Lack of competition

In every election cycle, between 80 and 90 percent of congressional districts are safe “blue” or “red” districts, making the general election a mere formality. Competitive general elections are necessary for keeping representatives accountable to voters.

Redistricting reform alone is not an effective vehicle for increasing competition. Even when more competitive districts are drawn, they can undermine fairness to political parties and racial and ethnic communities.

Opening primary elections to allow more voters to participate in them is no substitute for competitive general elections. Primary elections, however constructed, have low and unrepresentative turnout.

Without competition, an election is not really an election at all. Unfortunately, most Americans live in safe Republican or Democratic districts, where the results are easily known in advance, making the vote itself a mere formality. The many ways to measure and quantify competitiveness all demonstrate that, across our congressional districts, genuine competition is extremely rare. Most incumbents can comfortably win reelection if they want, and most do. Since 2016, politics has become less stable than before, and every election cycle since then surprised forecasters in various ways, yet in each of them more than nine out of ten incumbents were reelected.[1]

The reason is simple: Partisanship. The evidence shows that most voters have a preference for either Republicans or Democrats that outweighs almost anything about the particular candidates. Partisans for the two parties are not evenly distributed – some places are heavily Republican, and some places are heavily Democratic. Under these circumstances, it is inevitable that any reasonably compact district map will have a lot of heavily Republican or heavily Democratic districts.

We might admonish voters to look beyond party labels and assess candidates as individuals, but the reality is that partisanship overrides character, policy preferences, and constituent service. Candidate quality does matter some, but partisanship is the overwhelming factor in how people vote in congressional elections. Voters in the Texas panhandle district, where former President Donald Trump got about 80 percent support in 2020, will not elect a Democrat no matter how qualified the Democratic nominee or unqualified the Republican nominee.[2]

Likewise, voters in southern Houston, where President Joe Biden won about 75 percent of the vote, will not elect a Republican. Partisanship is highly correlated with the presidential vote, so those districts are safe for the incumbent party, as are the vast majority of other House districts. In 2022, the winner represented the same party as the presidential candidate who won the district in 421 out of 435 districts, over 96 percent of them,[3] demonstrating that most voters do not pay attention to congressional campaigns but simply align with a presidential nominee and then vote the same party in down ballot races.

Consider these other stats from the 2022 election:

  • Four out of five Representatives won by more than 12 points, and fewer than one in 10 won by six or fewer points.
  • The median margin of victory was 29 percentage points, meaning the median winner got more than 64 percent of the two-party vote.

In 25 districts, voters had no choice at all: Only one major party candidate appeared on the ballot.[4]

None of this is surprising: Predicting the outcomes of elections is a cottage industry, and all major horse race predictors (including Sabato’s Crystal Ball, The Cook Political Report, Inside Elections, and FiveThirtyEight), rated at least 300 districts safe for the incumbent party heading into Election Day in 2020.

Since 1997, FairVote, a nonprofit organization that promotes reforms to winner-take-all elections, has released its report Monopoly Politics every election cycle analyzing the competitiveness of congressional districts. Every report has demonstrated that more than 80 percent of congressional seats can be comfortably called for their incumbent party using data available two years in advance: specifically, the presidential vote in the district and the historical performance of the incumbent.[5]

FairVote’s report has called more than 300 seats every cycle, beginning with the 1998 edition. Its projections have been more than 99 percent accurate across the last six election cycles, despite not using any polling data or any other inputs other than voting history.

In other words, more than four out of five voters live in a district where their incumbent representative can sail to reelection, cycle after cycle, until the time comes for another member of the same political party to take their place. Redistricting or shifting partisan attitudes may upset the odd representative in a previously safe district, but they do not change this overall picture. For most of us, voting does not do very much, because our representatives were chosen for us with the adoption of a district map, long before Election Day.

This sad situation is the inevitable consequence of our choice to elect the U.S. House in single-winner districts, which prioritize geography over everything else. Geographic communities that are evenly split between people who prefer Democrats and those who prefer Republicans are rare. This is especially true given how few true swing voters exist today.

The most common suggestion to increase competition in single-winner districts is to either redistrict in a way that creates more competitive districts or to change how candidates are nominated to the ballot. However, as explained below, creating more competitive swing districts would undermine the fairness of the districts in other ways, while changing nomination rules simply does not increase competition. If we want elections where voters, not districts, decide representation, then we need to stop using single-winner districts to elect our representatives.

A district map cannot be both competitive and fair

In theory, congressional districts could be drawn in such a way that many more would be competitive. That, however, would not be a good thing for democracy. People tend to live in communities that share political values and interests, so districts that keep communities together rather than dividing them are rarely split evenly between voters who prefer Republicans and those who prefer Democrats.

Even if we could draw a much more competitive map, it would quickly be discarded as unfair. That is because competitive districts do not get distributed proportionally between the two major parties after an election. Rather, they tend to all swing together as a block, giving a disproportionate advantage to the party that wins them all. That is because the national character of congressional elections means that the nation as a whole tends to swing toward one party or the other each cycle. If the political winds favor Democrats, then we often see a nearly uniform swing in favor of Democrats across all districts. Closely divided districts tend to go to the party that is favored that year.

Consider, for example, the 2012 congressional elections in Arizona, then a red state. That year, GOP presidential nominee Mitt Romney won the state with about 54 percent support. About 52 percent of Arizona voters voted for Republicans in the congressional elections across the state as well. 

However, Democrats won five of Arizona’s nine House districts that year, despite losing the statewide vote. This was not due to intentional gerrymandering; Arizona’s districts were drawn by an independent redistricting commission, and the map it adopted actually had more Republican-leaning districts than Democratic-leaning districts. The problem is that Arizona’s Constitution instructs the Arizona Independent Redistricting Commission to maximize the number of competitive districts. The map that the commission adopted in 2011 included one safe Democratic district, four safe Republican districts, and four competitive districts. 2012 was a Democratic year overall, resulting in a swing in favor of Democrats across all districts. As a result, Democrats won all four of the state’s competitive districts, along with the sole safe Democratic district, for five seats total. Based on the statewide vote, a fair outcome would have been five Republican seats and four Democratic seats, but the competitiveness of the districts actually undermined fair representation

This problem takes on a different dimension in the context of racially polarized voting. Traditionally, communities of color earn representation through “majority-minority” districts where the minority group’s eligible voting age population comprises a majority of the district. In states with racially polarized voting, the Voting Rights Act (VRA) requires states to draw districts where communities of color have the numbers to elect their preferred candidates. 

To be effective, VRA districts cannot be competitive. If they were, the relevant community’s candidate of choice would lose about half of the time, diluting the community’s voting power. In practice, districts drawn to comply with the Voting Rights Act are usually very safe; in this context, they cannot be both competitive and provide fair representation.

Modern redistricting reform proposals require district maps to abide by a list of criteria: They must be compact and of equal population, they must not divide communities of interest, and so on. Competitiveness is rarely a criterion. However, maps are often required to be fair to both political parties. When Democrats in Congress sought to require nonpartisan redistricting in 2021 in the Freedom to Vote Act (S. 2747), the bill said nothing about drawing competitive districts, but it did forbid states from creating districts “drawn with the intent or [that have] the effect of materially favoring or disfavoring any political party.”[6]

Given the tension between fairness and competition in winner-take-all districts, modern redistricting reform proposals may, paradoxically, call for fewer competitive districts to promote statewide fairness.

Reforming primaries to increase competition does not help

Primary election competition is no substitute for general election competition. In almost all states, primary elections, where political parties choose their nominees, are technically internal political party matters. State-run primary elections do ensure that the party’s rank-and-file members select the nominee, rather than party leaders alone, but primaries are not fully public elections. Many states hold closed primaries, allowing only voters affiliated with a political party to participate.

However, primaries in districts that are safe for one party functionally serve as the election itself – whoever wins the primary will ultimately represent the district in Congress. With so few competitive districts, voters may resent being excluded from closed primaries. The movement for open primaries aims to remedy this exclusion. It essentially seeks to turn party primaries into public elections by allowing all voters to participate in them.

The apex of the movement for open primaries is the “Top Two” system used in California and Washington. Under this system, there are no partisan primaries at all; political parties cannot nominate candidates to represent them on the ballot. Rather, every candidate appears on the same blanket primary ballot, and voters may vote for any one of them, regardless of party. Then, the first and second vote-getters advance to the general election in November. They are not party nominees; in fact, they could both come from the same party.

But this method has not increased the competitiveness of congressional elections. Any eligible voter can participate in the first round, but few do. It takes place months prior to the general election, and it does not elect anyone, so it naturally draws less public attention and less media coverage than the general election. Ultimately, primary elections in California and Washington suffer the same fundamental problem as primary elections in other states: They are often the decisive election, yet they attract only a small group of more partisan and less representative voters.

In theory, the Top Two system could increase competition in the general election too by changing who appears on the general election ballot. A safe Democratic district, for example, may be more likely to advance two Democrats from the first round. A Democrat would definitely win, but the general election would decide which Democrat it would be.

In practice, this happens only rarely. In 2022, California had 36 strongly Democratic congressional districts and five that were strongly Republican. But only six of these 41 safe districts featured two Democrats in the general election, and none pitted two Republicans against each other.[7] And even when it does happen, it’s unclear how much competition intraparty races create. Five of California’s six Democrat-versus-Democrat contests were won by a greater than 10 point margin, suggesting that these elections were not all that competitive.

Normally, more competitive races draw higher turnout, with voters more motivated to vote in a contest where every vote could really matter. Yet turnout in intraparty races was lower than in otherwise similar districts with both a Democrat and a Republican. The median number of votes cast in the California districts with two Democrats was 20 percent less than the median number of votes cast in California’s other safe Democratic districts.[8] Congressional voting is dominated by national partisanship, so voters may see a choice between two Democrats as no choice at all, especially those who prefer Republicans.

Median votes cast in top two
Infogram

In 2020, Alaska voters adopted a novel way of conducting elections that attempts to improve on the top two system. The primary is the same, but the top four candidates instead of the top two.  These four then compete in a general winner-take-all election conducted by (single-winner) ranked choice voting. This approach doubles down on the primary election reform focus of top two, while attempting to improve voter choice in the general election, in part by promoting intraparty competition in general elections.

Analysis of Alaska’s first use of the system in 2022 suggests that this approach does increase intraparty competition compared to top two. Alaska’s at-large congressional district drew significant media attention as both a special election in August and the general election in November featured the same competitive field with Democrat Mary Peltola competing against Republicans Mark Begich and Sarah Palin. The inclusion of two Republicans certainly gave voters a more dynamic choice, but the competition was actually not between the Republicans, but between Palin and Peltola, with Peltola winning both the special and general elections.

In Alaska’s state legislative races, turnover was characteristically low. Altogether, 40 incumbents sought reelection across the state’s two chambers. Eight of those 40 lost to challengers, and three of those eight were losses to members of the same party. These numbers do not support the notion that Top Four made elections a lot more competitive, particularly given that redistricting had likely weakened many incumbents’ positions, but the system did apparently move a small amount of competition from the primary into the general election. 

Competition matters

Competition gives voters a more powerful vote in the general election, makes representatives more accountable to their voters, and elects representatives who are more likely to work with colleagues for sensible legislation rather than grandstand and obstruct. 

Safe districts allow small, ideologically extreme groups to dominate primary elections and effectively control who represents their districts. Congressional districts comprise about 750,000 residents each. Rep. Marjorie Taylor Greene of Georgia, who lost her committee assignments for promoting white supremacist and antisemitic conspiracy theories, earned her district’s Republican nomination in 2020 with fewer than 44,000 votes in her low-turnout primary and even lower-turnout primary runoff elections. Winning the Republican nomination in her safe Republican district ensured her ultimate victory, which means that about 6 percent of the population of Georgia’s 14th district chose its representative. 

Greene is an extreme example, but the dynamic that elected her plays out in many congressional elections. As a result, representatives in safe districts lack accountability to general election voters. Instead, they remain accountable to primary election voters, a significantly smaller, more partisan, and less representative group. Meaningful primary challenges are rare, but the threat of such challenges shapes decisions legislators make while in office.

Safe districts essentially convert a two-party system into a series of one-party systems. They create bubbles where rigid opposition to partisan opponents is prized above constructive legislative work. This dynamic plays out across different approaches to redistricting and to primary elections. In short, reform within the contours of the single-winner district system has failed to provide voters with competitive democracy.

Fortunately, there is an alternative. Multi-winner districts with proportional ranked choice voting would completely upend this dynamic. With proportional voting, there may be safe incumbents — those who have done the work of earning constituent support — but there is no such thing as a safe district. Every district would have meaningful competition between the major parties, within the major parties, and outside the two-party system.

Notes

[1] This only counts incumbents who actually sought reelection. The majority of congressional turnover comes from representatives not seeking reelection at all.

[2] In 2022, the most heavily partisan district to elect a member of the opposite party was Alaska’s at-large district. The state voted for Trump over Biden by about a 10 point margin, but elected Democrat Mary Peltola in 2022 in both a special election and in the general election.

[3] Nine districts were won by Biden despite electing a Republican, and seven were won by Trump despite electing a Democrat. Almost all were very closely divided. Full data is available on request.

[4] This includes on district (CA-10) in which the general election had a Democrat and a Green Party candidate. In an additional six California districts, the ballot had only two Democrats, and in one Louisiana district (LA-06) the ballot had two Republicans and a Libertarian (LA-06 reelected its Republican incumbent with over 80 percent of the vote).

[5] For full disclosure, FairVote is a funder of Our Shared Republic, and the principal author of this and other chapters worked on the referenced report as a staffer for FairVote over several election cycles.

[6] The Freedom to Vote Act, S. 2747 (117th Congress), Sec. 5003(c) No favoring or disfavoring of political parties.

[7] Here, “strongly Democratic” means that Biden outperformed his national performance by at least 6 percentage points. Washington did not have any intraparty contests, and Louisiana had one, though it was not meaningfully competitive, as the Republican incumbent earned more than 80 percent of the vote and his Republican challenger earned only 7 percent.

[8] The median number of votes cast in a strongly Democratic district with both a Democrat and a Republican was 198,464; the median number of votes cast in a strongly Democratic district with two Democrats was 163,377. Full data is available on request.