Disentangling gerrymandering and geographic sorting
As voters across the United States prepare to go to the polls in November, a sobering fact about the competitiveness of our elections is hard to ignore. According to the Cook Political Report’s 2024 House Race Ratings, only 69 out of 435 seats in the U.S. House of Representatives are competitive in the general election, with the remaining 85% of seats solidly in the Republican or Democratic column (see Figure 1). In these uncompetitive general elections, the winner was effectively chosen in the dominant party’s primary—meaning that the vast majority of U.S. House elections will have been all but decided before a single vote is cast in November. This lack of competitiveness is an important component of the “Primary Problem”—when general elections are uncompetitive the locus of competition shifts to the primaries, which tend to have lower turnout and be less representative of the population at large (especially in states with closed primaries). This reduces elected officials’ overall accountability as they are incentivized to focus only on maintaining the support of their “base,” which often involves moving away from the center to avoid being “primaried” by a more extreme member of their party.
Figure 1: Almost 85% of US House Seats Are Uncompetitive in the 2024 Election
Unfortunately, this lack of robust competition is not new. Historical data from the Cook Political Report indicates that at least since the mid-1980s, more than half of all U.S. House seats have been regularly uncompetitive in the general election (see Figure 2). Yet, this problem has become more acute over the past two decades. In fact, of the 13 cycles of U.S. House elections between 2000 and 2024, only two cycles have seen more than 100 competitive general election races (2010 and 2018). Even these elections were no boon for competition—only 31% of races were competitive in 2010, with an even more disappointing 26% being competitive in 2018.
Figure 2: Number of Competitive House Races Has Hovered Below 100 for Quarter Century
The dearth of real general election competition is not limited to a specific set of states or one region of the country (see Figure 3). In fact, 20 states —ranging from New England to the Deep South to the Mountain West—have precisely zero competitive districts in 2024. In an additional 18 states fewer than 25% of districts are competitive, while in another 9 states between one-quarter to one-half of districts are competitive. In only three states are a majority of districts competitive—Alaska’s single at-large seat, both of New Hampshire’s two seats, and three of four seats in Nevada.
Figure 3: Most States Have Few Competitive Congressional Races
The lack of competitiveness in most U.S. House elections is rooted in two features of our political system: gerrymandering (the drawing of district boundaries for partisan benefit) and geographic sorting (the natural geographic concentration of politically similar populations, whereby Republicans tend to live in rural areas, and Democrats tend to live in urban areas). Fortunately, two widely-considered election reforms stand to address these dynamics and dramatically improve the competitiveness of our elections—independent redistricting could reduce partisan gerrymandering by limiting politicians’ ability to draw their own districts, while all-candidate primaries would increase competition and ensure all voters have a say over who represents them. Given the unbelievably low amount of general election competition in U.S. House elections this year, considering these reforms appears more urgent than ever.
While uncompetitive districts are a significant challenge across the country, the problem emerges from two distinct dynamics—gerrymandering and geographic sorting.
Gerrymandering is the practice of drawing district boundaries with the intent to advantage one group over another. Gerrymandering has a long history in the United States, dating back to the early 19th century when the misshapen districts approved by Massachusetts Governor Elbridge Gerry gave the practice its name. While gerrymandering with the intent to dilute the power of particular racial groups is barred by the Voting Rights Act of 1965, doing so for partisan advantage is not. In fact, in 2019 the Supreme Court ruled in Rucho v. Common Cause that partisan gerrymandering is “nonjusticiable,” essentially barring courts from litigating cases where district lines are drawn to influence the partisan outcome of elections.
Partisan gerrymandering is carried out using two principal strategies—cracking and packing (see Figure 4). Cracking involves diluting the power of a group of voters by dividing them across many districts, while packing consists in cramming a group of voters into a small number of districts. For example, consider a hypothetical state with 60 “blue” and 40 “red” voters divided across five districts. A fair outcome in this context would likely result in three blue and two red districts. However, blue could sweep the board by “cracking” the red population across the map and giving itself the majority in all five districts (see left panel of Figure 4). By contrast, red could gain the upper hand by “packing” blue voters into two districts, leaving the remaining three with a red majority (see right panel of Figure 4).
Figure 4: Gerrymandering Strategies—Cracking and Packing
Cracking and packing are commonly used tactics, including during the 2020 redistricting cycle. For example, Illinois’ Sixteenth Congressional District snakes its way from the Wisconsin border in the north to the central part of the state, packing Republican-leaning populations into one district (see left panel of Figure 5). While this district is overwhelmingly Republican—its Cook Partisan Voting Index favors the GOP by 13 points—the majority of the state’s 17 districts are solidly Democratic. Likewise, the Democratic majority in Houston and the swing voters in its immediate suburbs have been cracked into more than half a dozen districts, most of which lean Republican by a 60–40 margin (see right panel of Figure 5).
Figure 5: Examples of Cracking and Packing
While gerrymandering is the most well-known and infamous cause of uncompetitive elections, solidly red or blue districts are also the product of geographic sorting. Through a wide range of political, economic, and social processes, left- and right-leaning populations often self-sort into geographically separate communities. For example, the Democratic Party tends to perform best in urban and suburban areas, while the Republican Party is often strongest in rural areas. Likewise, New England stands out for its large population of Democrats, while the Great Plain states are home to a high concentration of Republicans.
Although not the product of intentional manipulation, geographic sorting is also partially responsible for the lack of competition in U.S. elections. Perhaps the clearest example of this fact comes from U.S. Senate elections—since Senators are elected to represent an entire state and state lines are not redrawn for electoral purposes, any partisan advantage in these elections are better understood as a result of geographic sorting rather than gerrymandering. Of the 100 sitting Senators elected between 2020 and 2024, 36 were elected in competitive races and 64 were elected in uncompetitive races (see Figure 6). While the fact that only one-third of Senate elections in recent years have been competitive is no cause for celebration, this figure is remarkably higher than the 15–20% of House races during this same period that have been competitive.
Figure 6: Competitiveness of U.S. Senate Elections, 2020–2024
Data from the Gerrymandering Project at Princeton University provides an even more granular assessment of how our current congressional maps unfairly reduce competition and advantage one party or the other. To assess the extent of gerrymandering across individual states, researchers compiled various metrics to measure the geographic shape and partisan fairness of districts. For example, compact districts were considered less gerrymandered than those that snake their way across large areas, while districts that cut across county lines were viewed as more gerrymandered than those that respect pre-existing political boundaries. Likewise, states in which parties’ seat share is proportional with their statewide vote share were marked as less gerrymandered than those in which there was a large discrepancy between vote share and seat share. 1
These metrics were then used as the basis for a report card that assigns a letter grade to each state based on its relative level of gerrymandering (see Figure 7). Across the country, 16 states with 140 total seats were graded “A,” six states with 85 seats received a “B,” four states with 16 seats were given a “C,” a single state (Ohio) with 15 seats was assigned a “D,” and 12 states with 155 seats received an “F.” An additional 11 states with 24 seats were not graded—these states have only one or two districts, precluding the calculation of the aforementioned gerrymandering metrics. 2
Figure 7: Gerrymandering Across the United States
Diving deeper into the proportion of competitive races with each grade reveals that gerrymandering only accounts for part of the reason districts are uncompetitive (see Figure 8). For example, while one would expect to see high levels of competitive races in states given a grade of “A” or “B,” only 20% of races in these top-ranked states were competitive. Furthermore, only 13% of contests were competitive in the 31 districts from states graded “C” or D.” Not surprisingly, competitiveness was lowest in states graded “F”—of the 155 districts in this group, only 11% were competitive. Of the 24 seats in states without a grade, 12.5% were competitive. While competitiveness was certainly higher in states with better gerrymandering grades, it was still low overall.
Figure 8: Gerrymandering Captures Only Part of Uncompetitiveness
So, outside of gerrymandering, what else causes these pervasively low levels of competition? The phenomenon is best explained by geographic sorting—which makes it difficult to construct competitive districts in states that heavily lean toward one party or the other (see Figure 9). For example, Indiana’s nine districts were given an overall grade of “A” and also received an “A” for partisan fairness and geographic features—but the state received a “C” for competitiveness, given that only one district is competitive (IN-01). Likewise, Massachusetts received an overall gerrymandering grade of “A” with an “A” for partisan fairness and competitiveness, but it received a “C” for geographic features due to the large number of county splits. Despite receiving an “A” in most categories, all nine districts in Massachusetts are uncompetitive due to the high concentration of Democratic voters in the state.
Figure 9: Geographic Sorting in Indiana and Massachusetts
Gerrymandering and geographic sorting work in tandem to reduce the competitiveness of our elections. Therefore, policy solutions that seek to reduce uncompetitiveness should address both root causes.Independent redistricting is the clearest and most feasible way to reduce gerrymandering in the short term. This reform takes the power to draw legislative districts out of politicians’ hands and entrusts it to a commission with a mandate to draw competitive lines that do not unfairly advantage either party (see Figure 10). Already, six states with a total of 86 seats have empowered independent redistricting commissions to draw congressional maps without requiring approval from the state legislature. An additional six states with 49 seats have adopted advisory commissions with the power to recommend redistricting maps to the state legislature (although the legislature is free to ignore them and draw its own map), while three more states have adopted other types of commissions that reduce the role of the state legislature in drawing district lines. 3
Nevertheless, 28 states with 262 seats—almost two-thirds of the entire U.S. House of Representatives—entrust complete authority over drawing congressional districts to the state legislature. Alabama currently uses a court-imposed map, and the remaining six states have only one at-large district and do not have Congressional redistricting.
Figure 10: Independent Redistricting Across the United States
Overall, the evidence suggests that independent redistricting commissions do help mitigate the problem of uncompetitive maps (see Figure 11). States with fully independent and advisory committees have the highest rates of district competitiveness—22% and 27%, respectively. By contrast, competitiveness is lowest in states where the legislature (12%) or other types of politician-led commissions (12%) draw maps.
Figure 11: Competitiveness Highest in States with Independent Redistricting
Nevertheless, independent redistricting alone is likely to prove insufficient to ensure that most U.S. House districts are competitive—since geographic sorting produces some states and regions that are naturally less competitive due to the homogenous political orientation of their inhabitants. Furthermore, this reform solution would also be unable to improve the competitiveness of statewide elections, such as those for U.S. Senator, Governor, Attorney General, and State Treasurer. Thus, redistricting reform alone is unlikely to completely solve the problem.
However, pairing redistricting reform with open, all-candidate primaries offers a potential path to encouraging robust competition across all U.S. elections.
Here’s how it works: all candidates, regardless of party affiliation, run in a single primary election in which all voters may participate. Following the primary, only the top vote-getters proceed to the general election—in California, Louisiana, Nebraska, and Washington, the top two candidates move forward, while Alaska’s advances the top four (and the reform currently on the ballot in Nevada would send five candidates to the general election). 4
How does this help address the lack of competitiveness in elections?
First, by allowing all voters to participate in the primary process, all-candidate primaries would require candidates to compete robustly for support from a wide coalition of voters, including independents. Since multiple candidates from the same party could advance to the general election, candidates would have an incentive to develop broad-based support and not simply target the base of their own party.
Second, by allowing multiple candidates from the same party to compete in the general election, all-candidate primaries shift the locus of competition from the primary to the general. This would be particularly valuable in districts that are uncompetitive due to geographic sorting. Currently, elections in these districts are effectively decided in the dominant party’s primary election. Under all-candidate primaries, the general election would feature a realistic choice for voters—especially independents and voters from the less dominant party, who could still express their preference among the several competitive options.
Third, recent proposals promise to further enhance competition by linking all-candidate primaries with instant runoff voting (also called ranked-choice voting) in the general election. Providing voters the opportunity to rank the candidates in order of preference should increase competitiveness by incentivizing candidates to compete for voters’ second and third choices.
All Americans should have the right to vote in competitive elections—but across the country, this is currently not the case. The vast majority of voters live in the 85% of U.S. House districts that are uncompetitive in November, meaning that the result was effectively determined months prior, during the primary election.
Uncompetitiveness is rooted in both gerrymandering and geographic sorting. While gerrymandering represents an intentional attempt to reduce political competition, geographic sorting occurs more naturally and cannot be controlled by policy changes. Nevertheless, independent redistricting and all-candidate primaries stand out as election reforms that could realistically address both factors and enhance the competitiveness of our elections. By taking the power to draw districts out of politicians’ hands, independent redistricting can ensure fairer maps that do not advantage one party or the other. Likewise, by encouraging greater participation from candidates and voters across the ideological spectrum, all-candidate primaries should provide voters with a real choice during both the primary and general election.
Although no single reform proposal is a panacea, addressing the sources of uncompetitiveness is one of the most important tasks of election reform today. Succeeding in this task could initiate a virtuous cycle with positive downstream effects for many additional aspects of our politics.
1 A complete discussion of the methodology used is available on the Gerrymandering Project website.
2 These states are: Alaska, Delaware, Hawaii, Maine, North Dakota, Rhode Island, South Dakota, Utah, Vermont, and West Virginia. Additionally, Alabama’s court-mandated map was labeled “Pass” but not assigned a letter grade. These cases were grouped under “NA” for this analysis.
3 In Hawaii and New Jersey, redistricting commissions composed of politicians have the power to draw districts separately from the state legislature. In Virginia, a hybrid commission including politicians and citizens has power over redistricting.
4 In Nebraska, nonpartisan primaries are used for elections to the state legislature. In Louisiana, a second round among the top two candidates is only held if no candidate obtains a majority of the vote in the first round. In Alaska, the top four nonpartisan primary is followed by a general election using instant runoff voting (also known as ranked choice voting).
