Supporters of President-elect Donald Trump rejoiced across the nation on Election Night as their candidate defied the polls. (Monica Akhtar/The Washington Post)

We knew all along that Donald Trump drew his strength from the white working class. We knew this from the patterns in the primaries. We knew this from the nonstop polling conducted over the past 18 months. We knew this from all of the campaign-trail dispatches showing his anti-trade, anti-elite message thrilling crowds in the heartland.

Tuesday proved that this demographic remains a powerful force in U.S. politics — and the president-elect has thoroughly charmed the group. He vastly overperformed 2012 Republican presidential nominee Mitt Romney in the Midwest and the Rust Belt, in many cases flipping counties that had decisively voted for President Obama in 2012.

For instance, Obama won Iowa’s Howard County, a farming community that is 97 percent white by 21 points in 2012. On Tuesday, Trump took Howard County, which bills itself as “Iowa's Year 'Round Playground,” by 19 points. In Luzerne County, Pa., a place just outside of Scranton, Obama won by five points in 2012; Trump took it by 20 points. In Juneau County, Wis., smack in the middle of the state, residents voted for Obama by a seven point margin in the 2012 election. Voters there picked Trump over his Democratic challenger, Hillary Clinton, by 26 points.


The specks of red — where Trump won counties that previously voted for Obama — dot the Rust Belt. And these counties all had something in common. They were dominated by whites without a bachelor’s degree.

In the charts below, I have plotted the correlation. Even within the Midwestern states, it was the counties with higher populations of working-class whites who handed Trump his victory. Many of these areas have been going red for decades — but Trump went above and beyond. He performed significantly better than Romney in places dominated by working-class whites.


Even after controlling for income levels, employment levels, population size and the foreign-born population in a given county, the more white people there were who lacked a four-year degree, the more likely Trump outperformed Romney in that county. Roughly speaking, if you took an average county and increased the portion of less-educated whites by 10 percentage points, you would boost Trump’s winning margin by about three percentage points.

What made these white voters change their minds?

Certain economic factors were important. Trump tended to outperform Romney in places where median incomes were a little lower, places where people tended to be out of work and where middle-aged whites were more likely to die.

But these places have been like this for a while. In fact, Trump's victory doesn't seem to be linked to any recent declines in people's economic circumstances. The economy has been getting better over the past four years. Median incomes have risen. The unemployment rate has plummeted including in regions won by Trump:


In fact, if you just look at the simple correlation, Trump overperformed Romney in counties where the unemployment rate fell the most between 2012 and now. Part of it is that these places tended to have higher unemployment rates in 2012, so they had more room to improve. But it's clear it wasn't some recent decline in employment that affected how these counties voted. (The same is true of changes in median income.)


Voters often say that the economy is the most important thing on their minds, but this is more evidence that for politics, the economy is a state of mind.

As Gallup economist Jonathan Rothwell has pointed out, surveys show that Trump supporters are not necessarily poorer than average. It may be that many are probably doing pretty well, but they may see others in their neighborhood who are struggling and decide that the nation, as a whole, isn’t that great anymore.

That would accord with what political scientists have been saying for a long time. They emphasize that in politics, perceptions are important. People don’t always vote in their own self-interest — they think about what they believe will be good for the country as a whole. They also think about a candidate’s character. They judge whether a politician is capable of listening to them and their community.

Trump, on some level, understood the importance of making members of the white working class feel as though they were being heard. He tapped into deeper, slower-moving resentments.

Political scientist Kathy Cramer has spent nearly a decade talking to rural Wisconsin voters about their views on government. In her book, “The Politics of Resentment,” she describes the feeling like this:

The economic woes people communicated to me … were interlaced with their sense of who they are, who is a part of their community, what their values are, who works hard in society, who is deserving of reward and public support, and how power is distributed in the world. This complex set of ideas is the product of many years of political debate at the national level as well as generations of community members teaching these ideas to each other. This entwined set of beliefs was not something that any one politician instilled in people overnight — or even over a few months.

In other words, the tension was always there. Trump just found a new way to flick at it.