have also overperformed in a key indicator of the national environment: the Washington primary. Washington holds a nonpartisan blanket primary, which are primaries where candidates of all partisan affiliations compete in one primary election. In these primary elections, the top two candidates advance to the general election. What makes Washington’s primary an extremely good indicator is that the state has automatic voter registration and mails all voters a ballot, leading to high voter turnout. The primary election is also relatively close to the general election. Additionally, the state—compared to similar states’ elections, namely Alaska—is inelastic, so it does not see shifts in partisanship except in response to the national environment. The state’s inelasticity means that it is usually twelve percent more Democratic-leaning than the national environment. Returns show an approximate environment where a generic Democrat defeats a generic Republican by five percent, which is three percent more Democratic-leaning than in 2020 but not as Democratic-leaning as in 2018. This environment was just enough for Democrats to likely hold onto the Commissioner of Public Lands’s office. A whopping five Democrats ran whereas only two Republicans did, nearly splitting the Democratic vote that would have resulted in a Republican-only general election. Democratic candidate Dave Upthegrove only advanced by 49 votes, likely due to the more Democratic-leaning national environment. Notably, in 2016’s more Republican environment, the vote splitting was enough for two Republicans to advance in the Treasurer race by 2.9%. The results of this year’s Washington primary seem to be a good harbinger of this year’s general election results.
There are fairly reasonable concerns that overconfidence and complacency are what led to Trump being elected in 2016 as well, and Trump has overperformed in polling twice before. However, this ignores the underlying fundamental reasons why he overperformed in polling. In 2016, there were a substantial number of undecided voters before Election Day, with many “double negative voters”—voters who held a negative opinion of both Clinton and Trump. Clinton’s large popular vote lead was a Potemkin lead, meaning that the lead seemed stronger than it was in reality. These undecided voters, dissatisfied with the status quo, broke for Trump by a substantial margin. In the last weeks of the campaign, Clinton’s popular vote lead decreased significantly along with her leads in key battleground states, in line with undecided voters breaking for Trump. In 2020, though not definitively proven yet, there appears to have been a fundamental oversampling of Democratic-leaning voters because COVID-19 lockdowns led to them being more likely to answer polls than Republican voters. Clinton’s 2016 leads in polling, Biden’s 2020 leads in polling, and the respective polling errors were of completely different sizes. Polling errors are never guaranteed to be in favor of one candidate or the other, so it cannot seriously be argued that Trump is guaranteed victory again, especially when we have seen more recent polling errors to later favor Democratic candidates—such as 2022 polls indicating an unrealized “red wave.” There is evidence that a polling error in 2024 could be in favor of Harris this time around, because Harris’s favorability ratings are substantially better than Clinton’s, while realistic, and are consistently improving. It appears that complacency this time around is from the Trump campaign believing Virginia and New Jersey could be competitive—much like Clinton competing in states like Utah—rather than from Harris, who consistently asserts her underdog status despite all evidence to the contrary.
Confidence in a political campaign is something that should not be discouraged. My confidence in Harris’s chances is an understandable proposition. What that confidence should not affect is your enthusiasm and energy in a political campaign, which is complacency. Confidence provides the encouragement needed to actually fight the fight, rather than discourage one from even trying. In general, even in long-shot races where there are near zero odds, it is never wrong to keep fighting for a campaign. The Democratic Party is very much taking this to heart in this election cycle while the Republican Party is not. The former is learning from its mistakes in previous election cycles, while the latter continues to shoot itself in the foot. Treating this election as hyper-competitive may end up being the reality of this race and ensure that there is no complacency in the campaign, but this should never inhibit your confidence in a candidate. The future of this country is at stake in the upcoming elections, and I am confident in that future because I am confident that Harris will win the election. And that is something that no one should be afraid to admit.