Skip links


The 2012 election provided two powerful reminders about the electoral implications of overly-concentrated Democratic voters. First, the Republicans held their U.S. House majority, won in 2010, despite the fact that the Democratic candidates in the 435 House districts received more votes than their Republican opponents. Second, these House results were echoed by Democrat Barack Obama’s defeat of Republican Mitt Romney by nearly four percentage points nationally despite the fact that Obama carried fewer House districts than Romney did (211 to 224 based on the most recent congressional maps). Whether by dint of the nonpartisan self-segregation of voters or partisan gerrymandering, Democratic voters are distributed inefficiently in U.S. House districts.

Of course, House districts and state legislative districts can be redrawn each decade in ways that concentrate or diffuse voters to the electoral benefit of either (or neither) party. What do not change are borders for state, county, and local jurisdictions that elect officials and that also may happen to confer an advantage on one party or the other. For example, as I document in my latest book, The Stronghold, for the better part of a half-century Republicans have enjoyed inflated representation in the Senate by virtue of their greater strength in the small states. This was not always the case: In the mid-1950s, when Republicans regularly held Senate seats in California and New York, the Democrats enjoyed inflated Senate representation. Because state boundaries are fixed, however, barring the creation of new states via admission of new states or splitting of existing states into two, electoral advantages arise only if certain types of voters migrate or grow at disparate rates that favor one party.

Most remarkable are the variations in size and partisanship of America’s counties. The GOP’s recent, state-based advantage in the Senate pales against the far more dramatic partisan effects of population concentrations at the county level. Today, roughly half of all Americans live in the 144 largest counties, while the other half occupy the remaining 2,998 counties. In fact, the two largest counties — California’s Los Angeles County and Illinois’ Cook County — contain roughly the same share of the national population (4.82%) as the 1,437 smallest counties. Especially because they are home to concentrated communities of nonwhite voters, these large jurisdictions helped Obama win reelection even though he carried only 22% of all counties. (Throughout this piece, “county” will also include “county equivalents,” such as independent cities and parishes.)

Indeed, Obama carried 46 of the 50 most populous counties, many of them by overwhelming margins — and 12 more of the 50 most-populous counties than Al Gore (34) carried a dozen years earlier. Longer-term comparisons provide an even starker demonstration of the confluence of county-level population concentration and Democratic presidential performance: In his four-point, 2012 reelection victory Obama carried a mere 690 counties, whereas 24 years earlier 1988 Democratic nominee Michael Dukakis carried 819 counties despite losing to George H.W. Bush nationally by about eight points. Overall, modern Democratic nominees are receiving a greater share of their presidential votes from a shrinking set of rapidly-growing counties. Sure, Obama won twice, becoming the first Democrat since Franklin Roosevelt to win election and then reelection with national popular majorities both times. But in sub-presidential contests, his party is hamstrung by the geographic over-concentration of its base voters.

Consider Table 1, which lists the number of counties in all 50 states, as well as the number and share of counties in each state containing 50,000 or more residents, plus the subset containing 100,000+. The table summarizes the data for all 50 states as reported in Table 2 (at the end of the article), which sorts states top to bottom from least- to most-densely populated, and codes them either blue, red, or purple based on presidential results for the past four cycles: blue if it is one of the 18 states carried by the Democratic nominee each of those elections; red if it’s one of the 22 states carried all four times by the Republican nominee; and purple for the so-called 10 “swing states” that neither party won all four elections.

First, notice that the 18 Democratic-leaning states on average contain half as many counties (716, mean = 37.8) than the 22 Republican-leaning states (1,722, mean = 78.3), with the swing state mean in between but more closely approximating the Republican states (704, mean = 70.4). Overall, the red states contain 1,000-plus more counties than the blue states. In fact, the seven, solidly blue states of Connecticut (8), Delaware (3), Hawaii (5), Maryland (24), Massachusetts (14), New Jersey (21) and Rhode Island (5) combined have only 80 counties — fewer than a third of reliably red Texas’ highest-in-the-nation 254 counties. (Counties are also divided into various political subdivisions, including cities and townships.)

Table 1: Number and size of counties in blue, red, and swing states

Note: *The county totals include “county equivalents” such as independent cities and parishes.

Source: Compiled by author from U.S. Census Bureau.

Relatedly, the blue states on average feature higher shares of high-population counties. The share of counties with populations in excess of 50,000 persons in the 18 blue states (51.2%) is more than twice that of the 22 red states (20.6%), and the share of the subset counties with more than 100,000 persons in blue states is thrice that of red states (34.4% to 10.6%). Again, the respective shares for the 10 swing states fall in between: 35.9% and 21.0%, respectively, for 50K+ and 100K+ counties.

In the post-Civil War period, county configurations have changed some, but not much; typically new counties have been formed by breaking apart older, larger ones into smaller jurisdictions. But since World War II, changes in county structure are extremely rare. The partisan differences in county structure thus result neither from Democratic states consolidating their county systems nor Republicans expanding theirs. Rather, it is a bizarre artifact of modern political-electoral geography that states with fewer counties — some of which were once conservative Republican strongholds — tend to vote more Democratic today. However, that those fewer counties tend to have larger and more Democratic-leaning populations is no artifact: Population densities of major metropolitan areas, mid-sized cities, and densely-packed inner suburbs naturally coincide with liberal and Democratic preferences. Indeed, the 2012 exit polls reveal the unsurprising result that Obama also performed better the more consolidated the “size of place”: He won a bigger share of voters in big cities (69%) than mid-sized cities (58%), small cities (44%), or suburbs (50%).

Variations in county-level political geography would matter little if at all, of course, were it not for the contrasting preferences of voters in densely-populated areas compared to those in more sparsely-populated parts of their states and across the nation. But the significance of fewer, high-population counties in Democratic-leaning states is that concentration almost certainly affects the parties’ pool of potentially viable up-ballot candidates for state legislative and statewide offices.

Many county-level election contests are nonpartisan, and even where they are contested, the debates may not focus on the types of polarizing debates on issues like abortion and gay marriage that dominate statewide, congressional, and presidential elections. But the convergence of states’ partisanship and jurisdictional variance provides far more county offices to seek in Republican-leaning states. That said, if the extensive political science scholarship on progressive ambition and candidate quality[1] matter at all, by dint of sheer numbers contemporary Republicans should benefit from a deeper or at least more plentiful bench. Because the presidential candidate pool draws from governors, and the gubernatorial pool in turn draws from state legislators and other statewide politicians, who in turn draw from the stocks of county and municipal officials, electoral geography should provide the Republicans a more extensive farm system of “quality” electoral candidates, ceteris paribus.

In a thought-provoking essay for The Atlantic entitled “Red State, Blue City: How the Urban-Rural Divide is Splitting America,” Josh Kron examined the ways in which city-based Democratic voters in some states are able to exert sufficient political power, often through ballot measures, to expand gay rights or legalize marijuana — despite sometimes fervent resistance from majorities in many non-urban jurisdictions. Written before the Kim Davis gay marriage license episode in Kentucky, Kron foretold some of the tensions that we now see between the preferences of voters aggregated statewide relative to those aggregated into local jurisdictions. “Federalism’s dance is America’s great helix, and in due course a new national consensus will tend to emerge. But things might get more divided before they get better,” writes Kron.

Perhaps such tensions are precisely what the drafters of our national and state constitutions intended when they designed the geographic and temporal structures of elected offices. But as Democratic-leaning voters become increasingly concentrated in a small number of very large urban cities and inner suburban counties, tensions between city and state will persist — and the Republican electoral bench, so to speak, should only get deeper.

Table 2: State population densities and county size/shares

Note: *The county totals include “county equivalents” such as independent cities and parishes.

Source: Compiled by author from U.S. Census Bureau.

Thomas F. Schaller is a professor of political science at the University of Maryland, Baltimore County. He is the author of The Stronghold: How Republicans Captured Congress but Surrendered the White House and national political columnist for the Baltimore Sun.


[1] A few notable examples from this extensive literature include: Rohde, David W., “Risk-Bearing and Progressive Ambition: The Case of Members of the United States House of Representatives,” American Journal of Political Science, 1979:1-26; Fowler, Linda L. and Robert McClure, Political Ambition: Who Decides to Run for Congress. (1989: Yale); Maisel, L. Sandy and Walter J. Stone. “Determinants of Candidate Emergence in U.S. House Elections: An Explanatory Study,” Legislative Studies Quarterly, 1997:79-96; and Fox, Richard L., and Jennifer L. Lawless, “Entering the Arena? Gender and the Decision to Run for Office” American Journal of Political Science, 2004:264-80.