America's eight political tribes

Averaging the forecasts, Wisconsin polling woes, the fight over crowd size, how the election will affect the markets, the battle for religious voters, Census data misses migrants

No. 330 | September 13th, 2024

📊  Public Opinion

Each year, my team at Echelon Insights produces a Political Quadrants and Multiparty Democracy poll that goes beyond the traditional left-right divide to identify the country’s political tribes in a nuanced way. In a recent survey conducted in conjunction with Baron Public Affairs, we wanted to take this a step further, by not applying fixed definitions to these tribes, and simply letting the data sort them for us. 

How did we do that? Though a cluster analysis applied to our normal social and economic indices plus a new index that measures trust in institutions and the positions taken by most experts—from NATO to free trade to vaccines. This latter populist vs. establishmentarian dimension is the fuel for the rise of Trump to Brexit to populist parties worldwide. And, as it turns out, it has a lot of explanatory value, dividing members of the political parties themselves better than the normal social and economic questions. 

Check out the full presentation on the eight tribes we found and reply in the comments to tell us which one you think you are. 

Women aged 18 to 29 today are more liberal than young women in the past on specific issues, particularly the environment and abortion. This gender divide among young people supersedes any differences by education level. 

🇺🇲 2024

The average of the averages this week shows a slight shift back in Trump’s direction (pending more post-debate polling). 

• Silver Bulletin: Harris +1.9 (Trump+1.1)

• 538: Harris +2.8 (T+0.3)

• NYT Upshot: Harris +3.0 (H+0) 

• The Hill/DDHQ: Harris +3.4 (T+0.4)

• RCP: Harris +1.5 (T+0.3)

• Cook Political: Harris +1.2 (T+0.4)

• VoteHub: Harris +2.9 (T+0.6)

• RacetotheWH: Harris +3.1 (T+0.3)

Average of the Averages: Harris +2.5 (T+0.4)

Friend-of-the-newsletter Scott Tranter has also been rounding up an average of the forecasts.

Politico’s Steve Shepard always has a good beat on which polling trends to pay attention to, and which to ignore.  

Kyle Kondik at Crystal Ball also hits on an argument readers have seen me make: be highly skeptical of polls in Wisconsin.

We’re on different sides politically, but Mike Podhorzer’s Weekend Reading is a must-subscribe. And I think his new post on the substantial differences in crosstabs based on polling methodology is one to pay serious attention to. 

You can produce the same tied topline result with a poll that shows no gender gap as one that has a widening gender gap, but the meaning of these two polls is wildly different. 

Using data journalism to settle one of the dumbest scores of the campaign—the battle over crowd size.  (This mostly ends up being a catalog of the size of different venues, since both Trump and Harris have generally filled the exact same arenas where they’ve both held rallies.) 

The WSJ has issue polling on which candidate campaign promises are popular. 

Unsurprisingly, people like getting free things from the government. 

Traders are trying to figure out how a highly uncertain election outcome will affect the stock market. As usual, the answer is not much. 

In my book, I track the political history of white Catholics as a potential precursor to Hispanic voting behavior—from voting 3-to-1 for JFK to now being a Republican group. That gap has now widened to 23 points. Most groups with any religion at all, Black Protestants and Jewish voters being the biggest exceptions, lean further and further Republican as the rise of the religiously unaffiliated polarizes Americans politically by religion. 

👫 Demographics

The Census Bureau seems to be completely missing the recent spike in undocumented immigration, while the Congressional Budget Office projects net international migration 2 million people higher than the latest Census figures. A must-read piece by Jed Kolko at Slow Boring.