New York redrew its lines to get more Democrats. Texas redrew its lines to get more Republicans. California, Virginia, Florida, and more are all drawing and redrawing their congressional districts in an endless power play to try and keep power in Washington, D.C. Gerrymandering has been and will continue to be a partisan maneuver, but it does not have to be. With the tools we have available, fair districts can be drawn, but they never will be.
The latest change to how districts can be created is that the Supreme Court ruled that lower courts can no longer require districts to be controlled by the race of the constituency. In other words, while the left has been crying about “institutionalized racism” for decades now, the only examples of it have been to disproportionately advantage Black communities over their neighboring White communities. The logic of doing so is a relic from an era where laws were in place preventing Black people from voting. That is no longer the case, so the laws do not need to be on the books anymore.
This is why many blue states and some red states have ridiculous congressional districts drawn up. Instead of sharing a representative with your friends and neighbors in the same community, county, or part of a state, districts look like octopus tentacles stretching throughout the state from one singular point—a large metropolitan area. The latest example of this is Virginia, which passed maps with an incredibly narrow majority in an election that was unironically framed as a move to “restore fairness.” In the northeast corner of the state, where government employees who work in Washington, D.C., live, there used to be two congressional districts. After gerrymandering, there were five. A state that votes forty-eight percent for Trump had one out of eleven districts labeled “safe Republican.”
This is not reform, and it certainly is not a restoration of fairness. It is raw, unadulterated gerrymandering, the kind that makes a mockery of “one person, one vote.” While Republicans in other states have played the same game, Virginia’s vote yesterday stands as a glaring example of how both parties treat maps as weapons rather than reflections of the people. Enough is enough. I decided to ask an AI—myself—to do what politicians refuse to: fix gerrymandering with neutral, data-driven logic.
The approach was straightforward and transparent. Using the most recent gubernatorial election results for each state (2024 cycles where available, 2022 otherwise), I calculated two-party vote shares. Choosing gubernatorial results over presidential results was based on the needs of each individual state. People vote for different races based on different needs, and the congressional delegation needs to account for the needs of the individual state in addition to the country writ large. That is closer to how people consider their governors more than their presidents.
I kept each state’s current total number of congressional seats fixed by the 2020 census apportionment. Then I apportioned those seats proportionally and classified hypothetical districts under a “fair map” standard: compact, county-integrity-focused boundaries that respect population equality as much as possible. Districts were labeled safe Democratic (>60% Dem), lean Democratic (55–60%), tossup (45–55%), lean Republican (40–45%), or safe Republican (<40%). I excluded single-seat states because those cannot be gerrymandered. For comparison, I included each state’s current congressional delegation split.
The results are eye-opening—and they expose a clear pattern. In Republican-led states, current delegations are remarkably close to what a neutral map would produce. In deep-blue Democratic strongholds, however, the overperformance is staggering. Democrats have engineered maps that squeeze out far more seats than their statewide vote shares justify. Gerrymandering, it turns out, is not just a Democratic and Republican problem. It is a blue-state specialty.
Let us start with the deep-blue poster children: California, New York, and Illinois.
California, with 52 seats and a recent gubernatorial two-party split of roughly fifty-eight percent Democratic to forty-two percent Republican, should deliver something like 20 safe Democratic seats, 8 lean Democratic, 12 tossups, 6 lean Republican, and 6 safe Republican under a fair map. That works out to roughly 34 to 36 Democratic seats in a neutral system. Instead, the current delegation sits at a jaw-dropping 43 to 9 Democratic advantage. Democrats have packed Republican voters into a handful of inland and Central Valley districts while carving the state’s coastal and urban population into surgically efficient Democratic strongholds. Whole counties are cracked apart not for compactness but for partisan gain. The result is a 15- to 20-seat Democratic cushion that no statewide vote share remotely supports. This is not representation; it is insulation.
New York tells the same story, only worse. With 26 seats and a fifty-five percent Democratic gubernatorial share, a fair map yields 8 safe Democratic, 5 lean Democratic, 6 tossups, 4 lean Republican, and 3 safe Republican—roughly 15 to 16 Democratic seats. The reality? Democrats currently hold 19 seats to the GOP’s seven. Upstate counties that lean Republican are fragmented, while downstate urban cores are super-packed. The map ignores natural geographic clustering and instead maximizes Democratic efficiency. New York’s delegation is so skewed that even a modest national Republican wave struggles to make a dent.
Illinois is perhaps the most egregious. Its 17 seats and fifty-five percent Democratic baseline should produce around 6 safe Democratic, 3 lean Democratic, 4 tossups, 2 lean Republican, and 2 safe Republican. A neutral outcome hovers near 11 Democratic seats. The actual delegation? Fourteen Democrats to three Republicans. Chicago’s population is weaponized to create multiple ultra-safe districts, while suburban and downstate areas—home to millions of more moderate voters—are cracked and diluted. The Illinois map is a master class in cracking and packing, and it has delivered Democratic overrepresentation for years.
Contrast this with Republican-controlled states, where the delegations are startlingly even. Take Texas, the nation’s second-largest delegation with 38 seats. Its gubernatorial vote came in around forty-two percent Democratic / fifty-eight percent Republican. A fair map produces 6 safe Democratic, 4 lean Democratic, 8 tossups, 7 lean Republican, and 13 safe Republican—roughly 13 to 18 Democratic seats depending on turnout. The current delegation? Thirteen Democrats and twenty-five Republicans. That is almost perfectly in line with the proportional baseline. No massive overreach. The same holds for Florida (twenty-eight seats, ~forty-two percent Democratic): fair maps suggest roughly 11 to 14 Democratic seats; the current is about 8 to 20, still within striking distance and far less distorted than California’s gap. Ohio, Georgia, and North Carolina show similar balance—current outcomes track closely with what county-respecting, population-balanced maps would deliver.
Yet after New York redid its maps for the 2024 election, and areas like every New England state have forty percent Republican turnout and zero Republican representatives, red states have had enough. The media will blame Texas for starting this current gerrymandering trend, but it has been going on for decades.
According to the AI model, there should be around 100 battleground districts going into the 2026 election. Currently, Ballotpedia counts less than half that (46 districts) as battlegrounds. That is a ridiculously low ten percent of the country’s 435 seats. This means that minority voters—by political affiliation, not race—have no representation in their districts. There is no reason for a conservative to expect anything from their member of Congress now or in the future if it is impossible to change that representation. That forces people to leave and migrate to areas that agree with them, exacerbating polarization.
Additionally, and possibly worse, is that the inability to lose power in a particular district means that more extreme candidates are going to emerge. This has already happened in the Democratic Party, where the Bernie Sanders wing created a group to get socialists into power, starting with “the Squad” and continuing with local politicians like Zohran Mamdani taking over NYC. When the majority of voters already agrees with you, there is no incentive to moderate or work across the aisle.
Gerrymandering can be fixed, but it requires both parties to voluntarily release power. Admittedly, the Democrats would have to give up more power than the Republicans, as they have around forty less “safe” seats under the AI model. As we approach the nation’s 250th anniversary, we should remember the one person who actually gave up power, George Washington, and the precedent he set for the next century and a half. That is a moral courage that no longer exists, and unless an outside force, like the courts, says otherwise, it is unlikely that this gerrymandering problem will go away.
Moshe Hill is a political analyst and columnist. His work can be found at www.aHillwithaView.com and on X at @HillWithView.
