← Newsroom

Two thirds of courthouse elections had no opponent

A prosecutor decides who gets charged and who is offered a plea. A judge rules on bail, custody and eviction. A sheriff runs the jail. Across 6,358 elections for those three offices in 45 states in 2022 and 2024, 4,240 — two in three — had nobody running against the winner. These are not all tiny places: the typical one still drew 16,520 voters.

Commonwealth’s Attorney Vote for one

The incumbent Unopposed

No candidate filed

What 4,240 of these ballots looked like. Counted from certified precinct returns published by the MIT Election Data and Science Lab for 2022 and 2024.

Nobody publishes a national count of which American elections actually went uncontested. The results exist, but they sit in fifty states’ precinct files in fifty different formats. Contest Every Race and BallotReady both publish partial counts; neither breaks out the courthouse, which is where the pattern is sharpest.

One name, over and over

Share of general-election races with no opponent on the ballot, 2022 and 2024 combined. The three groups are mutually exclusive and sum to the total.

And it is getting worse, fast

The share is not steady. In 2022, 61.8% of these races had no opponent. In 2024, 71.4% did — nearly ten points in a single cycle, on a base of more than three thousand races each year. Two cycles is not a trend line, and this is the whole of what this source publishes; it is a change worth naming, not a curve worth extrapolating.

It is not a question of size

The easy explanation is that these are small places where one willing candidate is all there is. The data does not support it. Counting only the races that fill a single seat — where the vote total is genuinely the number of votes cast in that contest — the median uncontested courthouse race drew 16,520 votes. Nine in ten drew more than a thousand. Forty-nine drew more than a million.

At the top of the range sits a state supreme court. In November 2024 Steve González, then the sitting chief justice of Washington, was returned to the bench with no opponent on the ballot. Washington’s own certified results record 2,564,372 votes for him and 61,866 write-ins — 97.6% of everything cast. On the same ballot, Justice Sheryl Gordon McCloud was returned the same way on 2,534,783 votes. The third seat up that year, Position 2, was contested, and it was decided by about twenty-one thousand votes.

Source: Washington Secretary of State, certified November 2024 results. Washington elects its judges without party labels, so nothing here is a party contest.

At the other end the numbers stop looking like elections at all. In the town of West Almond, in Allegany County, New York, the 2024 contest for town justice drew 130 votes. The justice took 103 of them. The remaining 27 were write-ins, and the county’s own results sheet lists them by name: one of the people written in for the office was Harriet Tubman.

The gap is real — except on the bench

The pattern is not only that these races are quiet. It is who leaves them quiet. Counted per seat, a Republican won 71.1% of their 3,091 courthouse seats without an opponent; a Democrat, 64.0% of their 1,340. But that headline hides three very different offices.

Sheriffs are where the familiar story holds hardest: a Republican sheriff faced nobody in seven of ten elections, a Democratic one in fewer than half. Prosecutors show a real but smaller gap, and both parties sit above 80% — this is an office almost nobody challenges, whoever holds it.

Judges are the exception, and the exception matters. There the gap disappears entirely, and the Democratic figure is marginally the higher of the two. On top of that, 1,552 judicial seats — more than a third — carry no party label at all, which puts them beyond the reach of any party recruiting a slate. Any account of this that runs “one party is letting the other walk into these offices” is describing sheriffs, and to a lesser degree prosecutors. It is not describing the bench.

The national picture beyond the courthouse is starker, and it comes from the organisation that tracks it. Contest Every Race — a Democratic group, and one reporting on its own party’s record — found that in 2022 Democrats did not compete in 51% of all partisan races, some 14,894 contests, against 14.5%, or 4,224, left empty by Republicans. Those are that organisation’s own published counts across a far wider set of offices than this piece covers, quoted as theirs and not recomputed here.

Recruiting works where it is tried

Contest Every Race, founded in 2018 inside the organising outfit Movement Labs, exists to find candidates for exactly these elections. Its published tally is 12,807 candidates recruited and 5,660 elected — a 44% win rate.

And the model scales when a state party commits to it. In December 2025 Texas Democrats recruited 104 candidates and filled every state and federal race on the 2026 ballot — the first full slate either party has managed in modern Texas history, with nine in ten of the recruits never having run for anything before.

That is patient, unglamorous work — finding a person willing to put their name on a ballot in a place nobody is watching, thousands of times over — and the record shows it moves the number wherever it is pointed. The courthouse is simply somewhere it has not been pointed yet: prosecutor, judge and sheriff sit below the level most recruitment programmes reach.

You cannot recruit for a seat you cannot name

The reason no one has counted this is not that the results are secret. It is that they arrive as fifty states’ precinct files in fifty different shapes, and nothing joins them. Civly consolidated the 2022 and 2024 certified returns into 50,684 race-level results covering all fifty states and the District of Columbia, then flagged the contests nobody entered.

Across every office, not just the courthouse, 24,426 elections in those two cycles had no opponent — each one carrying the office, the district, the county and the name of whoever won it unchallenged. Of those, 973 seats drew no challenger in either cycle, and 6,047 were decided by fewer than a thousand votes. That layer — certified results joined to candidates, districts and party — is the same one underneath every Civly report and donor list.

How this was counted

Source

Certified precinct-level returns published by the MIT Election Data and Science Lab, collected from state and county election officials, for 2022 and 2024. The lab’s method for standardising these returns is written up in Scientific Data, and the datasets are archived at the Harvard Dataverse. Pulled and recomputed 13 August 2026. Precinct rows were consolidated into 50,684 race-level results, of which 47,495 are contests between people rather than ballot questions, and 6,358 are the three offices examined here.

Which offices count

Judges include trial, appellate and supreme courts, magistrates, surrogates, probate and orphans’ courts and justices of the peace. Prosecutors include district attorneys, commonwealth’s and state’s attorneys, prosecuting attorneys and elected solicitors; where a title could belong to either, prosecuting wins, so Georgia’s State Court Solicitor is a prosecutor rather than a judge. Court clerks, registers and recorders are excluded — they are elected in the same building but do not decide cases. Two traps had to be closed by hand: Kentucky’s county judge executive is a county administrator, not a judge, and the town of Courtland, Virginia returns a town council whose name contains the word “court”.

What “no opponent” means

No more names were printed on the ballot than there were seats to fill. Write-in votes count towards turnout but are not treated as an opponent — several states record every person written in, so an unopposed sheriff can otherwise appear to face hundreds of “candidates” on one or two votes each. West Almond’s write-in for Harriet Tubman is the clearest illustration of why that rule is necessary.

Seats and races are different counts

A few benches elect several judges in one contest, so 6,358 races produced 6,493 winners. Party figures count winners, one per seat. Of the 4,240 races with no opponent, 135 — 3.2% — filled more than one seat and so carried more than one name; the other 4,105 offered a single name. Where the size of a race is discussed, only single-seat races are used, because a multi-seat total sums the votes of every winner and is not the number of people who voted.

Party labels

The party shown is the one the winner’s own state recorded on the ballot. “Nonpartisan” is a legal designation in states that strip party from judicial elections; seats with no label filed are counted separately and never assumed to belong to either party.

How the individual races were checked

Every race named in this article was confirmed against the state or county that published it, not against the research file alone. The Washington figures come from the Secretary of State’s certified return and match to the vote. West Almond was checked against Allegany County’s own results sheet, which records 130 votes cast and 103 for the winner against the 132 and 104 in the certified research file — a two-vote difference between the county’s election-night file and the certified return, and the reason this check is worth running. Race totals were also validated against the real size of each legislative chamber rather than against the data itself.

What this cannot tell you

Stated up front, because a finding is only as good as the boundary around it.

An empty race is not proof of a problem

Judicial ethics rules discourage campaigning, some incumbents are simply respected, and some places are small enough that one willing candidate is all there is. Nothing here measures whether any of these officials did the job well.

A floor, not an exact count

A handful of states file scattered write-ins as though the person had been on the ballot, which makes a genuinely unopposed race look contested. Measured on these offices it is small — 31 races, about half a percentage point, 24 of them in Vermont — but the error runs one way only. The true share is 66.7% or a little higher, never lower.

No state ranking, deliberately

States publish wildly different amounts of their own ballot: Indiana names 1,945 distinct offices in these two cycles, Tennessee 5. Ranking states would measure how much each one discloses, not how competitive it is. That is also why the office totals cover 30 to 38 states rather than all 51.

Two cycles, and no long trend

2020 is not published in this series, so the comparison is 2022 against 2024 and nothing earlier. Two points do not establish a direction, and nothing here says what is on the 2026 ballot or when filing closes.

One number hides three different offices

Across all courthouse seats, Republicans won 71.1% of theirs without an opponent against Democrats’ 64.0%. Almost all of that gap is sheriffs, where it runs 71.5% to 45.9%. Among judges there is no gap at all — 64.6% Republican, 65.5% Democratic — and more than a third of judicial seats carry no party label to begin with. Used on its own, the 71.1-against-64.0 figure reads as a statement about the whole courthouse. It is a statement about sheriffs.

Known holes in the source

Nine of the 6,358 courthouse races — all Indiana in 2022 — arrive with no vote total at all, and are excluded from anything about size. Nebraska files many offices without a district, which splits some contests across counties. Both are documented rather than repaired, and neither moves the headline: dropping every zero-vote race leaves the rate at 66.7%.

Classification is ours

Which office names count as a judge, a prosecutor or a sheriff is Civly’s rule, applied to the 2,557 distinct office titles that mention a court, a prosecutor or a sheriff. It is not a designation any state supplies. Reasonable alternatives move the top line by a few tenths of a point — treating county attorneys as prosecutors, for instance, raises the overall rate to 67.5% — and never change the ordering of the three offices.

---