The maps on this ballot were drawn mostly by judges, not voters

Most of the people who voted in House races this week did so inside district lines that neither they nor, in several of the most consequential cases, their own state legislatures got the final word on. The 2026 midterms are the product of the most intensive mid-decade redistricting push in more than a century — ten states redrew congressional maps between the 2024 and 2026 elections, according to a tracker maintained by the National Conference of State Legislatures — and in the biggest fights, it was not legislatures or voters who settled what the maps would look like. It was judges, often by narrow margins, often on grounds that had little to do with which party's voters were being helped or hurt.

That matters for reasons beyond who wins which seat. A legislature that loses a redistricting fight can be voted out. A ballot measure that fails can be put to voters again. A map that a court throws out on a technicality, months after voters approved it, leaves no comparable recourse before the next election is already underway. This cycle produced that outcome more than once, and examining why reveals less a coherent doctrine than a patchwork shaped by which court happened to hear which case.

Two voter-approved maps, two different fates

California and Virginia both asked voters directly to approve new, Democratic-favored congressional maps in the past year. Only one map survived.

In California, voters passed Proposition 50 at a special election in November 2025, authorizing the legislature's map for the next three election cycles in place of the one drawn by the state's independent redistricting commission. California Republicans, backed by the Trump administration, asked the Supreme Court for an emergency injunction blocking it. The justices declined, in a brief, unsigned order with no dissents noted on the docket in Tangipa v. Newsom. California's secretary of state confirms the map is now in effect, drawn to help Democrats pick up as many as five seats.

Virginia's experience ran the opposite direction. Voters there approved a Democratic-backed redistricting amendment at an April 2026 referendum, by a margin of roughly 51.7 percent to 48.3 percent on about 3.1 million votes cast. The Supreme Court of Virginia then threw the result out, ruling 4-3 that the legislature had violated the state constitution's procedural requirements by authorizing the referendum after early voting had already begun for a prior election. The court found the legislature had missed a constitutional deadline, not that the map itself was unlawful or that voters had been misled about what they were approving. Virginia Democrats took the case to the U.S. Supreme Court, which declined to intervene, leaving the state's 2021 map — drawn under a bipartisan process after that commission deadlocked — in place for this week's elections.

Put plainly: one state's voters got the map they chose at the ballot box. Another state's voters, who approved a similar reform by a similar margin, did not, because of a filing-deadline dispute that had nothing to do with the substance of their vote. Nothing about the underlying legal theories in either case compelled that divergence; the outcomes turned on which court each dispute landed in and how each court chose to read its own state's procedural rules.

A weakened Voting Rights Act made this terrain more volatile

The redistricting fights unfolding this cycle were also shaped by a Supreme Court ruling that had nothing to do with partisan control on its face. In Louisiana v. Callais, decided 6-3 in April 2026, the court struck down Louisiana's second majority-Black congressional district and significantly narrowed the circumstances under which Section 2 of the Voting Rights Act can be used to challenge a map as racially discriminatory. Justice Samuel Alito wrote for the majority; Justice Elena Kagan's dissent, joined by Justices Sonia Sotomayor and Ketanji Brown Jackson, did not mince words about what she saw the ruling accomplishing:

I dissent, then, from this latest chapter in the majority's now-completed demolition of the Voting Rights Act.

Whatever one makes of the merits, the practical effect showed up almost immediately in Texas. A three-judge federal panel had ruled in November 2025 that Texas's redrawn map likely amounted to an unconstitutional racial gerrymander and ordered the state back to its 2021 lines. Justice Alito stayed that order on the emergency docket, allowing the new map — expected to hand Republicans as many as five additional seats — to be used this cycle. In April 2026, the full court made that stay permanent, summarily reversing the lower court 6-3, with Justices Sotomayor, Kagan and Jackson again dissenting. Legal analysts tracking the Callais fallout have noted that the ruling effectively raised the evidentiary bar for any plaintiff trying to prove a map was drawn along racial rather than purely partisan lines — a distinction the court treats as constitutionally decisive even when the practical racial impact of a map is identical either way.

None of this required Congress to act, and none of it was decided by voters. It was decided by nine justices interpreting a 1965 statute, with consequences for how millions of ballots were drawn into districts this week.

The one case where a court protected, rather than overrode, a reform voters chose

Utah complicates any simple story about courts versus democracy. In 2018, Utah voters passed a ballot initiative, Proposition 4, meant to curb partisan gerrymandering. The Republican-controlled legislature later repealed large parts of it and adopted its own map anyway. A state judge, Dianna Gibson, ruled in November 2025 that the legislature's map violated the standards voters had enacted and ordered a court-drawn remedial map used instead for 2026 — a map expected to make at least one additional district more competitive for Democrats. The Utah Supreme Court declined to block that order, and a separate federal lawsuit brought by Republican members of Congress seeking to restore the legislature's map also failed, a result the League of Women Voters of Utah celebrated as a vindication of the 2018 initiative.

Utah is the mirror image of Virginia: a court enforcing what voters had directly chosen, against a legislature that tried to override it. That it is the exception rather than the pattern this cycle is itself the point. Across the ten states that redrew their maps — eight in ways that favored Republicans, two, California and Utah, in ways that favored Democrats, according to the NCSL's running tally — courts were the actual decision-makers far more often than legislatures completing an ordinary process or voters ratifying one at the ballot box.

• Texas: legislature's map upheld by the Supreme Court after a lower court found it likely an unconstitutional racial gerrymander.
• California: voter-approved map upheld after the Supreme Court declined an emergency challenge.
• Virginia: voter-approved map struck down by the state's top court on a procedural technicality; federal appeal rejected.
• Utah: legislature's map struck down by a trial judge enforcing a voter-passed 2018 reform; that ruling upheld on appeal.

Estimates of the cumulative partisan effect of this cycle's redraws vary — analyses from Bloomberg Government and Republican strategists tracking the maps have put the net advantage to Republicans anywhere from roughly five to well over a dozen seats nationally, depending on assumptions about which litigation resolves which way before votes are counted. The imprecision itself is instructive. In a House that will likely be decided by a handful of seats, the outcome of individual redistricting lawsuits — decided on procedural grounds in one state, evidentiary standards in another, a voter initiative's exact wording in a third — plausibly mattered more to control of the chamber than any single candidate's campaign.

Representation by litigation

None of this is an argument that any particular ruling was wrongly decided; reasonable judges can and did disagree, sometimes sharply, within the same courts. It is an argument that the aggregate effect of this cycle's redistricting litigation was to make unelected judges — federal and state, appointed under wildly different selection systems from one state to the next — the practical authors of how tens of millions of Americans are represented in Congress, in a cycle when the stated justification for intervention shifted case by case: a missed filing deadline in Virginia, an evidentiary standard in Texas, a flatly enforced ballot initiative in Utah, a one-line denial with no reasoning offered at all in California. A system that produces that much variance is not obviously applying a single principle about democratic legitimacy. It is applying whichever principle happens to be available in whichever court gets there first.

That is not a partisan complaint; the pattern cuts against both parties' preferred maps in roughly equal measure this cycle. It is an institutional one. The redistricting wars of 2025 and 2026 did not resolve the question of who should draw America's districts. They mostly confirmed that, for now, the answer is whoever is still litigating when the map has to be printed.