The return of racial realigment
White college grads shift left, religious voter groups to watch, working class politics, appreciating Pew's crosstabs, Harris's TV advantage narrows, my take on recalled vote weighting
No. 334 | October 11th, 2024
🇺🇲 2024
A slight tightening in the averages for the week.
Silver Bulletin: Harris +3.0 (T+0.4)
538: Harris +2.4 (T+0.1)
NYT Upshot: Harris +3.0 (H+1)
The Hill/DDHQ: Harris +2.9 (T+0.5)
RCP: Harris +1.8 (T+0.4)
Cook Political: Harris +2.2 (T+0.3)
VoteHub: Harris +3.3 (T+0.1)
RacetotheWH: Harris +3.0 (T+0.5)
Average of the Averages: Harris +2.7 (T+0.2)
After a pause following Joe Biden’s exit from the race, racial realignment discourse is back.
A reminder from the FT that Realignment is a global trend: ethnic voters in the U.S. and U.K. are shifting away from the mainstream left.
White Democrats used to resemble Black and Latino voters in their policy preferences. Now they’re much further to the left.
In NBC polling, the Democratic advantage in party ID with Hispanic voters has gone from 41 points in 2012 to as low as 12 points in the most recent Telemundo poll of Hispanics.
On the flip side, Harris gained in September with white college graduates.
But it’s worth looking at other aspects of realignment, beyond race and education.
Religion plays a role too. Under Trump, Republicans have gained the most with white non-college voters with low levels of religious attendance.
Meanwhile, Trump appears to be gaining back the Catholic support he lost to Joe Biden in 2020, himself a Roman Catholic. It’s been harder to discern shifts among white voters this cycle, but there’s evidence that white Catholic voters are shifting, with outsized implications for Pennsylvania and Wisconsin.
Jacobin (I know) has a fascinating look at the occupational dimensions of working class politics. Manual workers in Pennsylvania are for Trump, while service workers are for Harris. Trump also leads among current or former union members, people with more job insecurity, and people who say they’ve been unfairly terminated.
I’m usually not one to cite individual polls, but Pew’s crosstabs here are immaculate.
Pro-Harris sponsors have spent over $109 million in television and radio ads in the last two weeks compared to about $71 million for pro-Trump sponsors.
While Harris still has a spending advantage, Trump’s add airings have ticked up.
Trump has held more campaign events than Harris.
Crystal Ball has the counties to watch in the Sun Belt.
Last year’s court battles over redistricting might have more to do with who controls the House than this November’s results.
Important Nate Cohn piece on how pollsters are dealing with weighting on recalled 2020 vote. Those polls that do tend to show a 2020 re-run while those that don’t tend to show more interesting and potentially variable results.
My take: I don’t think the rule that the last election winner gets a boost necessarily applies anymore. The last two presidents have been extremely polarizing and relatively unpopular throughout their terms, so there doesn’t seem to be a tendency to over-report voting for them.
In fact, I think it may now be the opposite following Biden’s exit from the race: fewer people across partisan groups are reporting having voted for Biden now that he’s no longer a candidate they have to support, however grudgingly. That means the question comes out Trumpier relative to partisanship, and a Biden +4 target may be too left-leaning.
🗺️ Data Visualization
A piece from earlier this year newly relevant this week: Home insurance rates have almost nothing to do with risk from natural disasters.