There is a bumper sticker cynicism that reads “If voting mattered, they’d make it illegal.” We don’t know whether that’s true, partly because, for the vast majority of Americans, there’s no risk that voting matters much — the use of single-member districts and the natural tendency of urban areas to pack in voters in a concentrated area produces a reliable GOP tilt to American elections for state legislative seats and for the US House of Representatives. As FairVote observes, if elections don’t have consequences, democracy is defeated. And in most of America, elections no longer have consequences for the party in power. The way we elect determines which party will win, not the issues or the funding or the individual virtues of the candidates.
Deficient Democracy: District Partisanship, Disappearing Competition, and Skewed Representation in the US House
The decades-long trend of hardening partisanship among individual voters has coincided with a similar rise in the partisanship of congressional districts. Just as voters have become less likely to shift their support between parties from one election to the next, Democratic- and Republican-leaning districts have tended to elect candidates from their preferred parties by wider and wider margins.
The rising partisanship of both voters and districts has led to a dramatic decline in electoral competition and left millions of Americans without hope of winning representation in their district, or even their state. On top of this, Democratic and Republican-leaning districts have become polarized to different degrees, leading to a built-in partisan bias that distorts the relationship between seats and votes in U.S. House elections.
Shifting district partisanship has fueled both the decline in competition and the growing structural advantage of Republican candidates (see Figure 5.1). The number of swing seats has declined since 1992, in concert with a steady increase in the number of seats that are safe for one of the two major parties. Meanwhile, the number of safe Republican seats has increased more rapidly than the number of safe Democratic seats, even as Democrats have been dominant nationally, winning the popular vote in six of the seven presidential elections from 1992-2016.
The Vanishing Competitive District
As was established in chapter three, for recent U.S. House elections, the only districts that can be consistently contested by either party are “swing districts” with partisanship within three percentage points of an even 50%-50% split. In districts outside of this narrow com- petitive range, election outcomes are highly predictable — the party with majority partisanship in these remaining districts won 97.4% of House elections in 2012 and 2014.
The number of districts that fall in the small competitive range has been declining steadily for decades. As Figure 5.2 illustrates, after elections in 1992, 134 districts (30.8%) had majority partisanship within three percentage points of 50%. By the year 2000, the number of districts in this competitive range had fallen to 78 (17.9%). After redistricting in 2010 and the 2012 presidential election, the number of swing districts in the U.S. House fell to just 47, or 10.1% of all districts. Today that number is 23, or just 5.3% of districts.
New court-ordered maps in Florida and Virginia, created to resolve illegal gerrymanders, actually reduced the number of competitive districts even further in the lead up to 2016. There are now just 45 competitive districts in the United States House. In most of the remaining 390 U.S. House districts, the outcome of general elections is effectively predetermined. Supporters of the minority party, whether they make up 5% or 45% of the electorate, have little chance of winning representation.
FairVote’s projections for 2018 U.S. House races, discussed in chapter one, further illustrate the dearth of competition in congressional districts across America. The model uses only partisanship and recent U.S. House election results to make high-confidence projections in 374 districts (86%) for 2018. These high-confidence projections were accurate in 1,061 of 1,062 races projected in Monopoly Politics 2012, 2014 and 2016. There are only 61 districts in which out- comes are not projected for 2018 (highlighted in yellow in Figure 5.3 above).
Inherent partisan bias
In the 2012 election, Democratic candidates for the U.S. House of Representatives received over a million more votes than Republican candidates, but Republicans won a large majority of seats — 234 (54%) to 201. The election’s skewed outcome distorted the will of American voters, who, at the national level, voted for one party and elected the other. While this “wrong-way” outcome was a powerful illustration of what can go wrong under our current system of elections, the more important point is that this was not a freak occurrence, but a natural outcome of a consistent partisan bias that is now built into the structure of our elections itself.
Today, there is one underlying geographical phenomenon that makes partisan skew toward the Republican Party inevitable under our current winner-take-all system: Democratic voters tend to be clustered in cities, while Republicans are more spread out across the suburbs and rural areas. As a result, Democratic-leaning districts are more Democratic than Republican-leaning districts are Republican. In 2012 and 2014, the average partisanship of Republican-leaning districts was 61.5% Republican, while the average partisanship of Democratic-leaning districts was 65.0% Democratic. Because of this discrepancy, an equal number of votes will earn Repub- licans a greater number of seats, because more Democratic votes are “wasted” running up large margins of victory in heavily Democratic districts.
The implications of this inequality are apparent in Figure 5.4 below, which plots the Democratic Party’s share of national two-party support against their expected share of House seats in 2018, using FairVote’s full projection model. Under any level of national support (the x-axis), the share of House seats won by Democrats (represented by the purple line) is projected to fall well below what they would achieve if there were a proportional relationship between a party’s support among voters and their share of seats (represented by the orange line).
Given an even 50%-50% split in national support between the parties in 2018, Democrats would be projected to win just 44% of House seats to Republicans 56%, a margin of 54 seats. For Democrats to win even a one-seat majority, our model suggests that they would need to win at least 55.4% of the national two-party vote share. Meanwhile, we project that in a Republican wave year in which the GOP earns 54% of the national two-party support, they would win 58% of seats, leading to an 73 seat margin over Democrats in the House.
While a portion of the Republican advantage illustrated in these projections comes from the incumbency bump and the fact that more Republican incumbents than Democratic incumbents will be running in 2018, a strong bias would persist even in the absence of any incumbents. As described in chapter 1, in a 50%-50% year in which all seats were open, Republicans would still be favored to win 237 districts, with Democrats favored in the 198 remaining districts.
Monopoly Politics at the State Level: Partisan Minorities Shut Out from Representation
In many states, partisan skew and declining competition combine to distort outcomes and consistently thwart any hope of representation for supporters of one of the two major parties. While the states highlighted below constitute some of the most egregious examples, similar outcomes can be found in many states across the country. In the states highlighted below, and in U.S. House elections more broadly, scarce competition and unrepresentative outcomes create a disconnect between the will of voters and election outcomes that undermines our democracy.
Even as Republicans are consistently overrepresented at the national level, the use of winner- take-all elections means that Republican voters can just as easily be underrepresented within states in which they are in the minority. Massachusetts, for example, is often thought of as a solid bastion of support for the Democratic Party. In reality, however, the state’s partisanship suggests that 37% of its voters lean Republican.
Despite such a substantial proportion of its voters favoring Republicans, all nine of Massachu- setts’ congressional seats are currently held by Democrats (Figure 5.5). Republicans have not won a U.S. House seat in Massachusetts since 1994. The partisan imbalance and the lack of competitive districts is so severe that we project all nine seats as safe Democratic seats. Unrepresentative outcomes are so ingrained in the structure of Massachusetts U.S. House elections that Republicans did not field a candidate in five of the state’s nine House races in 2016.
As in Massachusetts, House elections in Oklahoma consistently exclude a sizable portion of the state’s voters from representation. While the state’s partisanship is nearly one third Democratic, all five of the state’s U.S. House districts lean Republican by wide margins.
Oklahoma’s last remaining “Blue Dog” Democrat left office before the 2012 election. As none of the state’s congressional districts have partisanship anywhere near the narrow competitive range, representation for the Democratic minority will remain unlikely in any of the state’s districts for the foreseeable future.
Massachusetts and Oklahoma are thought of as highly partisan states—the former Democratic, and the latter Republican. In reality, a large share of voters in Massachusetts prefer Republicans, and a significant portion of Oklahomans consistently prefer Democrats. However, of the 14 districts across these two states, not one is held by the state’s minority party. These states are not unique: Arkansas, Connecticut, Hawaii, Idaho, Kansas, Nebraska, New Hampshire, Rhode Island, and Utah are also completely dominated by one party despite having minority party partisanship of at least 30% and multiple congressional districts. The interaction of increasingly partisan voters and districts with our winner-take-all system of elections means that these minority party voters, and millions of others in uncompetitive districts across the United States, are continually precluded from any chance at representation in Congress.
Monopoly Politics at the State Level: Partisan Skew
Across the United States, Republicans’ structural advantage in House elections results in large part from the geographic distribution of voters. Many Democratic votes are wasted in highly Democratic urban districts, while Republicans candidates tend to win by smaller margins, often in suburban and rural districts. No state illustrates this better than Pennsylvania.
Pennsylvania is an evenly divided state in terms of partisan- ship, and has been won by the Democratic candidate in six of the last seven presidential elections. Democrats, however, control only five of the state’s 18 congressional seats and are projected to have just four safe seats in 2018. Predictably, this skew results from a significant difference in the partisan margins of the state’s Democratic and Republican districts.
The average partisanship of Pennsylvania’s Democratic dis- tricts — concentrated in the Philadelphia and Pittsburgh metropolitan areas — is 69.3% Democratic, while the av- erage partisanship of the state’s GOP-leaning districts is just 59.1% Republican. The gap in average margins of victory is even larger. In 2016, the average margin of victory for Republican winners in Pennsylvania’s U.S. House races was 35.2% of the two-party vote, while Democratic winners won with an average margin of victory of 60.2%.
While Pennsylvania presents an extreme example, the same phenomenon occurs in House elections across the United States. Their cumulative effect is an entrenched partisan bias that, in today’s electoral environment, skews results in favor of the GOP. The changing nature of political geography means that these structural biases can benefit different parties at different points in time. However, disproportionate results at both the state and national level will remain an enduring feature of U.S. House elections as long as the system of winner-take-all elections remains.
The Democratic Disconnect
The stagnant and consistently skewed results of U.S. House elections constitute a breakdown in the democratic linkage between citizens and government. The House was envisioned by the framers of the Constitution as the most representative and directly democratic element of the American system. Yet today, in nearly all of the 391 House districts (89.9%) that lean significantly towards either Republicans or Democrats (with majority partisanship of at least 53%), the outcome of elections is effectively predetermined. The tens of millions of voters that support the minority party in these districts have little chance of electing a representative that shares their views, and the power of gerrymandered districts and political geography ensure that one party will retain an arbitrary chamber majority in all but their most lopsided electoral defeats.
Legitimate democracy requires that elections have consequences – winning more votes should mean winning more seats, and representatives that disappoint their voters should risk electoral defeat. When these links are broken, political polarization is compounded, as safe incumbents are left with even fewer incentives to compromise with the other side.
A second casualty is voter engagement. Low voter turnout results, in part, from the belief among some citizens that their vote is unlikely to have an impact. Unfortunately, in the vast majority of U.S. House districts, these citizens are right. Even in competi- tive districts, voters are faced with the reality that only in the strongest Democratic wave years is their local election likely to impact partisan control in the chamber.
The next section of this report will explore the different forces driving the rise of partisanship in congressional districts across the United States, and the partisan skew and scarce com- petition that result. Ultimately it will be shown that uncompetitive districts and consistently skewed results are an inevitable outcome in our system of winner-take-all elections in single-winner districts. Regardless of the presence of gerrymandering or the political geography of the nation, any system that, like ours, awards groups earning 0% and those earning 49.9% of the vote in a district the same level of representation (none) will produce a narrow range of potential competition and a dubious relationship between seats and votes. True restoration of the link between voters and the U.S. House will therefore require bold reforms to the structure of our elections.