ALG Research peaks over the picket fences to examine the newly forming Democratic coalition

March 20, 2019
Following the 2016 Election - and persisting into the 2018 Midterms - pollsters, pundits, and handicappers have placed a microscope on the significant gains that Democrats have made in suburbia, as well as the significant losses incurred in rural portions of the Midwest. In the 2018 Midterms, many of the same trends from 2016 persisted - Democrats continued to grow their margins in the suburbs but were unable to produce Obama-era successes in more rural communities.
Suburbia broke toward Democrats, GOP made gains in rural Midwest
2012 to 2016 Presidential Democratic margin shift (ALG Research)
Using CityLab's widely cited Congressional Density Index (CDI) - which breaks down congressional districts into six density-based categories (Ranging from "Pure Rural" to "Pure Urban") - and Daily Kos' 2012 and 2016  Presidential election data, we looked at what the 2018 Midterms told us about density's increasingly important relationship with congressional districts and partisan politics.

O f the 43¹  House seat pickups in 2018, nearly two-thirds of Democratic flips were found in suburban district s. ²  Conversely, factoring in Republicans flipping Minnesota's 8 th Congressional District, Democrats only netted one seat in the most rural district classification. And while congressional Democrats' highest average shift in 2018 relative to 2016 was found in rural districts, this is because these districts so dramatically underperformed historical benchmarks in 2016 - as opposed to  Democrats making any significant inroads with rural voters. In fact, congressional Democrats actually lost ground in rural districts in 2018 relative to Obama's 2012 margins.

In addition to supplying Democrats with the majority of their newly-held seats, suburban districts have also experienced sizable shifts from 2012  and 2016, with suburban districts even further expanding their already-considerable gains from 2016 to 2018.
Tide turning in suburban and rural areas: 2018 pickups by district type   
*The table above excludes districts that were uncontested by a major party candidate in the 2018 Midterms order to maintain consistency and compare apples-to-apples with presidential numbers. A total of 394 districts were included in this analysis. For the flipped seats column, we've excluded flipped districts in Pennsylvania due to recent redistricting. (ALG Research)
From both the observations we've made using CityLab's district classifications and the observations we've made from other measurements of d e nsity ³,  we can say with a relatively high degree of confidence that  urbanity and density are increasingly predictive of Democratic performance, and 2018's results solidified many of the trends we saw in 2016 rather than regressing to the 2012 baseline.  We anticipate that population density will continue to be a determinative component of electoral politics in the 2020 elections and beyond.
Looking forward
In order to hold and maintain majorities, Democrats will need to build a coalition of both suburbanites and rural voters. And while winning these two groups is not mutually exclusive, they are two groups that often differ on key policy positions - ranging from trade to immigration to domestic spending.

So what does this mean for candidates?
  • Know your district: The urban-rural divide is not binary, and there is a diverse array of persuadable voters living in small-to-medium metros, exurbs, and working-class suburbs that can - and often do - determine who ultimately gets elected.
  • Ride the wave: While it's unclear if the new suburban Democrats are here to stay, or just seeking refuge from Trump's brand of Republicanism, capitalize on the trends that we know are occurring. There will be serious electoral dividends for Democrats by engaging, persuading, and turning out voters in suburbia. 
  • Stop the bleeding: Although it is clear rural voters have materialized behind Republican lines, this doesn't mean that mitigating losses won't determine an election's outcome. There are numerous Democratic members of Congress who would not have won in 2018 had they not made substantial improvements on the 2016 Democratic performance in rural parts of their districts. 
¹  This excludes Pennsylvania's four districts due to redistricting in 2018, so the proportion of 2018 House seat pickups is calculated out of 39 total Democratic pickups.

²  Of the six CityLab Congressional Density Index categories used, "Dense Suburban" and "Sparse Suburban," are the most suburban district types, which are what is being referenced here.

³  We conducted a bivariate regression analysis using logarithmically transformed population per square mile data to predict 2018, 2016, and 2012 election margins, which yields a statistically significant relationship in all cases. The relationship (R²) was strongest in 2016 (.4818) but was stronger in 2018 (.4374) than in 2012 (.4051). 2016's stronger relationship with density over 2018 is partially a function of candidate quality and incumbency advantages (For instance, in MN-07 and WV-03 - both rural districts - Democratic candidates performed much better than a "Generic Democrat" might have).