Demographics, not Destiny?

In November 2021, Republican Glenn Youngkin won the Virginia gubernatorial election amid record high voter turnout for an off-year election despite President Joe Biden winning the state by double digits just a year before. The result shook the political world: for over two decades, the conventional wisdom in American politics held that higher voter turnout benefits the Democratic party. Since the mid-1990ss, many experts echoed philosopher Auguste Comte’s famous maxim that “demographics are destiny”, arguing that nonvoters skewed left politically and that higher turnout would by extension advantage the Democratic party. Youngkin’s victory is the latest in a string of recent elections that have undermined assumptions about the effects of higher turnout. With the implementation of new voting restrictions in many Republican-led states, it is more important than ever to determine how and if voter turnout levels influence election results. Is it possible that by modifying election laws to make it more difficult to vote, Republicans are actually shooting themselves in the foot?

Following decades of sluggish turnout, the 1990s and early 2000s saw a slew of studies assessing the demographics of nonvoters and their implications. A 1999 study from a team at Texas A&M found that areas with higher levels of racial and ethnic diversity tended to have lower voter turnout. Since nonwhite voters typically vote overwhelmingly for Democratic candidates, the study indicated that higher voter turnout in these more diverse areas would likely benefit the Democratic party. Likewise, a 2002 study by a team in California titled “New Perspectives on Latino Voter Turnout in the United States” found that Latino voters tended to vote at lower levels than other ethnicities, and that those who did vote tended to have lived in the United States longer and were generally more wealthy. As a result, the study implied that the large “unmobilized” population of Latino voters skewed heavily Democratic, which contributed to the formation of the “Demographics are Destiny” argument. During the time these studies were conducted, American elections were largely won and lost in white, blue-collar areas of the country, with Republicans dominating in well-educated suburbs and rural areas while Democrats overwhelmingly won urban cores, areas with high rates of worker unionization, and areas with high Black populations. In 2008, Barack Obama was elected President in a near-landslide, mobilizing young and nonwhite voters at record rates while still doing well with more traditional Democratic constituencies. The narrative seemed true: as more people voted and the country became more diverse, Democrats would gain an ever-larger advantage in US elections.

However, such an advantage never materialized. In 2020, despite the highest turnout in a century, Joe Biden barely won the 2020 election. In counties with some of the largest increases in voter turnout, former President Trump held his own or even outpaced Biden. The long-term signs were even more dire among nonwhite voters; though Biden still won Hispanic voters by a 3-2 margin, he lost significant support in areas like South Florida and Texas’s Rio Grande Valley, which are traditionally low-turnout areas with supermajority Hispanic populations. Despite these headwinds, Biden did still win the election, largely because he made inroads amongst more affluent and educated voters in urban areas; his overperformance in the Atlanta and Phoenix suburbs put him over the top in Georgia and Arizona and won him the election.

Given these geographic variations in election outcomes, it is worth assessing if the politics of nonvoter populations differ by region, a question tackled by the New York Times and Siena College in some landmark polling conducted in 2019. They found that nonvoters skewed younger, nonwhite, and less educated – all demographics that have historically been strong for Democrats but where Republicans are making increasingly effective inroads. The most notable finding was that, though the partisan leanings of nonvoters are largely evenly split nationwide, they vary widely by region. In the sunbelt (a group of rapidly-growing states in the southern US), nonvoters were far younger and less white than their cohort nationwide, and were also more likely to identify as Democrats. Conversely, nonvoters across the Midwest and Northeast were much more likely to be white and working-class than their states overall, skewing heavily Republican. Since the Times-Siena survey, other pollsters have produced similar data. The broad takeaway is that higher turnout nationwide would indeed benefit Democrats in states they are targeting (Florida, Texas, Arizona, etc), but potentially with the cost of states they currently win by slim margins in the Midwest. The 2020 election bore out this theory; amidst high turnout, Biden overperformed when compared to Hillary Clinton nationwide by about 2.5 points, but that effect was far more pronounced in the Sunbelt and nearly nonexistent in the Midwest.

The most obvious counterpoint to this trend is Virginia in 2021. The state’s nonvoter demographics skews Democratic and Democrats expanded voting access in the state during their time in power. Yet despite high turnout, Republicans won. The explanation for the unexpected result is differential turnout – voter turnout increased more decisively in rural areas than in urban ones, allowing Governor Youngkin to win despite the state’s Democratic lean. Had voter turnout increased in urban areas at the same level as it did in more rural constituencies, Democrats would likely have comfortably retained the governorship. This differential turnout raises broader questions, namely had Democrats not made mail-in voting more accessible, would Youngkin have been able to ride the surge in rural turnout that he did? Voting tends to be more difficult in rural areas than urban ones, as numerous studies have shown that proximity to polling places is correlated with higher voter turnout. Without more readily-available mail-in voting, it is quite possible that rural turnout in Virginia simply wouldn’t have risen as much as it did under the new rules.

Given the observed benefits of differential turnout facilitated by easier voting, it seems illogical that Republican legislatures across the South are now pushing to make voting more difficult, particularly by reining in mail-in balloting. State lawmakers crafting the new laws claim that they are apolitical efforts to ensure election integrity and not designed to benefit any party. If, however, the true reason behind the new restrictions is (as has been widely reported) to suppress the Democratic vote, these Republican lawmakers may be sleepwalking into a self-inflicted blunder, writing laws based on a flawed understanding of coalition and turnout data that hasn’t been true since the mid-2000s. It is highly possible that in the absence of mail-in voting, Democrats in urban areas will find an alternative way to vote, while more rural and/or older (and less mobile) Republican voters are hung out to dry. Georgia Republicans already fell into this trap; after passing a set of restrictive voting measures, the state party watched in horror as rural turnout tanked in the 2021 special elections for US Senate, resulting in Democrats flipping both seats and handing the party control of the US Senate at large. Given this consequential miscalculation, it is perplexing not only that Republicans in other states are now following Georgia’s lead, but that Georgia itself has doubled down and imposed even more restrictive rules. Republicans in the Midwest and Northeast are also attempting to make it more difficult to vote, which could be an even bigger unforced error than the one currently being made by their counterparts in the South: in the North, nonvoters skew Republican. Making it more difficult to vote would actively harm Republicans by disenfranchising voters that would support them at the polls if they could.

Democratic governments claim legitimacy by representing the popular will. In a country with historically sluggish turnout that has recently seen huge surges, knowing what the new popular will could look like should be important to all Americans. Further research is needed on the precise geographic impacts of new voting restrictions, but the data we have is clear: the country is evenly divided, and tampering with who can and can’t vote easily is liable to backfire in both directions. Many scholars expect record-breaking turnout in the 2022 midterm elections - where exactly those surges materialize could leave both parties kicking themselves the morning of November 9th.