Episode Summary
Executive Summary: Derek Thompson and Nate Cohn dissect Joe Biden’s weak polling, arguing it reflects more than generic “Biden old, Republicans crazy” explanations. They focus on an engagement gap: highly engaged voters look more Democratic, while less engaged, younger, Black, and Latino voters are far more disaffected with Biden. The conversation also covers polling limits, the economy, and why 2024 may be unusually susceptible to year-ahead polls.
Main Topics: Why Biden polls badly despite Democratic wins (Priority: 5/5): The episode contrasts Biden’s poor national and swing-state numbers with recent Democratic off-year victories, asking whether media panic over polls is disconnected from actual election outcomes. Polling methodology and modern error sources (Priority: 5/5): Nate Cohn explains low response rates, sampling bias, and weighting, emphasizing that polls are imperfect but still informative when interpreted cautiously. Biden’s approval problem: economy and age (Priority: 5/5): The discussion frames Biden’s unpopularity as a mix of economic dissatisfaction and a pervasive perception that he seems too old and less effective on television. The engagement gap (Priority: 5/5): A central thesis is that highly engaged voters differ sharply from less engaged voters; this gap may explain why Democrats can win lower-turnout races while Biden performs poorly in broader polling. Demographic realignment among nonwhite and working-class voters (Priority: 4/5): Cohn argues that Black, Latino, and white working-class voters are weakening older Democratic loyalties, partly because Trump’s populist brand has captured some anti-establishment and working-class themes. 2024 turnout dynamics may differ from prior elections (Priority: 5/5): The conversation suggests higher turnout could help Trump if it brings in less engaged voters who are more anti-Biden than anti-Trump, reversing an older Democratic turnout advantage. Reading 2023 off-year elections carefully (Priority: 4/5): Cohn warns against over-interpreting Kentucky, Ohio, and Virginia results because they were atypical contests, though they do fit with broader patterns in turnout and engagement.
Key Arguments: Modern polling is harder because response rates are extremely low, creating large risks of nonresponse bias even after weighting and targeted sampling. Polls are not perfect predictors, but they still contain real, actionable information about voter attitudes and demographic shifts. Biden’s weak standing is driven less by Trump strength than by Biden’s own erosion, especially among younger, Black, Latino, and low-engagement voters. The economy is unpopular not just because of inflation but because post-inflation prices remain high and other conditions like interest rates and supply constraints shape public pessimism. Biden’s age matters because voters judge what they see, and the White House likely limits his appearances because he often reinforces concerns about being too old. The most important frame for 2024 may be the engagement gap: highly engaged voters are more Democratic, while less engaged voters are more susceptible to Biden skepticism and may not show up consistently. High turnout in 2024 could be a mixed blessing for Democrats if it expands the electorate with more low-engagement, less anti-Trump voters who are also less committed to Biden. Recent demographic shifts may reflect the weakening of historical party attachments and the success of Trump’s populist appeal among less affluent voters across racial groups.
Data Points: Swing states in NYT poll: Trump led in 5 of 6 battlegrounds - Biden trailed Trump in Nevada, Georgia, Arizona, Michigan, and Pennsylvania, while leading in Wisconsin. Average battleground margin: Trump 48% to Biden 44% - Across the six battleground states in the cited New York Times poll. Telephone survey completion rate: About 1% - Cohn said only about 1% of attempted telephone calls yield a completed political interview. 2016 national poll error: Clinton +4; actual result Clinton +2 - Used to show that even flawed polls can still be directionally informative. 2020 national poll error: Biden +8; actual result Biden +4.5 - Another example of polls being off but not useless. Biden too old share: 70% of voters - Cohn cited a poll finding that about 70% think Biden is too old to be an effective president. Approval rating comparison: Below every postwar president not named Jimmy Carter - Derek described Biden’s approval roughly 1,020 days after inauguration as historically low for an incumbent. Black/Latino/nonwhite movement: Support declined across nearly every cohort - Referenced Cohn’s analysis showing nonwhite support for Democrats falling across age, gender, income, and education groups except high-income nonwhite voters. High-income nonwhite exception: Income over $100,000 - The only nonwhite subgroup where Democratic support held steady in Cohn’s analysis. Off-year election examples: Kentucky, Ohio, Virginia - Democrats won key contests in these states shortly after the NYT poll caused backlash.
Pivotal Quotes: "The political experience of the Biden era for Democrats is extended periods of intense anxiety about terrible polling, occasionally punctuated by strangely positive election results." — Chris Hayes (quoted by Derek Thompson): Used to summarize the recurring disconnect between polling anxiety and actual Democratic election performance. "Republicans are crazy and Joe Biden is old. That's it." — Derek Thompson: A deliberately simplified theory Derek presents and then challenges as incomplete. "The thing that showed was that Biden, in particular, was far weaker than he used to be." — Nate Cohn: Cohn clarifies that the notable finding of the NYT poll was Biden’s weakness more than Trump’s strength.
Implications: The episode suggests 2024 may hinge less on conventional turnout wisdom and more on whether Democrats can re-engage disaffected low-participation voters. Polls a year out should be treated seriously, but as indicators of real attitudes, not certainties.