Commentators like to categorize American politics by space. The United States, we are told, is comprised of conservative rural areas, liberal urban areas, and hotly contested suburbs. But just like maps that color whole states red or blue, these labels paint a false uniformity. It’s the variation among them we most need to see to understand political change.
Our divides are widening, and density is only part of the story. In some rural spaces, information-economy consultants freelance using broadband. In other rural places, stressed moms drive miles for work shifts that change without notice. The communities that fall between urban and rural are even more varied in ways “suburban” doesn’t begin to capture. The suburbs conjures images of mown lawns and two-car garages. But while some suburbs have subdivisions and golf courses, others have foreclosure-scarred streets and strip malls. There are places where vacation bible school is thriving—and places where even churches have closed.
Reports that the suburbs swung left in 2018 thus shed little light. Atlantic Media’s CityLab offered a little more detail, categorizing congressional districts based on density ranges. According to this metric, “sparse suburban” districts swung to Democrats in the midterms by more than any other category. “Rural-suburban mix” districts and “dense suburbs” were close behind. “Pure urban,” “urban-suburban,” and “pure rural” districts saw smaller leftward shifts. But these labels still don’t show how fortune and hardship shape the politics of a district.
To try and capture how they do, analysts like Dave Wasserman point to the divide between Whole Foods and Cracker Barrel counties. In 2016, Donald Trump won 76 percent of counties with a Cracker Barrel and just 22 percent of counties with a Whole Foods. That’s a vivid mental picture. But most of America falls outside it. Roughly 200 of the 3,200-odd counties in the United States have a Whole Foods. No more than 650 have Cracker Barrels. A swingy, politically diverse state like Pennsylvania has fourteen Whole Foods, twenty-five Cracker Barrels, and sixty-seven counties. Six Pennsylvania counties have both, and forty-three have neither.
Here’s a better way to track hardship across space: SNAP-authorized dollar stores. There are over 30,000 dollar stores in the United States, and about two-thirds of them are authorized to accept Supplemental Nutrition Assistance Program payments. Each SNAP-authorized dollar store location reflects a series of specific locational assessments: that rent here is low, low-income demand is high, and the number of folks on food stamps who will rely on this store for groceries will make up for the cost of stocking goods that meet USDA authorization criteria. It’s no surprise, then, that congressional districts with more SNAP-authorized dollar stores are less well off, with lower life expectancies, lower educational attainment rates, and lower median incomes.
There are fewer SNAP-authorized dollar stores in dense districts in part because more people live within range of any given store. But other factors are also responsible for keeping the number of dollar stores low in some areas, including high retail rent, strong local merchants, and public policies that promote better food options. Because SNAP-authorized dollar store locations crystallize all these dimensions, classing districts by the number of stores they contain creates sets that are more homogenous than categorization by density, while still capturing density as a key component.
This is evident in the above charts, which both utilize the Social Science Research Council’s income index to measure wealth. The first chart groups districts by CityLab categories. As it shows, each CityLab category includes districts with a wide range of median incomes. The second chart groups districts by the number of SNAP-authorized dollar stores. It creates sets with a far smaller spread.
In the second chart, the dots’ colors reflect which party held the seat after the 2016 election, conveying the political stakes of hardship in non-urban America. There’s a clear pattern. Districts with many dollar stores are more likely to be conservative.
That’s true even within suburban areas.Virginia’s second and eleventh congressional districts, for example, are both classed by CityLab as “dense suburban.” But according to 2008-18 data, VA-11—Washington D.C.’s bedroom suburbs—has only five SNAP authorized dollar stores. VA-02, home to Norfolk, Virginia Beach, and the two poorest counties in Virginia, has forty-three. In VA-11, life is good. In VA-02, it’s not great. The median household income in VA-02 is only two-thirds as high as in VA-11, and the portion of adults with advanced degrees is less than half as high. In 2016 congressional races, Democrats carried VA-11 by a forty-four point margin. Republicans took VA-02 by twenty-three points.
Republicans’ slow rise to dominance in the kind of places dollar stores have colonized—more rural, less wealthy, more white, more evangelical—stuttered with the 2008 recession but then surged after Barack Obama’s election. Donald Trump’s campaign, with its populist rhetoric and rejection of Republican orthodoxy on trade and entitlements, furthered the trend. Trump earned a smaller vote share than Mitt Romney in low-dollar-store districts, but made up for it in places where the number of dollar stores is high.
The rural districts to the right of this chart have long been home both to voters who depend on programs like SNAP to help put dinner on the table, and to voters who resent thoseprograms. The former tend to vote Democrat when they vote at all. The latter are thisnation’s staunchest Republicans. In 2016, Donald Trump convinced enough of the former that he might be their champion, or at least that Hillary Clinton would never be, that he carried these districts by unprecedented margins.
This chart shows the number of SNAP-authorized dollar stores for congressional districts, sorting each district by its presidential votes in 2012 and 2016 and color-coding each by its 2016 congressional districts. Starkly, the districts that Hillary Clinton sliced away from the Romney were systematically those home to the fewest dollar stores. Surprisingly, Obama-Clinton districts cover a pretty broad range, in part because there are a handful of heavily African American rural districts in the Deep South with high numbers of dollar stores. Meanwhile, the 2012 Obama districts that Donald Trump won over were largely in the 30 to 60 dollar store band. These were New England and Midwestern districts with small cities and waning numbers of union jobs.
Above, we have mapped out how these districts voted in the midterms. Purple diamonds represent seats that Democrats flipped and orange ones represent seats that Republicans flipped. At first glance, these results might suggest Democrats increasingly speak mostly for places with low numbers of dollar stores. The party entered the 2018 midterms holding two-thirds of the 218 districts with thirty or fewer stores, and it picked up an additional twenty-nine seats in this category. By comparison, Republicans went into the midterms holding three-fourths of the 217 districts with over thirty SNAP-authorized dollar stores and lost a net of just eleven seats in that range.
But Democrats flipped few seats in high-dollar-store districts because they started from far behind. In two dozen of them, Democrats did not even field candidates in 2016. But in 2018, that changed. The chart above highlights those newly contested seats with diamonds. Most of them are disadvantaged rural districts where Democrats have not been competitive since Ronald Reagan first won office. Many are southern districts that went red as party positions on civil rights switched. In 2018, Democrats stepped back up to compete in these places, and did so with a nationalizedbrand that is avowedly progressive on issues of racial justice. These candidacies are a leading indicator of Democrats’ new dynamism in places far beyond affluent suburbs.
But it isn’t just that more Democrats ran in dollar store country. The above plot shows congressional districts where Democrats ran in both 2016 and 2018. Each dot’s color shows the party that held the district going into the midterm election. Each dot’s location shows the magnitude of the congressional vote swing. The breadth of the shift is clear. Very few districts moved towards the GOP in 2018. Those that did were almost entirely in (and remained in) Democratic hands. Rather, even in districts with many dollar stores, congressional votes totals moved somewhere between a little and a lot towards the Democratic candidate.In fact, in 2018, Democrats improved their vote share as much in high-dollar-store districts as they did in ones with the fewest stores. The party’s vote share improved most in the mid-to-high dollar store districts in between. They even managed to win in VA-02.
Up through the 2016 elections, the ongoing geographic concentration of prosperity drove a widening political divide. Democrats were positioned as caring about the kinds of people who live in urban areas, and the kinds of poverty and inequality they face. That left Democrats vulnerable to Republican claims that they didn’t care about the kinds of people who live in small town and rural areas or the hardships they face. The social infrastructure through which Democrats once made their case in dollar-store country, like unions and working-class churches, was battered by the same grim trends that favored dollar stores’ arrival.
So how did Democrats make a comeback? In place after place, in the wake of Donald Trump’s election, local progressives decided they could no longer wait for someone else to fix a political system they saw as broken. They stepped forward, found each other, created and used online resources, and took hands-on political action. Where Democrats’ local infrastructure had most atrophied, the new presence was most impactful.
New or re-energized progressive groups in red districts have repopulated local Democratic committees and altered the ecosystem for campaigns up and down the ballot. These groups aided candidate recruitment and fundraising, knocked on doors and made calls, and encouraged campaigns to come hold events in locales they might otherwise have skipped.
Within Iowa’s deeply conservative fourth congressional district, for example, a dozen separate grassroots groups joined the Indivisible network. Meanwhile, another new group (one not even listed on the Indivisible site) helped Democratic candidate J.D. Scholten host a campaign stop in Pocahontas, Iowa, population 1,700. The stop featured a rally with forty attendees as well as a coffee-shop phonebank. That may not sound like a revolution, but it’s local activism like this that helped Scholten almost defeat fifteen-year incumbent Steven King.
What happened in Scholten’s seventy-eight dollar store district is happening in places like it across the country. Regional politics have been renewed by activists’ passion, time, and treasure. They supported Democratic candidates who ran for positions large and small—with locally framed messages to match. As a result, GOP-skeptical younger voters far outside metro cores found persuasive Democratic alternatives in front of them.
The distance between countryside and country club has not collapsed. Choosing cans carefully on the dollar store aisle to make sure you don’t go over your EBT limit remains a common experience in some communities. It remains unknown in others. But Democratic organizing is now underway in local hands in districts at both ends of this gulf, and across its vast middle. For Democrats, the task ahead is translating this into sustained electoral gains and policy action that will reduce the disparities dividing America.