In 2009, Andrew Gelman published a fascinating and informative book called Red State Blue State Rich State Poor State: Why Americans Vote the Way They Do. Gelman’s main thesis was nicely summarized by Leo Carey in The New Yorker: “Andrew Gelman persuasively argues that the poor in both red and blue states still mostly vote Democratic and the rich, nationally speaking, overwhelmingly vote Republican.” At the time of publication, the data clearly upheld Gelman’s thesis. But with the Republican Party’s turn to rightwing populism under Donald Trump, a reassessment is necessary. If we compare exit-polling data from the elections of 2004 and 2020 (see the edited Wikipedia table posted below), the answer is clear: poor voters have been trending in a Republican direction, although they still primarily supported Joe Biden in 2020. For the affluent (a category limited in the surveys to those earning more than $200,000 a year), support for the Democratic Party has surged. In 2020, the vote of this income category was evenly divided between Biden and Trump.
Precinct-level mapping, made available by the New York Times for 2016 and 2020, shows a more nuanced situation. As can be seen on the map posted below, the extremely wealthy, old-money, and formerly Republican city of Atherton in the San Francisco Bay Area strongly supported Joe Biden over Donald Trump in 2020. Yet at the same time, Atherton remained distinctly “less blue” than neighboring cities, particularly Palo Alto. Palo Alto is an affluent town, but it is not as rich as Atherton. It is also more closely associated with the new-money of Silicon Valley. Also notable on this map is the fact that East Palo Alto, a relatively and poor and ethnically diverse community, gave a slightly lower percentage of its vote to Biden than did Palo Alto. This pattern would not have been seen in elections from previous decades.
The map of the Bay Area vote in 2020 reveals one of the problems associated with precinct-level mapping. A surprising feature of this map if the scattered “red islands” of Trump support in a very blue region. But when one clicks on these red precincts on the interactive Times map, it is quickly revealed that all of them have very few voters, often just one.
The precinct-level maps of New York City reveal similar patterns and tendencies. The Times vote-shift map from 2016 to 2020 shows that relatively poor areas, such as the Bronx, moved in a red direction, even though their total vote-counts remained overwhelmingly Democratic. Wealthy areas in Manhattan, in contrast, showed a modest blue shift, deepening their already dark blue coloration on the actual election map. The 2020 election map also indicates that the wealthiest areas of Manhattan, such as the Upper East Side, are not quite as strongly Democratic voting as somewhat less affluent nearby neighborhoods, such as the Upper West Side. Similar patterns can be found in other American cities. Gellman’s thesis thus appears to be valid to a limited degree: at the precinct-scale, the country’s wealthiest neighborhoods in deep-blue metro areas are still somewhat more Republican-oriented than nearby less-affluent areas.
It will be interesting to see what the 2024 election brings in this regard. I can only hope that Times continues to produce these fascinating precinct-level maps.