Since at least the 2000 presidential election, pundits, scholars, and the general public have conceptualized the country’s partisan landscape using the blue states, red states, and swing states framework. But despite its ubiquity, this structure ignores how intrastate regional tensions and political competition imbue the divisions between red and blue America. Differences within states also anchor the long-standing urban-rural divide—a salient feature of American politics since the country’s founding.
Last year, the Brookings Institution Press published our book, Blue Metros, Red States: The Shifting Urban-Rural Divide in America’s Swing States. We argue that movement toward the Democratic Party in rapidly urbanizing suburbs is shifting America’s partisan fault line from a long-standing urban-rural divide to an emerging split between metro areas and the rest of state. The book covered 13 swing states that had a 2016 presidential race margin of 10 points or less and contain at least one metropolitan area exceeding 1 million residents.1
Within-state splits derive from four interrelated influences. First are sociocultural factors often determined by initial settlement patterns and that shape values and attitudes related to diversity acceptance. Next is demographic and economic sorting that concentrates diversity and economic productivity in the country’s largest metro areas. There are also sharp attitudinal differences due to rising negative partisanship and the growing saliency of cultural, racial, and other diversity-related issues. The attitudinal differences also have implications for perceptions of status loss, particularly among white residents. Finally, institutions such as the U.S. Senate and the Electoral College; policies and processes such as redistricting, home rule, and preemption; and metro area fragmentation often disadvantage urban and suburban interests.
The end result, as journalist Ron Brownstein noted after the 2016 election, has been a decoupling of demographic and economic power from political power.2 Blue Metros, Red States placed these tensions in a common geographic framework—the million-plus population metro area compared to the rest of a state’s population—and considered how these economic and demographic powerhouses navigate their state’s often hostile political terrains. By focusing our analysis on swing states, we examine these dynamics in contexts that determine the partisan balance of power at the federal level.
In this brief, we assess how well the book’s main hypotheses fared in the 2020 presidential and the 2021 Georgia U.S. Senate runoff elections. We find that the elections accelerated the trends we identified. Indeed, while it has long been the case that density and diversity predict Democratic support,3 the results suggest that another “D” can be added to that axiom: degrees. Democrats’ increasing support among higher-educated voters helped the party push further into the suburbs—particularly those that are diverse, urban, and well educated. In the election’s aftermath, the role of density and the exodus of college-educated suburbanites from the GOP has received significant attention.4 The same can be said for the migration of Democratic-leaning voters from places such as California into red state major metro areas (such as Austin, Texas) and the impact the new residents have on statewide election outcomes.5
Within-state geographic shifts are significant given that in many swing states, the metro area vote share is so large that if the Democrats win enough urban and suburban voters, Republicans cannot count on enough exurban and rural voters to win statewide elections. As such patterns strengthen over successive election cycles, the urban-rural divide that for generations nourished Republican majorities will be eclipsed by a divide between major metropolitan areas and the rest of the state—an emerging dynamic that favors Democrats.
A contrast exists between Sun Belt swing states, where the new pattern is ascendant, and Rust Belt swing states, where the components of the emerging Democratic coalition are less prevalent due to fewer diverse and college-educated voters. To prevail in the Rust Belt, Democratic candidates must cut Republicans’ margins in smaller cities and rural communities, while running up large wins in the million-plus population metro areas.
In 2016, Hillary Clinton was unable to patch together enough Rust Belt and Sun Belt states to secure an Electoral College victory. Four years later, Joe Biden successfully did so. Clinton was trapped between a weakening Rust Belt “blue wall,” and the yet-to-emerge Sun Belt blue states.6 Biden’s coalition, on the other hand, secured the blue wall in the Midwest while adding two Democratic-trending, fast-urbanizing Sun Belt swing states: Arizona and Georgia.
Metro area swings between the 2016 and 2020 elections
The analyses below summarize shifts in the 2020 presidential vote compared with 2016. The main unit of analysis is metro areas with a population over 1 million. We define metro areas using the census-designated metropolitan statistical area (MSA)7 and with the exception of Northern Virginia, refer to the MSA using the leading principal city instead of the formal MSA title.8 For each of the 13 swing states, data was collected from secretary of state websites after election results were certified. The data was aggregated for the counties in each of the 27 million-plus population metro areas in these states to generate the partisan voter margin. We also use the data to measure each metro area’s share of the total state vote. Data for counties that are not part of the metro areas was aggregated to generate the “rest-of-state” measure.
Table 1 summarizes the partisan margin for the 2016 and 2020 presidential elections in 27 swing-state metro areas and the inter-election shift in partisan support and change in vote share. Metro areas are grouped in terms of state and region using a Sun Belt/Rust Belt framework. For Sun Belt states, we group Florida and Texas together as “Big Sun Belt” states due to their large scale and the fact that each contains four metro areas with populations over 1 million. Eastern (Georgia, North Carolina, and Virginia) and western (Arizona, Colorado, and Nevada) “New Sun Belt” states also are clustered together. Table 1 further groups the “Rust Belt East” (Michigan, Ohio, and Pennsylvania) and “Rust Belt Central” (Minnesota and Wisconsin) states. Figure 1 uses data from the “difference” column in Table 1 and orders the 27 metro areas by their inter-election partisan shift from most Republican to most Democratic.
Among the 13 states, Florida clearly bucked the expectations of our Blue Metros, Red States thesis. While Biden’s margin did increase relative to Clinton’s in Orlando and Tampa, these improvements were offset by President Donald Trump’s surge in Miami, delivering him the Sunshine State’s 29 electoral votes by a greater margin compared to 2016. The result in Jacksonville was also a surprise; we predicted that it was poised to move toward Democrats, and in so doing, nudge Florida blue. While Biden narrowly won Duval County, including the city of Jacksonville (where Trump won in 2016), Republicans carried the metro area due to overwhelming support in surrounding suburban counties. Consequently, Jacksonville had the second-largest swing to Republicans among the 27 metro areas, following Miami (see Figure 1). Among Florida’s other million-plus population metros, only Orlando’s share of the state vote increased relative to 2016—meaning that Biden’s stronger performance in Orlando provided a bump in Democratic support compared to Clinton.
Texas provides the clearest example of our thesis. In 2016, Clinton only won Austin, while in 2020 Biden carried all four of the Lone Star State’s million-plus population metro areas. He picked up nearly 8 points in Austin and Dallas-Fort Worth, while making modest gains in Houston and San Antonio.
As detailed in Figure 1, Biden’s improvements in Austin and Dallas-Fort Worth were the largest metro area swings for Democrats in 2020. Moreover, Biden’s surge in Texas’s million-plus population metro areas carried even more punch given that the total vote shares increased in each. But even though Democrats cut Republicans’ margin from 16 percentage points to less than 6 over the course of the last two presidential elections, Biden fell short of winning the state, which we speculated was possible, but unlikely. Heavily Latino- or Hispanic-populated counties along the Mexican border—long a Democratic stronghold—saw Biden’s margins decline sharply compared with Clinton in 2016. Had the border county margins held steady, Biden would have lost Texas by perhaps less than 4 percentage points. If Texas eventually moves to the Democratic column, then the GOP’s path to 270 electoral votes narrows significantly, particularly given that Texas (like Florida) is poised to gain multiple electoral votes after the 2020 census-based reapportionment.
Among the three “New Sun Belt East” states, Biden further strengthened the Democrats’ hold on Virginia due to an even stronger showing in its three metro areas with populations over 1 million—two of which (Northern Virginia and Richmond) also gained in statewide vote share. Biden’s win is the party’s fourth consecutive presidential election victory in the state, a string that was last accomplished by Democrats in the 1930s and 1940s, when Franklin Roosevelt and Harry Truman carried Virginia five times in a row. In North Carolina, Biden cut Trump’s 2016 margin by nearly 2.5 percentage points, but still narrowly lost the state. Biden’s improved performance was driven by 5-point increases in the state’s two metro areas with populations over 1 million (Charlotte and Raleigh), each of which accounted for a larger share of the total vote compared to 2016.
Georgia is one of the five states that Biden flipped. Democrats nearly doubled the party’s margin in metro area Atlanta—the third-largest Democratic gain (see Figure 1) among the 27 million-plus population metro areas in the 13 swing states. In total, the party improved its performance by 5.5 percentage points, making Biden the first Democratic presidential candidate to win Georgia since Bill Clinton in 1992. Biden’s support in Atlanta was even more potent due to the increase in the metro area’s vote share relative to 2016. Georgia is on track to add an electoral vote this year, driven almost entirely by population growth in the booming Atlanta metropolitan area.
Farther west, Biden won all three of the “New Sun Belt West” states and became the first Democratic presidential candidate to win Arizona since 1996, when Bill Clinton carried the state with a plurality of the vote. Biden’s Arizona victory was highlighted by 5-point swings in the Phoenix and Tucson metro areas, as well as nearly doubling the raw vote advantage in Coconino County (home to Flagstaff) and improving upon Clinton’s margin in Apache County (including the Navajo and Hopi Native American reservations). Winning metro area Phoenix was even more significant given its increased vote share and the fact that more liberal Tucson’s statewide vote share decreased. Like Georgia, Arizona is due another electoral vote after reapportionment. And, as with Atlanta, most of the state’s population increase occurred in metro area Phoenix.
Biden more than doubled the Democratic margin in Colorado due in large part to a nearly 6-point swing in metro area Denver. Like Virginia (which also features high shares of college-educated voters and growing diversity), Colorado’s status as a swing state appears to be waning given the pace at which the Democrats are consolidating their support. And, again, Colorado will gain an electoral vote based mostly on suburban growth around Denver.
Despite favorable demography and strong Democratic showings in recent elections, Biden’s margin of victory in Nevada replicated Clinton’s. This inter-election stability is even more surprising given that Biden lost ground in Las Vegas, the state’s big blue metro area. Biden’s stronger showing in the state’s rural counties and in Washoe County (home to Reno) offset the increase in Trump’s Las Vegas support, even as its share of the vote increased by 1 percentage point. Las Vegas has both density and diversity but lacks degrees, thus failing to hit on all cylinders that drive Democratic gains.
Biden carried all of 2020’s Rust Belt swing states except Ohio, the only swing state besides Florida where Trump’s margin improved compared to 2016. Among the three “Rust Belt East” states, Biden made modest gains in the Detroit and Philadelphia metropolitan areas and larger gains in Grand Rapids and Pittsburgh on his way to flipping both Michigan and Pennsylvania. Trump won Ohio even though he lost support in two of the state’s three metro areas with populations over 1 million. Trump did perform better in Cleveland, but the metro area’s vote share decreased compared to 2016. Biden returned Wisconsin to the Democratic column due to increased support in Milwaukee and the smaller, but still relatively large and heavily Democratic Madison metropolitan area.
Although the Trump campaign targeted Minnesota after narrowly losing the state in 2016, the campaign’s efforts were to no avail. Biden won Minnesota by nearly 6 percentage points and boosted the Democratic margin in the Minneapolis-Saint Paul metropolitan area by nearly 7 points as turnout in the Twin Cities surged.
To put the inter-election data in context, Figure 2 compares the Democratic metro area margins against the Democratic margins in the rest of the state (all counties outside of million-plus population metro areas) and the state margin. Positive values indicate a Clinton or Biden advantage, while negative values indicate a Trump advantage. For states with multiple million-plus population metro areas, the “Metro” values combine the data for all such metro areas.
The 2020 and 2016 presidential election data demonstrates the degree to which Democratic victories in the 13 swing states depend upon the largest metropolitan areas, and just how much these regions’ partisan support differs from the balance of their state. In four of the five states that Biden flipped (Pennsylvania being the exception), his aggregate support increase in the million-plus population metro areas exceeded the statewide margin. Equally important, though, is the fact that Biden cut into Trump’s voter margins outside of the states’ largest metropolitan areas. While the shift was 1 point or less in Arizona and Wisconsin, Biden’s “rest-of-state” gains equaled or exceeded 2 points in Georgia and Michigan, and were more than 2.5 points in Pennsylvania—roughly twice as large as his million-plus population metro area percentage point gains. Only in Ohio and Texas did Trump’s margin improve outside of those states’ largest metropolitan areas.
In sum, while outsized support in million-plus population metro areas powers Democratic vote totals, the party’s candidates also need pockets of support in smaller cities and towns to push them over the top in statewide races. As we detail in Blue Metros, Red States, where these votes come from reflects the uniqueness of each state’s political, demographic, and economic landscape. For instance, in Minnesota and Wisconsin, the Twin Ports region on Lake Superior continues to deliver Democratic majorities. In Arizona, strong support among Native Americans in the northern part of the state was critical to flipping it blue, while high-amenity resort towns in the Rockies fortified Democratic support in Denver to deliver Colorado for the party. Biden’s victory in Pennsylvania—largely regarded as the 2020 “tipping point” state—was aided not just by swings in Philadelphia and Pittsburgh, but by squeezing out majorities in Lackawanna County (home to Scranton) in the east and Erie County in the west.
Georgia delivers the U.S. Senate for Democrats
Coming just nine weeks after the presidential election, the runoff elections to fill Georgia’s two U.S. Senate seats provide another opportunity to evaluate the partisan divergence between metro area Atlanta and the rest of Georgia. To facilitate these comparisons, the data presented in Table 2 reports the margin of victory for the state, the Atlanta metropolitan area, and the rest of Georgia, as well as the share of the vote that was cast in metro area Atlanta and the rest of state for the November presidential and January Senate contests.
While participation in the runoff elections approached 90% of the turnout of the general election, the drop-off was greater outside of Atlanta. As a consequence, metro area Atlanta’s share of the vote increased by just over a half a percentage point, as turnout in some Trump strongholds declined. Democratic candidates Jon Ossoff and Raphael Warnock won by larger margins than Biden, even though both candidates’ vote shares in Atlanta decreased. Ossoff and Warnock improved their showings outside of Atlanta relative to Biden by making small but consistent gains in vote share in the counties that contain Georgia’s smaller cities, including Augusta (Richmond County), Columbus (Muscogee County), Macon (Bibb County), and Savannah (Chatham County). Still, without double-digit margins in the Atlanta metropolitan area, neither Democratic candidate would have won and secured the party’s narrow majority in the U.S. Senate.
This data also suggests the degree to which turnout among some rural voters may be dependent upon Trump’s presence on the ballot. Akin to the 2018 midterm elections—when Democrats gained control of the U.S. House of Representatives, flipped two Senate seats, and picked up seven governorships—without Trump atop the ticket, some of his supporters stayed home. For Republicans contemplating if the party should double down on Trumpian-style populism or work to win back voters who moved toward Democrats in the past four years, turnout patterns in 2018 and in the Georgia runoff elections should be front and center in these conversations.
A new dynamic emerges for statewide elections
In 2016, Hillary Clinton cut Republicans’ margins in Arizona and Georgia, but she was unable to move both states’ electoral votes into the Democratic column. At the time, Republicans held all four Senate seats in Arizona and Georgia. Four years later, Biden won both states, and Democrats hold their Senate seats. Democrats also gained Senate seats in two other Sun Belt states—Nevada (2018) and Colorado (2020)—both of which have been carried by the party in the last four presidential elections.
As of 2020, Democrats hold all U.S. Senate seats in Arizona, Colorado, Georgia, and Nevada. In contrast, among the three Rust Belt seats that flipped back to the Democrats in 2020, only Michigan is represented by two Democratic senators.9
Biden would have won the presidency without taking Arizona and Georgia. Yet, without the inroads Democrats have made in the Sun Belt during Trump’s presidency, Republicans would have maintained majority control of the Senate. Indeed, even with a net gain of six Sun Belt Senate seats since 2016, the Democratic majority rests on Vice President Kamala Harris’s authority to break tie votes.
As we argue in Blue Metros, Red States, the outcomes in Arizona and Georgia are consistent with the emergence of a pan-metro-area identity anchored by a single large, dense, and more politically cohesive metro space. The same level of metro area cohesion does not exist in Florida or Texas (each with four metro areas with populations over 1 million), or North Carolina, where the states metro areas—Charlotte, Raleigh, and the Piedmont Triad—are splintered across a nearly 200-mile swath along I-85.
In fact, among the 13 states in our analysis, Biden won the five states—Arizona, Colorado, Georgia, Nevada, and Minnesota—where a single metro area constitutes over half of the state population. This also goes for blue states such as Illinois, New York, Oregon, and Washington.
In 2016, Clinton won every state in the conterminous United States where a single metropolitan area accounted for more than half the state’s total population—the only exceptions being Arizona and Georgia. In 2020, Biden swept all such states. In Blue Metros, Red States, we predicted that swing states with a major metro area exceeding half the total population were more likely to vote Democratic than states with more divided urban systems such as Florida, North Carolina, and Texas. The 2020 election validated our point, with deep ramifications for future statewide elections as the urban-rural divide that anchored many Republican victories gives way to a metro area/rest-of-state dynamic more favorable to Democrats.
Report Produced by Brookings Metro
- The 2016 presidential election margins in Iowa, Maine, and New Hampshire were less than 10 percentage points but these states do not have a metro area with a million plus population.
- Ronald Brownstein, “The Prosperity Paradox is Dividing the Country in Two.” CNN, January 23, 2018 (https://www.cnn.com/2018/01/23/politics/economy-prosperity-paradox-divide-country-voters/index.html).
- For an early analysis of the relationship between diversity, density, and Democratic voting, see Robert Lang, Thomas Sanchez, and Alan Berube, “The New Suburban Politics: A County-Based Analysis of Metropolitan Voting Trends since 2000,” in Ruy Teixeira ed., Red, Blue, & Purple America, (Washington DC: Brookings Institution Press, 2008), pp. 25-50.
- See, for example, Richard Florida, Marie Patino, and Rachael Dottle, “How Suburbs Swung the 2020 Election.” City Lab, November 17, 2020 (https://www.bloomberg.com/graphics/2020-suburban-density-election/).
- Matt Levin, “How California Expats are Helping Turn Texas Into a Battleground State.” Cal Matters, October 8, 2020 (https://calmatters.org/politics/2020/10/california-expats-texas-battleground-state/).
- Robert E. Lang and David F. Damore, “The End of the Democratic Blue Wall?” Brookings Mountain West, December 2016.
- For the six MSAs that extend into multiple states, we only include data for the counties that are in the state associated with each MSA’s principal city.
- For instance, we use Las Vegas instead of Las Vegas-Henderson-Paradise MSA. The only exception is we use Northern Virginia to label the Virginia components of the Washington-Arlington-Alexandria, DC-VA-MD-WV MSA.
- Wisconsin Republican Senator Ron Johnson will be up for reelection in 2022. Pennsylvania Republican Senator Pat Toomey, whose terms ends in 2023, declared last October that he will not run for reelection.