Two-thirds of the House seats Democrats picked up in 2018 shared a common characteristic: They were in suburban areas.
That the suburbs would probably be a battleground in 2020, though, was obvious from the results in 2016.
You’ve no doubt seen some iteration of the map below at some point. It shows the results from four years ago at the precinct level, using data compiled by researcher Ryne Rohla. It suffers from the standard land-isn’t-voters problem of other such maps, but, for the sake of this demonstration, let’s set that aside. Lots of red areas that backed Trump. A lot of blue regions that preferred Hillary Clinton.
If you look closely at that map, which fades to lighter colors in more closely contested areas, you can probably see where this is headed. When we zoom in on the upper Midwest, a region central to Trump’s victory, that’s made more obvious. Islands of blue ringed by lighter shades — closer contests.
But let’s be explicit, presenting that same information in a different way. The national map below, instead of showing results, highlights places where the race was most contested. White areas had either Trump or Clinton winning by more than 10 points. Purple areas had narrower margins, with the closest contests unfolding in the darkest-colored precincts.
Here’s that same Upper Midwest region.
And here, that region with the boundaries of cities added.
Even that level of detail, though, doesn’t make the point entirely clear.
Drilling down further, we look at northeastern Ohio and western Pennsylvania.
Here are the areas where the 2016 race was the closest.
And here’s how those regions compare with nearby cities.
An animation illustrates the point even more clearly.
In this region, the places where the results were closest were almost uniformly just outside city boundaries. They were, in other words, in the suburbs.
Let’s move farther west. Here’s the Chicago region.
And here’s how city boundaries compare with close precincts.
Chicago itself is almost entirely free of contested precincts. Head away from the lake, though, and you quickly hit a scattershot ring of places where the Clinton-Trump vote was narrowly determined.
If we take all 172,000-odd precincts nationally and divvy them up by the National Center for Health Statistics (NCHS) index of county types, we see that counties on the outskirts of large metro regions were the most common location of contested precincts (the darker bar) and that those counties also had the highest density of close precincts.
As that graph suggests, though, the correlation between suburbs and close precincts isn’t perfect. In some places, the effect is stark, like near Denver.
In other places, it’s spottier, like in the Southeast.
(Clicking these maps will allow you to explore them in closer detail.)
In Florida, the delineation isn’t as neat. A number of precincts in Tampa, Melbourne (east of Orlando) and Port St. Lucie (north of West Palm Beach) were closely fought.
In Southern California, things looked a bit different. While Los Angeles wasn’t home to many contested precincts, San Diego — a more conservative city — was. But the whole surrounding region, including immediate suburbs and the more rural parts of the Inland Empire closer to the Nevada border, saw closer contests.
That was, to some extent, a function of how even more-Republican Southern California was open to Clinton’s candidacy. That led to closer contests in places that had, in the past, been solidly Republican.
In 2018, it also meant that Orange County, an area between L.A. and San Diego that was once reliably Republican, saw its delegation become solidly blue. A year later, its voter registration numbers followed suit.
While the county overall is a “large central metro,” according to the NCHS, four of Orange County’s six congressional districts are identified as suburban in CityLab’s density index. If those suburban areas have gotten less friendly for Trump since 2016? That’s potentially a problem.
If they’ve gotten worse than they were in 2018, it’s a crisis.