An Archer Daniels Midland plant sits on the banks of the Mississippi near downtown Clinton, Iowa. (Michael S. Williamson/The Washington Post)

As he ran for president, Donald Trump promised to lift up regions of the country that had been left behind by the economy. “I want to go into the neglected neighborhoods, the failing schools, the forgotten stretches of this nation, and unlock their potential for all of our people,” he said in September 2016.

The argument proved persuasive — many of the nation’s most economically distressed regions voted for Trump.

Now the early returns are in: In the first year of the Trump presidency, places that voted for Trump are doing better economically than at the end of the Obama administration. But that is not totally surprising, because the overall economy has continued to add jobs.

What may be more surprising is that not only are these counties adding jobs, but also job growth has accelerated the most in counties where Trump earned the most votes, according to a Washington Post analysis of Labor Department data.

In other words, even if the counties that supported Trump most are still struggling relative to the rest of the country, they’ve experienced the largest reversal of fortune.

It’s too soon to declare victory, however, and the current set of gains may well be short-lived.

But for now, the data indicate that the 10 percent of Americans living in counties that gave Trump the most support in 2016 saw average job growth of 0.6 percent in 2015-2016. They saw job growth of 1 percent in Trump's first year. Expand to counties that were home to the top 20 percent of his supporters — places where he got 60.6 percent of the vote or higher — and the acceleration was a still-perceptible 0.2 percentage points.

Meanwhile, the 10 percent of people living in counties that showed little support for Trump — which tend to do better economically overall — saw job growth slow slightly, going from 1.9 percent in 2015-2016 to 1.7 percent in 2017. The broader top 20 percent of anti-Trump counties saw job growth effectively stagnate.

(We're measuring February through January of each year, because presidents begin their tenures on Jan. 20. The late 2017 and January 2018 figures are preliminary and subject to revision).


When we base the analysis on a simple ranking of counties, not adjusted for population, the difference sharpens. Like the election maps that the president prefers, it gives equal weight to both lower-population small towns and rural areas, which tend to be red, and massive, dense coastal cities, which trend blue.

Those rural areas, towns and smaller cities that make up Trump's geographic base saw relatively little job growth in 2015-2016, but they jumped closer to the national average in 2017.


Trump's America

Trump attracted a different cross-section of America than previous Republicans, one that was heavily skewed toward counties with the slowest job growth.

As you can see in the chart below on Trump's share of the vote in 2016, there's a wide gap between the counties that are home to the 20 percent of people who provided Trump with the highest vote shares in 2016 (red lines) and the counties that are home to the 20 percent who supported him least (teal lines). Linger for a moment on the red line, which shows that these hardcore Trump-supporting counties have barely recovered in terms of jobs from the Great Recession.

Then consider the places that voted for Mitt Romney, the second chart. On the whole, they weren't defined by their economic distress. There's not much of a gap in the economic recovery between those who supported him most (red line) and those who supported him least (teal). Romney's message didn't resonate as much with Americans living in counties that are struggling the most.

If you want to drive home the difference, just consider the third chart. The counties with the biggest shift from Romney to Trump (red line) are still trailing far behind the rest of the country and may never regain jobs lost during the recession.


Why the shift

Economists broadly agree a politician’s ability to create jobs in the short term is overblown. (The exception is a large spending bill, such as the 2009 fiscal stimulus, but that did not occur in 2017.) And Trump hasn't yet had time to fully see the effects of long-term policy that might bring sustained growth to rural America. He initiated a process of regulatory overhaul and passed tax cuts at the end of the year, but any gains from those policies probably won't show up in any meaningful way in the 2017 figures. It's nonetheless possible Trump's victory has contributed at least in part to the faster improvement in regions where residents voted for him most.

That's because economic expectations of Republicans turned drastically more positive after Trump's election.


Those higher expectations alone might nudge growth rates upward as people are more willing to spend and invest, said Jed Kolko, chief economist at the job-search site Indeed.com.

“The industries that have seen a rebound ... tend to lean red,” Kolko said.

It's not just expectations

According to government data, six industries are overwhelmingly concentrated in Trump territories: mining, support activities for mining and oil and gas extraction, livestock production, logging, and the manufacturing of textiles and wood products.

The prominence of mining and its support services on that list, in particular, hints as to why those territories are doing better.

Oil and commodity prices collapsed in the second half of 2014, driven by slowing global demand, and U.S. production plummeted. Counties that depended on mining and drilling shed jobs throughout 2015 and 2016.

“In some of these smaller areas, the economy outside of oil and gas is growing pretty slow,” said Jason Brown, assistant vice president and economist at the Federal Reserve Bank of Kansas City. “When oil prices declined by over 70 percent between 2014 into 2016, oil and gas activity declined pretty quickly in areas that were less productive.”

Oil and commodity prices crawled up off the floor in 2017 as global economic growth picked up pace and, more importantly, production started to recover. Companies became more efficient and shifted their efforts to more cost-effective wells and regions. As Brown explains, production, not price growth, leads most directly to jobs in the industry.

The return of mining and petroleum-related jobs goosed growth rates in Trump’s favor. The list of counties that bounced back the most last year is dominated by mining towns and shale basins, from Wyoming and South Dakota to Texas, Oklahoma and even parts of Appalachia.

It’ll be tough to sustain

Oil, gas and mining are cyclical, but at least cycles include, by definition, some upswings. As the U.S. economy continues a long structural shift toward industries and services that tend to be focused in urban areas, it’s one of the few rural industries that can foster a 2017-like boost.

Of the six industries that are most concentrated in Trump territory, the Labor Department estimates five will lose jobs over the coming decade. There's just one lone, cyclical, exception.


So while 2017 was relatively strong, the rural areas that supported Trump most lack the underlying economic structure to overcome the ever-present head winds of an aging population and receding industries.

“The trend over the past year or so is, in many ways, the opposite of what has — and will probably return to be — the longer-term trend,” Kolko said.


Remember that Trump’s apparent success is relative to a particularly anemic baseline. His strongest supporters are still seeing the country’s slowest job growth, even if it’s faster than what they had become accustomed to by the end of the Obama years.

And if they’re growing slower than the rest of the country, then not only are they not making up lost ground, but also they’re falling further and further behind.

Methodology:

For most of the analysis, we split U.S. counties into five groups based on the share of their votes that Donald Trump earned in 2016. They are weighted by the size of county labor forces. This makes the groups about equal in population and corrects for the fact that Trump won a large number of low-population counties, while Hillary Clinton won a much smaller number of high-population counties.

Select counties and equivalents have been combined for historical comparability. Because of how the state's votes are reported, Alaska's boroughs were given an equally weighted portion of the state's overall vote.

Most of the data comes from the Labor Department's Local Area Unemployment Statistics. They are the most recent figures available, but they're revised each month, with annual benchmark revisions occurring in mid-April. Industry-specific data at the county level comes from both the Census Bureau's County Business Patterns and the Labor Department's Quarterly Census of Employment and Wages.

To correct for seasonal variation, the numbers are all reported as annual averages. In our annual-change analyses, we defined “years” as beginning in February and continuing through January of the following year, because presidents don't begin their terms until late January.