Readers of American Nations often ask if something changed with regard to the regional cultures to allow the 2016 Republican nominee, Donald Trump, to capture the Electoral College votes of several key “blue” and swing states and, thus, the White House.

The answer, which initially surprises many, is no. Voting behaviors among the 11 “nations” I identify in the book actually corresponded to what one would have expected. The critical difference wasn’t that some regional cultures had suddenly become “more conservative” or “less liberal,” whatever those terms are supposed to mean these days. What changed was the political program offered by the Republican nominee, who promised the most communitarian-minded agenda of any such nominee in the past four decades. His complete reversal of those promises will likely allow Democrats to roll back his gains in the very places that proved decisive last time around.

Here’s what happened in 2016, as seen through the lens of America’s regional cultures, which, as I’ve previously demonstrated in this series, have more influence on electoral outcomes than the differences between rural and urban life or between the states, whose borders were often drawn with little thought to actual settlement history and cultural fissures. (If you’re new to this paradigm, you can find a summary here.)

First, the top-line results: Just as one would expect, the more communitarian-minded candidate—Hillary Clinton, in this election—won Left Coast, New Netherland, and Yankeedom by wide margins. Like her Democratic predecessor, Clinton also had a wide margin of victory in Tidewater and El Norte, two onetime swing regions that have, for differing reasons, become reliable members of the “blue” coalition over the past 15 years. She also captured the southernmost part of Florida, which is part of a wider Spanish Caribbean regional culture that (like Hawaii and Newfoundland) is not treated in American Nations. In almost all of these “nations,” Clinton actually matched or outperformed Barack Obama’s 2012 results.

Trump, as in the past four elections, won the three “nations” that comprise the current Republican coalition — Deep South, Far West, and Greater Appalachia—as well as the now deeply conservative New France enclave in southern Louisiana. He accomplished this despite deviating from the liberty-minded platform of his recent predecessors, with its promise that fewer taxes, regulations, and social programs would deliver more freedom—a program that resonates with the individualistic ethos of those three regional cultures. It’s perhaps not surprising, then, that Trump didn’t really perform better than 2012 Republican nominee Mitt Romney in these ruby-red regions, registering only a two- to three-point gain in each, save the Far West, where he performed two points worse.

Much of the action, as history would predict, was in the Midlands, the great swing region of U.S. politics and the only one that was the least bit competitive in 2016. This region—communitarian-minded but wary of top-down government action—voted for the Democratic candidate for the third election in a row, but by a greatly reduced margin that proved catastrophic for Clinton. Whereas Obama’s margin of victory in the Midlands was 11 points in 2008 and six in 2012, Clinton won by just 0.4 percent — in effect a tie, and a doomsday result in three traditional Electoral College swing states, where Democrats need strong margins in Midlands areas to compensate for the substantial, highly individualistic Greater Appalachian sections, Pennsylvania and Ohio. It also took Iowa out of competition and ensured there would be no hope of becoming competitive in two states, Indiana and Nebraska, where Obama captured Electoral College votes in 2008.

This development alone ensured Trump’s victory, but a second important shift gave him 27 extra Electoral College votes, which a conventional laissez-faire Republican would have had little hope of capturing.

This second—and not unrelated—development was that Trump significantly improved on Romney and McCain’s results in Yankeedom, losing by eight points rather than 16 or 19, respectively, a swing equal to that in the Midlands. Significantly, this shift was overwhelmingly concentrated in rural areas that traditionally vote for the more community-minded candidate.

As I’ve argued before earlier in this series, the central rift in U.S. politics is not rural versus urban. A key case in point is that in both 2008 and 2012, Obama won scores of rural, white, relatively poor counties—the very sort that are supposed to hate Democrats—across Yankeedom. In 2012, Obama won 15 of 16 counties in my home state of Maine—one of the whitest and most rural states in the nation—along with all the rural counties of Vermont, Massachusetts, and Connecticut and all but one in New Hampshire. Obama also won a great swath of rural Yankee counties in the Upper Mississippi Valley, where Wisconsin, Minnesota, Iowa, and Illinois come together, and dozens of others in (very rural, very Yankee) northern New York, northern Minnesota, and western Wisconsin. Trump flipped a great many of these counties, tipping Michigan, Wisconsin, and Maine’s more-rural second district into his column. Overall, rural Yankee counties went for Trump by more than 18 points, after having voted for Obama in 2008.

Election map by county in 2012 (left) and 2016 (right). Source: Wikipedia

These “Trump Democrats” were almost entirely a phenomenon of Yankeedom and the Midlands, communitarian nations where the traditional Republican message — that fewer taxes, regulations, and public investments will result in more freedom — have never found purchase. Trump’s campaign message resonated here, especially in more homogenous white rural counties, for several reasons.

First, Hillary Clinton was vulnerable on this front. Though she ran as a successor to Obama, both her primary opponent and Trump were able to tie Clinton to her husband’s neoliberalism, a policy legacy that included welfare reform, deregulation of the financial industry, passage of the North American Free Trade Agreement (NAFTA), and the famous pronouncement that the “era of big government” is over, which some of his top aides interpreted as a repudiation of the very New Deal politics that once gave the party a lock on blue-collar workers. This vulnerability was best demonstrated in the primaries, where an obscure Democratic Socialist from a tiny state made Clinton sweat for weeks as he won contests across Yankeedom and up and down the Left Coast and sparred with her in the Midlands. The Democratic base has had enough of the “third way” and was willing to sit out the contest, vote for Russia-linked spoiler Jill Stein, or even rush to a fresh-sounding voice in the GOP.

Democratic primary map. Source: Wikimedia

Recall that Trump, alone among the 17 candidates for the Republican nomination, did not run on a laissez-faire agenda, but rather as a European-style ethno-nationalist, promising robust government intervention and social programs for a subset of Americans and extralegal or extra-constitutional punishment for others. With this agenda, he destroyed the field, including many candidates who were far better qualified, more experienced, and better funded. In office, Trump’s administration has supported oligarchy-friendly policies, from the GOP’s give-to-the-rich tax “reform” to their failed medical-system-destroying 2017 health care bill. But what he promised supporters on the campaign trail was government-led protectionism, industrial intervention, infrastructure spending, and the replacement of the Affordable Care Act with “something terrific.” Social Security and Medicaid would be protected, the swamp of Washington would be drained of its lobbyists, and more taxes might be levied on the rich. It was, in libertarian versus collectivism terms, the most communitarian-sounding set of campaign promises from any Republican nominee since Richard Nixon.

And it worked, especially in rural parts of Yankeedom and the Midlands, where most people belong to Trump’s “in” group of white, native-born Christians and don’t have many friends and neighbors on his list of undesirables: Mexican-Americans; Muslims; immigrants from Haiti, El Salvador, or sub-Saharan Africa; and professional journalists. While Clinton’s margin in Urban Yankee counties slipped by 3.6 points and in Midland urban counties by 3.2, compared with 2012, Trump increased his party’s margin of victory by 18.3 points in both rural Yankeedom and the rural Midlands, winning the latter category by more than 40 points.

Not surprisingly, these areas of my American Nations map are also where many observers of the American political scene have gone looking for answers about 2016’s surprising and dangerous outcome. Katherine Cramer, director of the University of Wisconsin’s Morgridge Center for Public Service, spent years studying rural Wisconsin’s political behavior via listening tours. She concluded that a lot of people felt disrespected and ignored by the proverbial big-city people, whereas they felt Trump saw and respected them. “They were doing what they perceived good Americans ought to do to have the good life, and the good life seemed to be passing them by,” Cramer later wrote. “Part of Trump’s appeal was that he gave people a story, however false or partial, about whom the good life is going to.”

A year into the Trump presidency, however, there are indications that voters in these areas are none too happy with how the president has acted or delivered. A December Selzer-Des Moines Register poll of voters in Iowa—the only state in the country dominated by Midlanders—found the president’s approval rating at 35 percent and disapproval at 60 percent, a staggering reversal for a man who won the state by nine points. Disapproval levels were even worse among women (69 percent), those making less than $50,000 a year (68 percent), and even independents (62 percent). Morning Consult’s state-by-state polling found that in his first nine months in office, the president’s disapproval rating had grown by 12 points (to 52 percent) in (Yankee, largely rural) Maine, 16 points in Vermont (to 62 percent), 12 points in Wisconsin (to 53 percent), and 15 points in Michigan (to 54 percent.) Trump was also underwater in Ohio and Pennsylvania. This very likely indicates a critical loss of support in the areas of these states that swung to him in 2016.

A note for the data geeks among you: The electoral results above do not include those from Alaska, because that state doesn’t release its election results by county. As a result, First Nation isn’t included in this analysis, but we know based on precinct results that these areas voted overwhelmingly for Democratic candidates in all three elections. The rest of the state is divided between Far West and Left Coast, but the vote counts and party preferences are unlikely to alter the overall count “by nation” for these contests.

This analysis also adds to Republicans’ concerns about their prospects in this year’s midterm elections, especially in the U.S. House, where Democrats are likely to benefit from a reverse wave across the Midlands and Yankeedom. How many Trump allies are swept away in the storm remains to be seen.