Skip to main content
Menu
Where America Lost And Gained Population Could Help Democrats In Redistricting

With the release of block-level data from the 2020 census, we now have a much clearer picture of how and where the U.S. population has grown — and shrunk — over the past 10 years. And while Republicans are largely setting the terms of the redistricting process that will ensue from this announcement, the data throws a much-needed lifeline to Democrats.

It’s not because of the country’s increasing racial diversity, though. Sure, under current electoral coalitions (where white voters are more likely to vote Republican and voters of color are more likely to vote Democratic), it’s arguably better for Democrats if the nonwhite population grows. But even in a country where only 58 percent of residents are non-Hispanic white, the 2020 presidential election was still very competitive. And electoral coalitions can change — for instance, Republicans may continue to gain ground among nonwhite voters in future elections.

Instead, the good news for Democrats came in which parts of the country gained and lost population since the 2010 census. Despite the country’s overall population increasing by 7.4 percent, rural areas — the reddest parts of the country — have steadily lost residents over the past 10 years. The average county with a FiveThirtyEight urbanization index below 81 lost 3.1 percent of its population between 2010 and 2020. This encompasses the 1,430 most rural counties in America — 1,302 of which voted for former President Donald Trump in 2020, and only 127 of which voted for President Biden.

By contrast, the fastest-growing parts of the country are the suburbs. The average county with an urbanization index between 11 and 13 (spanning from sparse suburban areas like Roanoke County, Virginia, to dense suburban areas like Union County, New Jersey) grew by 9.6 percent between 2010 and 2020. Most of these counties (194 out of 264) voted for Biden, and if the trends of the 2018 and 2020 elections continue, they will only get bluer.


I don’t buy that the 2020 census is good news for Democrats: Nate Silver

America’s cities grew at a healthy rate too, which was something of a surprise compared with pre-census estimates. The 18 most urban counties or county equivalents in the country — those with a FiveThirtyEight urbanization index above 13 (including San Francisco, Philadelphia, four around Washington, D.C., and eight around New York City) — grew by an average of 8.4 percent — and all but one of them voted for Biden in 2020.2

Overall, the average county that voted for Biden boosted its population by 3.4 percent between 2010 and 2020, while the average Trump county grew by just 0.2 percent. Strikingly, 370 out of the 538 Biden counties (69 percent) gained population, while 1,468 out of the 2,574 Trump counties (57 percent) lost residents.

These population shifts are important because they will shape the nation’s political geography for the next 10 years, as congressional district lines will be drawn based on this data. As such, it will now be that much harder to draw congressional maps that favor Republicans: If a heavily Democratic metropolitan area is overflowing with people, map-drawers need to put them somewhere — which may necessitate creating an extra (blue) district for that city, or at least including more blue precincts in a neighboring Republican district than the GOP would like.

At the other end of the spectrum, because congressional districts must have roughly equal populations, many depopulated rural congressional districts will need to expand — either cannibalizing a neighboring rural district (which could reduce the number of safely red seats) or encroaching on the outer suburbs. Because America’s exurbs are still fairly red (three-quarters of counties with a FiveThirtyEight urbanization index between 10 and 11 voted for Trump in 2020) this could reduce the number of Republicans living in a current swing district — a sort of redistricting domino effect.

Rep. Conor Lamb standing at a podium with a crowd behind him announcing his retirement

related: What All Those House Retirements Mean For Democrats So Far Read more. »

We can see how this domino effect might unfold by looking at which congressional districts, rather than counties, are over- and underpopulated. To see this, we can compare the 2020 population of each current district to what its population should look like after the reapportionment process. (Remember, because each state’s districts must have the same population, the target population of a state’s districts is the state’s 2020 population divided by the number of congressional districts it received in reapportionment.)

In the map below, orange districts will need to add population (i.e., they will need to expand), while purple ones will need to lose population (i.e., they will need to get smaller).

The most obvious examples of districts that will need to expand are those in states that lost a seat in reapportionment, like West Virginia. But take a look at New York, which also lost a seat. The districts that will need to expand the most are all in upstate New York, while four New York City-based districts — the 5th, 8th, 10th and 12th — will actually need to shed residents. Since those districts are all extremely blue, that will provide mapmakers with an opportunity to add Democratic votes to nearby districts that might not be as safe, such as the 11th. 

Similarly, in Illinois, the districts around Chicago are those that are already closest to their target post-redistricting population, whereas the downstate 12th and 17th will need to add the most people. And because Illinois also lost a seat in reapportionment, it’s most likely that downstate will lose representation. Since five of Illinois’s six current downstate representatives are Republicans (and Democrats control the map-drawing process in Illinois), that makes it a safe bet that Republicans will draw the short straw. 

At the same time, patterns in population growth could blunt Republican gains as well. For example, the two most overpopulated current districts in North Carolina — which is gaining a seat — are the deep-blue 2nd (which covers Raleigh) and 12th (which covers Charlotte). Republicans are in charge of redistricting in North Carolina, but they would have to engage in some complicated gerrymandering gymnastics if they want to blunt the political power of these metro areas. Similarly, the Cook Political Report’s Dave Wasserman has estimated that explosive population growth in urban areas like Austin could limit Texas Republicans to a map where Republicans have the advantage in only 25 of the state’s now-38 districts, whereas previously they might have been able to draw 27 Republican-leaning seats.

However, as Wasserman’s estimate makes clear, the population trends from the 2020 census are not enough to cancel out any gerrymandering efforts. Rather, the census numbers are merely a silver lining for Democrats. They still remain at a serious disadvantage in the 2021 redistricting cycle simply because Republicans will control the redrawing of more congressional districts. Republicans will have full control over the redrawing of 43 percent of congressional districts, and Democrats will have full control over the redrawing of just 17 percent. (Neither party enjoys full control of redistricting for 38 percent of districts, while the remaining 1 percent are at-large districts and won’t be redrawn at all.) 

So Democrats got some good news with the specific census data released earlier this month, but they are still in an unenviable position. As long as redistricting remains such a subjective process, who draws the maps will remain more important than the underlying data.

Geoffrey Skelley contributed research.


Why has Biden’s approval rating fallen? | FiveThirtyEight Politics Podcast


Cuomo is resigning. Now what? | FiveThirtyEight Politics Podcast

Footnotes

  1. FiveThirtyEight’s urbanization index is a measure of how densely populated a county or county equivalent is based on the average number of people living within a five-mile radius of a given resident. The analysis in this article includes Washington, D.C., but excludes Alaska, which does not have county-equivalent-level election results available.

  2. The Trump-voting exception was Richmond County, New York, which is Staten Island.

Nathaniel Rakich is a senior elections analyst at FiveThirtyEight.

Comments