Last updated: · 17 min read
In brief
Democratic House candidates won 1.4 million more votes than Republicans in 2012, yet Republicans won 33 more seats. Much of the trick was in the lines: pack opponents into a few landslides, scatter the rest into narrow defeats. Mathematicians built tools to catch this, from the efficiency gap to millions of computer-drawn rival maps. They can flag an outlier. But they cannot say what a fair map looks like, because every test assumes a definition of fairness first. That choice is political, and it can be written down openly or left unstated.
Contents
The 2012 House election produced a result that looks backwards. Democratic candidates received 1.4 million more votes than Republicans nationwide, and Republicans won 33 more seats. No ballots were destroyed and no precincts closed early; the votes were counted accurately.
The obvious explanation is gerrymandering. That explanation is a big part of the truth; the rest, as we will see, is geography. But “gerrymandering” has become a word that substitutes for understanding: the name is familiar, the mechanism mostly opaque. Once you see the machinery as geometry applied to vote tallies, you can see why the sophisticated tools designed to stop it keep hitting the same limit: they can show that a map is unusual, but they cannot say what a fair map looks like.
What a wasted vote is
Start with five hundred thousand voters, split exactly in half: 250,000 for Party A, 250,000 for Party B. The state gets five congressional seats, so five districts of 100,000 voters each. In a system that translates votes into seats proportionally, the natural outcome is somewhere around 3-2.
Draw Map 1. Three districts go 60,000-40,000 in Party A’s favor; two go 65,000-35,000 against them. Three seats for A, two for B. The map reflects the underlying split.
Now draw Map 2. Four districts go 55,000-45,000 for Party A. One district goes 70,000-30,000 for Party B. Party A wins four of five seats, or 80%, with exactly half the votes.
The voters and their totals are identical in both maps; only the lines have moved. The mechanism is visible once you look at where the losing votes went.
In Map 2, Party B’s 45,000 supporters in each of the four losing districts contributed nothing to the outcome. Those votes don’t accumulate or transfer to a seat elsewhere; they expire at the district boundary. And in the one district Party B wins overwhelmingly, 19,999 of its 70,000 votes were surplus. You win a seat with 50,001 votes; every vote beyond that buys nothing.
Those are wasted votes, and they come in two kinds: the losing votes cast in a district you never had a chance to win, and the surplus votes piled uselessly onto a win you already secured. A deliberate gerrymander engineers an asymmetry in wasting. Pack your opponents into a few blowout districts, and their surplus votes are gone. Distribute their remaining voters thin across many hostile districts where they lose narrowly, and those losing votes are gone too. Meanwhile, spread your own supporters across as many districts as possible, winning each by a comfortable but not extravagant margin. Your side wastes almost nothing while theirs wastes a great deal. A 50-50 electorate becomes an 80-20 seat advantage, and since 2019 no federal court will strike it down.
The arithmetic can also be optimized. Redistricting software can take precinct-level demographic and voting history data and model, before a single boundary is drawn, which configuration will produce the most asymmetric wasting. The map-drawer can then iterate until the map is essentially optimal.
The arithmetic, worked
Map 2 in the example above produces these totals. Districts 1 through 4: Party A 55,000, Party B 45,000 each. District 5: Party B 70,000, Party A 30,000. Each party ends up with exactly 250,000 votes. Party B loses four districts and wins one lopsided landslide.
Wasted votes, Party B: 45,000 per losing district × 4 = 180,000 votes spent on losses. In the landslide win, 70,000 − 50,001 = 19,999 surplus votes. Total Party B wasted: 199,999, call it 200,000.
Wasted votes, Party A: 30,000 in the one loss, plus 4,999 surplus per winning district × 4 = 19,996. Total Party A wasted: 49,996, call it 50,000.
Party B wastes four times as many votes despite identical total vote counts. The efficiency gap, the difference in wasted votes divided by total votes, registers at 30% in this fictional state (150,003 ÷ 500,000). Real gerrymanders are subtler but work the same way: the Wisconsin Assembly map at the heart of Gill v. Whitford produced efficiency gaps of roughly 13% in 2012 and 10% in 2014.
Around the 2010 census, Republicans set out to apply that arithmetic on a national scale.
How REDMAP turned the machine on
For most of American political history, partisan redistricting was a matter of opportunity. Control the state legislature and governorship in a census year and you drew favorable maps; lose that control and the other party drew them instead. It was cynical and widely understood, bounded by what a roomful of politicians with paper maps and staff could accomplish before the deadline.
REDMAP, the Redistricting Majority Project, was the Republican State Leadership Committee’s strategy for the 2010 cycle, built around a simple structural insight: the 2010 midterm elections would determine which party controlled state legislatures when those legislatures drew congressional maps following the census.
Win the right state chambers in the right swing states, and you control redistricting in those states for an entire decade. Karl Rove said as much in public, in a Wall Street Journal column in March 2010, under the subheading: “He who controls redistricting can control Congress.”
REDMAP targeted 107 state legislative seats in 16 states (Wisconsin, Michigan, Ohio and Pennsylvania among them). The Republican State Leadership Committee raised about $30 million for the 2010 cycle, by far its best year, and poured it into down-ballot races most voters never noticed.
Republicans flipped enough state chambers in 2010 to control redistricting where it mattered. New maps were then produced using statistical and mapping software capable of modeling partisan outcomes at the precinct level — calculating expected vote-share under a proposed district configuration before any line was finalized. In Wisconsin, the maps that became Act 43 of 2011 were drafted in a private “map room” at the Michael Best & Friedrich law firm rather than in a public committee room. The drafters worked on state-supplied computers running the redistricting program autoBound, and scored each draft’s partisan performance with a regression model built with political scientist Keith Gaddie.
In 2012, Republican House candidates nationally received just under 48% of the popular vote and won 54% of seats. In states where they controlled redistricting, the conversion was 53% of votes to roughly 72% of seats. The gap between those two ratios is the wasted-vote asymmetry operating at national scale, and the new lines widened it: under the old districts, the Brennan Center estimated, Republicans would have won about 65% of those seats.
What Democrats did
Democrats redrew Maryland's 6th Congressional District in 2011 to turn a reliably Republican seat into a safe Democratic one, by attaching suburban Washington D.C. precincts to rural western Maryland and carving away the Republican-leaning communities that had defined the district for decades. Illinois Democrats did the same at statewide scale in 2021: their new congressional map merged two Republican-held downstate districts and was drawn to yield 14 Democratic seats out of 17, the exact split Illinois sends to Washington today. The Princeton Gerrymandering Project graded it F for partisan fairness.
Both are partisan manipulations. Neither was a nationally coordinated, multi-state operation with a documented strategic brief and roughly $30 million in reported funding.
The court’s long evasion
The Supreme Court has been looking at this problem for half a century. The cases follow a pattern: the problem is real and harmful, and the Court will not provide a remedy.
The modern line of cases begins with Davis v. Bandemer, in 1986. Indiana Democrats challenged state legislative maps that gave Republicans a substantial seat advantage despite Democrats winning a statewide vote majority. The Court ruled 6-3 that partisan gerrymandering claims were justiciable under the Equal Protection Clause — federal courts could hear them. But the majority immediately fractured on what a winning claim would require. Justice White gathered four signatures for an opinion on the merits, not five, and no majority agreed on a standard. The case found Indiana’s maps constitutional and deferred the harder question, leaving the threshold for a successful challenge undefined.
Eighteen years later, Vieth v. Jubelirer narrowed the options to two. Pennsylvania Democrats challenged post-2000 congressional maps. Justice Scalia’s plurality of four argued the question was nonjusticiable, a political question courts had no business touching. Four other justices disagreed but couldn’t agree on a standard among themselves. Justice Kennedy, the deciding vote, occupied a position of suspended judgment: the question was justiciable, he believed, but no workable standard had yet been proposed. His concurrence held out hope that a “limited and precise rationale” might yet be found, and observed that the same computer technology driving gerrymanders “may produce new methods of analysis” capable of exposing them. The efficiency gap was built to answer that concurrence.
In Gill v. Whitford in 2018, Wisconsin plaintiffs challenged Act 43 using the efficiency gap as the constitutional measuring stick — exactly the kind of standard Kennedy had indicated would matter if one could be found. The Court was unanimous that the plaintiffs had not yet proved standing, and it ruled neither for them nor against them on the merits. Seven justices sent the case back for another try; Justices Thomas and Gorsuch would have ended it outright. The plaintiffs had argued about statewide partisan injury; the Court held they needed to show district-specific harm to each named plaintiff personally.
Kennedy announced his retirement nine days later, and in 2019 Rucho v. Common Cause shut federal courts out of partisan gerrymandering claims altogether.
Chief Justice Roberts, writing for a 5-4 majority, held that partisan gerrymandering claims present “political questions beyond the reach of the federal courts.” Federal courts lacked a “clear, manageable, and politically neutral” standard and therefore could not adjudicate how much partisan advantage crosses a constitutional line. Roberts acknowledged that excessive partisanship in districting “leads to results that reasonably seem unjust” and is “incompatible with democratic principles.” Then he handed the problem to Congress and the states.
The efficiency gap had been constructed to answer a specific constitutional challenge that a swing-vote justice had placed on the record, and that justice left the bench before the test was ever judged on its merits.
Where racial and partisan gerrymandering intersect
Racial gerrymandering operates under different law. For four decades after Thornburg v. Gingles in 1986, federal courts had a working framework for judging whether a map diluted minority voting power under the Voting Rights Act: whether a state had packed or cracked minority voters so they could not elect their preferred representatives. In April 2026, Louisiana v. Callais sharply narrowed that framework, imposing new and tougher proof requirements on vote-dilution claims. Partisan gerrymandering, since Rucho, has no federal standard at all.
The two categories do not separate cleanly in practice. In states where race and party affiliation are highly correlated (as across much of the South; nationally, Black voters back Democrats by margins of five to one or more), drawing "Republican" districts and drawing "white" districts can be the same geographic act accomplished with the same line. The legal strategy in response is now routine: assert partisan motivation when racial motivation is the challenge, or the reverse, depending on which claim is more legally precarious. Courts have left the line between permissible partisan manipulation and prohibited racial manipulation blurred.
The efficiency gap and its rivals
Kennedy had asked for a standard. Nicholas Stephanopoulos, then a law professor at the University of Chicago (now at Harvard), and political scientist Eric McGhee tried to give him one in 2015, formalizing the wasted-vote asymmetry into a single formula.
The efficiency gap is compact: subtract one party’s total wasted votes from the other’s, divide by total votes cast. Wasted votes are all losing votes in a district, plus all votes beyond the bare majority needed to win for the winning candidate. The result expresses the asymmetry in how efficiently the two parties convert votes into seats.
Stephanopoulos and McGhee proposed that a gap of 8% or more in a state-house plan, or two or more seats in a congressional plan, should make a map presumptively unconstitutional; in the Wisconsin case, the plaintiffs’ expert Simon Jackman argued for a 7% trigger. Act 43’s gaps of 13% and 10% cleared both bars.
The formula has real virtues. It captures the mechanism and can be calculated from public election results without access to the map-drawers’ intent. Because the result is a single number, a court can evaluate it, which is what Kennedy had asked for.
The rivals followed. The mean-median difference measures the gap between a party’s median district vote share and its mean district vote share: when votes are bunched into a few blowout wins, the median drops below the mean. Of the common metrics, it is among the least accurate options: broadly applicable, but worse at flagging gerrymandered elections than efficiency-gap variants or the declination. The reason is that extreme mean-median values do not reliably line up with extreme seat outcomes. Gregory Warrington, a mathematician at the University of Vermont, developed the “declination” metric: the angle between two lines, one for each party, that encode how many districts the party wins and its average vote share in the districts it wins. It is geometrically elegant and has the same underlying problem.
Every single-number metric is sensitive to where voters happen to live, independently of how lines are drawn, and when one party’s voters cluster in cities, that sensitivity is large.
Massachusetts provides the clearest demonstration. Republicans there routinely win 30 to 40 percent of the statewide vote, yet their voters are spread so evenly across the state that it is all but impossible to draw them a district: for several statewide elections, every one of the astronomically many valid plans built from whole towns or precincts produces a 9-0 Democratic delegation. A neutral map drawn with no reference to partisan data, following county lines and optimizing for equal population, might register an efficiency gap that would trigger constitutional scrutiny under the Stephanopoulos-McGhee threshold. Nobody would have engineered that result; it would follow from how evenly Massachusetts Republicans are spread. A 10% efficiency gap produced by deliberate packing-and-cracking is arithmetically indistinguishable from a 10% gap produced by voter geography alone.
This gives map-drawers cover. Map-drawers in heavily urbanized states know that natural voter concentration generates a baseline efficiency gap before they’ve moved a single line. Draw a map that scores near that baseline, then argue the gap reflects geography rather than manipulation. The argument is sometimes true, and the metric cannot tell when.
Beneath the geography lies a deeper problem. Each metric measures deviation from an implicit baseline: what a “fair” seat distribution is supposed to look like for the given voter geography. But the baseline isn’t in the data. It is chosen before the measurement begins.
Wisconsin's numbers
In Gill v. Whitford the plaintiffs built their case on two election cycles under Act 43. In 2012, Republican candidates won 60 of 99 Wisconsin State Assembly seats while Democrats' average two-party share of the district vote, with uncontested seats estimated, came to about 51.4%. The plaintiffs argued that the efficiency gap in both 2012 and 2014 exceeded the 7% threshold proposed by their expert, Stanford political scientist Simon Jackman, and that the gap's persistence across two election cycles with different turnout patterns showed it was built into the map rather than a single-election artifact. Because the Supreme Court decided the case on standing, that argument was never evaluated on its substance.
The ensemble and the arms race
Single-number metrics also have a practical weakness: one number wears its assumptions so openly that an opposing expert can attack the baseline directly in court. The next generation of tools replaced the single number with a comparison against alternative maps.
Moon Duchin, a mathematician then at Tufts University, founded the Metric Geometry and Gerrymandering Group in 2016. The ensemble approach her team developed produces a distribution rather than a single score. Draw millions or billions of alternative maps that satisfy the same legal constraints as the enacted map (contiguous districts, equal population, Voting Rights Act compliance, reasonable compactness), then ask where the actual map sits in that distribution.
If 99.9% of randomly generated maps satisfying the same legal requirements give the disadvantaged party more seats than the enacted map does, the enacted map is a statistical outlier among the alternatives the ensemble sampled. The finding says the map is anomalous; it passes no judgment of fairness under any contested metric. That result is much harder to blame on natural geography, because the randomly generated maps are drawn from the same state, constrained by the same rules, and they still produce different results from the enacted plan.
GerryChain, the open-source Python library maintained by MGGG, implements this with Markov chain Monte Carlo methods, specifically the ReCom algorithm developed around 2018. Ensemble evidence carried two major rulings in January 2018. A federal panel in North Carolina leaned on simulated-map evidence when it struck down that state’s congressional map, a ruling the Supreme Court later undid in Rucho, and in Pennsylvania Carnegie Mellon mathematician Wesley Pegden testified that the state’s congressional map showed more carefully crafted partisan bias than more than 99.99% of the alternatives his Markov chain explored. The state supreme court struck the map down. GerryChain itself grew out of MGGG’s 2018 Voting Rights Data Institute; an early version was used that same year to analyze remedial maps for Virginia’s House of Delegates, and ensemble methods have been a fixture of redistricting litigation ever since.
What ReCom does
Earlier ensemble approaches generated new maps by flipping individual geographic precincts one at a time between adjacent districts. The problem: the chain moves slowly, staying near the starting configuration rather than exploring the full space of valid maps. For large states, getting a representative sample was computationally prohibitive. ReCom works differently: it selects two adjacent districts, merges them into a single region, builds a random spanning tree of that combined region, then cuts the tree at one edge to produce two new valid districts. Each step makes a large structural change instead of an incremental adjustment. The chain mixes faster and samples more broadly. This matters because ensemble analysis is only meaningful if the randomly generated maps represent what was available to the map-drawer, not just maps that happen to resemble the one already drawn.
The method could also be run in reverse. A map-drawer could generate an ensemble before submitting a plan and, from among the maps that sit comfortably inside the distribution, pick the one most favorable to their party, so the detector can double as a design tool. The software is open source, so that option is open to anyone with a laptop and the precinct data.
Then there is the constraint problem. Generating a random distribution of maps requires specifying which legal constraints to randomize within. Prioritize compactness and you get one distribution; require that county lines be preserved where possible and you get another; interpret VRA compliance strictly and you get a third.
Different constraint specifications produce different distributions of random maps, different baselines, and potentially different verdicts on whether the enacted map is an outlier. Setting up the test requires deciding which values the map should serve, and that is exactly the political question the test was designed to sidestep.
The baseline problem
What does a fair redistricting map look like? The question has at least three distinct answers in serious current use, and they can pull against each other.
Proportionality: a party that wins X% of the votes should win approximately X% of the seats. Competitiveness: as many districts as possible should be contestable, neither outcome determined before anyone votes. Partisan symmetry: if the parties’ vote shares were exchanged, the seat outcomes would also exchange, so neither side holds a structural advantage that persists regardless of its actual support.
A map optimized for proportionality often requires creating safe seats for each party, which eliminates competitive races. A map optimized for competitiveness can produce non-proportional outcomes when one party’s voters are geographically concentrated: you get close races, but the seat split diverges from the vote split. Partisan symmetry can be achieved by making every district a safe seat symmetrically, which scores perfectly on the symmetry metric while producing an election in which nothing is contested.
Every metric embeds some definition of fairness as its assumed baseline, and not always one of these three. The efficiency gap’s baseline is a fixed seat bonus: with equal turnout, a gap of zero requires a party to gain about two points of seat share for every point of vote share above 50%, so 55% of the votes should bring about 60% of the seats. Symmetry metrics, the mean-median difference among them, assume symmetry directly.
The ensemble method encodes the same judgment in the choice of constraints: which definition of compactness, which VRA interpretation, whether county lines bind. In every case the baseline is chosen before measurement begins.
Chief Justice Roberts wrote in Rucho that federal courts lacked a “clear, manageable, and politically neutral” standard because adjudicating partisan gerrymandering would require courts to “make their own political judgment about how much representation particular political parties deserve.” The criticism of that holding as judicial evasion is fair: there was a judgment to make and the Court refused.
But Roberts was also pointing at something structurally real. Each metric stands in for a contested political value, written into its baseline as arithmetic, and a court that adopted one would be making the political judgment he described.
The metrics also fail on their own terms. Two popular symmetry metrics, the mean-median difference and partisan bias, need not grow more extreme as a map’s seat outcome grows more extreme; tested on 18 real U.S. maps, they can miss extreme seat outcomes entirely. And the mean-median difference, efficiency gap, declination and GEO metric can each be gamed: a computer search can find maps that hand one party an extreme number of seats while the metric stays inside the “fair” range the researchers set in advance.
Mathematics can expose a gerrymander, but it cannot define fairness. California’s Citizens Redistricting Commission ranks its criteria in explicit priority order, which puts the value judgment on the public record instead of leaving it to a formula. Its congressional map is on hold for now: in November 2025 voters approved Proposition 50, swapping in a legislature-drawn map until the commission redraws the lines after the 2030 census. The measure answered Texas’s mid-decade redraw, another round of the arms race.
Proportional representation shrinks the leverage by allocating seats by vote share, and removes it when seats are allocated from a single statewide or national tally. Both approaches make the political choice in the open.
None of this changes the 2012 result: 1.4 million more votes for Democratic candidates, 33 more seats for Republicans. The mechanism behind much of that outcome, the wasted-vote asymmetry modeled in software before a single vote was cast and optimized across targeted swing states, is what the efficiency gap and the ensemble methods were designed to expose, and they exposed it well.
Whether that outcome was unjust is a judgment that rests on a prior commitment to what redistricting is supposed to accomplish: the commitment a commission can write down as a ranked list of criteria, and a legislature drawing its own map can leave unstated.
Related articles
- The countries that chose autocracy democratically
- The paradox of tolerance: what Karl Popper actually said and why everyone is using it wrong
- The countries that never needed a revolution — and what protects them
Further reading
As an Amazon Associate, Eikleaf earns from qualifying purchases. The book links below are affiliate links: if you buy through them, we may earn a small commission at no extra cost to you.
Shape: The Hidden Geometry of Information, Biology, Strategy, Democracy, and Everything ElseIts substantial treatment of gerrymandering explains how geometry, randomness and competing fairness standards shape electoral maps.
One Person, One Vote: A Surprising History of Gerrymandering in AmericaTraces district manipulation from colonial politics through REDMAP, court challenges and independent redistricting commissions.
Ratf**ked: Why Your Vote Doesn’t CountReconstructs the strategists, money and state-level campaigns that turned the 2010 redistricting cycle into lasting partisan advantages.
Antidemocratic: Inside the Far Right’s 50-Year Plot to Control American ElectionsInvestigates the legal campaign around voting rights and John Roberts that helped remove constraints on partisan and racial district manipulation.Recommended videos
11:40Weird maps win elections – Gerrymandering explainedMap Men’s funny, accurate explainer of how district lines decide elections.
3:53Gerrymandering: How drawing jagged lines can impact an election – Christina GreerTED-Ed animation of packing and cracking, the core gerrymandering techniques.
2:42Gerrymandering, explained | The Washington PostWashington Post short showing how the same votes yield different results under different maps.Cite this article
Owen Parker. “The mathematics of gerrymandering: why democracy’s greatest flaw is geometric.” Eikleaf, 6 May 2026, updated 2 November 2026. https://www.eikleaf.com/the-mathematics-of-gerrymandering-why-democracys-greatest-flaw-is-geometric/
Gen AI Disclaimer
Some contents of this page were generated and/or edited with the help of a Generative AI.
Media
Second round of the French presidential election of 2007. – Wikipedia
Key Sources and References
Brennan Center for Justice (Sundeep Iyer). “Redistricting and Congressional Control Following the 2012 Election.” November 29, 2012. https://www.brennancenter.org/our-work/research-reports/redistricting-and-congressional-control-following-2012-election
PolitiFact. “Steny Hoyer says House Democrats won majority of 2012 popular vote.” February 19, 2013. https://www.politifact.com/factchecks/2013/feb/19/steny-hoyer/steny-hoyer-house-democrats-won-majority-2012-popu/
Eric Petry. “How the Efficiency Gap Works.” Brennan Center for Justice. 2017. https://www.brennancenter.org/sites/default/files/legal-work/How-the-Efficiency-Gap-Works.pdf
Karl Rove. “The GOP Targets State Legislatures.” The Wall Street Journal. March 2010 (paywalled); discussed in Justin Levitt, “Karl Rove Is Right About Importance of Local Elections,” Brennan Center for Justice, March 24, 2010. https://www.wsj.com/articles/SB10001424052748703862704575099670689398044 · https://www.brennancenter.org/our-work/analysis-opinion/karl-rove-right-about-importance-local-elections
FactCheck.org. “Republican State Leadership Committee.” August 2010. https://www.factcheck.org/2010/08/republican-state-leadership-committee/
ProPublica. “How Dark Money Helped Republicans Hold the House and Hurt Voters.” December 21, 2012. https://www.propublica.org/article/how-dark-money-helped-republicans-hold-the-house-and-hurt-voters
WBUR / NPR. “‘Gerrymandering On Steroids’: How Republicans Stacked The Nation’s Statehouses.” Here & Now. July 19, 2016. https://www.wbur.org/hereandnow/2016/07/19/gerrymandering-republicans-redmap
Whitford v. Gill, 218 F. Supp. 3d 837 (W.D. Wis. 2016). District court opinion. https://www.scotusblog.com/wp-content/uploads/2017/04/16-1161-op-bel-dist-ct-wisc.pdf
Capitol News Illinois. “New congressional maps signed into law.” November 24, 2021. https://capitolnewsillinois.com/news/new-congressional-maps-signed-into-law
Davis v. Bandemer, 478 U.S. 109 (1986). Supreme Court of the United States. https://www.law.cornell.edu/supremecourt/text/478/109
Vieth v. Jubelirer, 541 U.S. 267 (2004). Supreme Court of the United States. Syllabus and Kennedy, J., concurring. https://www.law.cornell.edu/supct/html/02-1580.ZS.html · https://www.law.cornell.edu/supct/html/02-1580.ZC.html
Gaffney v. Cummings, 412 U.S. 735 (1973). Supreme Court of the United States. https://www.law.cornell.edu/supremecourt/text/412/735
Gill v. Whitford, 585 U.S. 48 (2018). Supreme Court of the United States. https://www.law.cornell.edu/supremecourt/text/16-1161
Rucho v. Common Cause, 588 U.S. 684 (2019). Supreme Court of the United States. https://www.law.cornell.edu/supremecourt/text/18-422
Thornburg v. Gingles, 478 U.S. 30 (1986). Supreme Court of the United States. https://www.law.cornell.edu/supremecourt/text/478/30
Louisiana v. Callais (U.S. Supreme Court, April 29, 2026). National Constitution Center. “The Supreme Court’s Callais decision sets new framework for racial gerrymandering.” April 30, 2026. https://constitutioncenter.org/blog/the-supreme-courts-callais-decision-sets-new-framework-for-racial-gerrymandering
Pew Research Center. “Behind Trump’s 2024 Victory, a More Racially and Ethnically Diverse Voter Coalition.” June 26, 2025. https://www.pewresearch.org/politics/2025/06/26/behind-trumps-2024-victory-a-more-racially-and-ethnically-diverse-voter-coalition/
Nicholas O. Stephanopoulos and Eric M. McGhee. “Partisan Gerrymandering and the Efficiency Gap.” University of Chicago Law Review, Vol. 82, No. 2, 2015. https://chicagounbound.uchicago.edu/uclrev/vol82/iss2/4/
Karthik Seetharaman. “A Comparison of Metrics for the Identification of Partisan Gerrymandering.” arXiv:2111.02540. 2021. https://arxiv.org/abs/2111.02540
Gregory S. Warrington. “Quantifying Gerrymandering Using the Vote Distribution.” Election Law Journal, Vol. 17, No. 1, pp. 39–57. 2018. https://arxiv.org/abs/1705.09393
Moon Duchin et al. “Locating the Representational Baseline: Republicans in Massachusetts.” 2018. https://arxiv.org/abs/1810.09051
MGGG — Metric Geometry and Gerrymandering Group. GerryChain. https://github.com/mggg/GerryChain
MGGG — Metric Geometry and Gerrymandering Group (DeFord, Duchin, and Solomon). “Recombination: A Family of Markov Chains for Redistricting.” Preprint arXiv:1911.05725, 2019. Published in Harvard Data Science Review, Vol. 3.1, 2021. https://arxiv.org/abs/1911.05725
Carnegie Mellon University. Pennsylvania Supreme Court gerrymandering decision cites Wesley Pegden’s testimony. January 2018. https://www.cmu.edu/news/stories/archives/2018/january/gerrymandering-decision.html
ScienceDaily (Washington State University). “Open source tool can help identify gerrymandering in voting maps.” May 6, 2021. https://www.sciencedaily.com/releases/2021/05/210506105420.htm
Daryl DeFord and Ellen Veomett. “Bounds and Bugs: The Limits of Symmetry Metrics to Detect Partisan Gerrymandering.” Election Law Journal 25(2): 127–153 (online 2025). https://doi.org/10.1089/elj.2024.0038 · https://arxiv.org/abs/2406.12167
Thomas Ratliff, Stephanie Somersille, and Ellen Veomett. “Don’t Trust a Single Gerrymandering Metric.” La Matematica (Springer Nature), 2025, Vol. 4, pp. 764–809. https://link.springer.com/article/10.1007/s44007-025-00172-y · https://arxiv.org/abs/2409.17186
Matthew R. Francis. “Math Can’t Solve Gerrymandering.” Scientific American. March 14, 2024. https://www.scientificamerican.com/article/math-cant-solve-gerrymandering/
CBS News. “California voters pass Prop 50 to redraw the state’s congressional maps.” November 5, 2025. https://www.cbsnews.com/news/california-prop-50-results-2025/
California Secretary of State. “Proposition 50: Analysis by the Legislative Analyst.” Official Voter Information Guide, Statewide Special Election, November 4, 2025. https://vigarchive.sos.ca.gov/2025/special/proposition/50/analysis.htm
California Citizens Redistricting Commission. Mapping Playbook, Draft 3.1. October 6, 2021. https://wedrawthelines.ca.gov/wp-content/uploads/sites/64/2022/10/Mapping_Playbook_-_v.3.pdf












