In counties where a smaller percentage of residents held bachelor's degrees, voters were more likely to support Donald Trump. In counties where a greater percentage of residents held bachelor's degrees, voters were more likely to prefer Joe Biden. (Deesmealz/Dave Leip's Atlas of U.S. Presidential Elections, U.S. Census)

Way back in 1976, political scientist Everett Carll Ladd, Jr., proposed that liberalism had turned “upside down.”

Coming out of the 1930s New Deal, Ladd explained, the American political system had clear class lines: Democrats represented the working stiff. Republicans were the party of the bosses, the professionals, the owners of small businesses, the families with college degrees. But beginning in the 1960s, Ladd found, there had been a slow “inversion of the relationship of class to electoral choice….” — a change that is readily apparent in the latest Presidential election.

In 1960, only 38% of white voters with the highest income and education voted for Democrat John Kennedy, according to Ladd. In the third of white voters with the lowest incomes and least education, 61% voted for Kennedy. That was the New Deal order — workers voted Democratic and bosses and college grads backed Republicans.

By 2020, the New Deal order was not only upside down, but inside out, topsy turvy and bassackward. The counties that have the richest families and the best-educated people are Democratic. Those with lower incomes and where college degrees are few and far between voted Republican.


The story here is simple: The higher the percentage of people with college degrees in a county, the higher the Democratic vote. (The 2020 results are at the top of this post, and the 2016 results are just above.)

We sorted the nation’s 3,110 counties according to the percentage of adults living there who have a bachelor's degree or more. The counties with the lowest percentage of diploma-holders are on the left.

You can see the steady increase in the Democratic vote (blue bars) as the percentage of adults with college degrees increases. There is a 28 percentage point difference in the Democratic vote in 2020 between the most educated group of counties and those counties with the lowest percentage of people with college degrees.

That gap widened by 4 percentage points from 2016 to 2020.

Trump (red bars) won 63% of the vote in the counties with the lowest percentage of people with college degrees this year. The President’s percentage dropped 28 points in the counties with the highest percentage of college graduates.


These charts show the election results by county median household income. Low-income counties (under $40,000 in median household income in 2018) are on the left. High-income counties (a median over $80,000 per household) are on the right.

Again, there is a nice stairstep up in the Democratic vote as the median income of the county increases. In 2020, Joe Biden won nearly 60% of the vote in the counties that had the highest median income. Donald Trump won only 38.6% of the vote in those counties.

Trump, however, won 54.6% of the vote in the counties with the lowest income; Biden won 44.2% of the votes in those counties.

Democrats increased their share of the vote in the richest group of counties from 55% in 2016 to 59.4% in 2020.

And so the political gap also increased between high and low-income communities. The counties with the highest income in 2016 were 9.5 percentage points more Democratic than the counties with the lowest median income.

In 2020, the difference in Democratic vote between these two groups of counties grew to 15.3 percentage points.

We are a long way from the New Deal. Now the Democratic Party represents the richest, most educated places. As earlier Yonder analysis found, these places are largely in the central cities or the suburbs of the major metro regions.

The one area where the old voting order prevails is with race. Deesmealz will show how the racial makeup of counties affected votes this year in a later post.

Creative Commons License

Republish our articles for free, online or in print, under a Creative Commons license.