Did uncontested races cost the Democrats the House popular vote?

Yet readers have bombarded me with tweets and emails, questioning whether these vote totals accurately measure which party Americans preferred in 2022.

The complaint: Democrats failed to run any candidate at all in more districts than Republicans did. The implication is that if both parties had competed in every race, the Democratic vote total would rise — maybe allowing them to surpass the GOP.

So I did the math. And I don’t think those uncontested seats tipped the national popular vote to the Republicans.

Here are the gory details.

According to my count, there were 35 “problem” districts in 2022 — seats where one party failed to run a candidate. Democrats did not field a candidate in 23, while Republicans were not on the ballot in 12.

These districts were not all technically uncontested. Some featured a runoff between two Democrats or a third-party candidate facing off against just one major-party politician. But in each case, one of the two parties was locked out of vote counting on Election Day.

So, to estimate what might have happened had there been a major-party candidate in each race, I did a statistical two-step.

First, I used the 400 races where there were candidates from both parties to estimate how many people would have voted if both parties ran candidates in the 35 problem seats. For example, no votes were counted in Arizona’s 9th District, a Trump +26 seat where Republican Rep. Paul A. Gosar ran unopposed. But in Texas’s 4th District — another Trump +26 seat where GOP Rep. Pat Fallon faced a Democratic opponent — about 250,000 people voted for one of the two major party candidates.

So I estimated that roughly 248,000 people might have voted for Gosar or his Democratic opponent, if Democrats had run a candidate.

After estimating turnout in each district, I used a similar method to project who these hypothetical voters would have supported. Gosar, for instance, won zero votes in his uncontested district. But if Gosar had faced a Democratic opponent, he might have won 67 percent of the two-party vote. (Note that Fallon won 68 percent, after excluding third parties, in his equally red district.)

In Arizona’s 9th District, this math resulted in extra Republican votes. But most uncontested seats are more like Alabama’s 1st District — where Republican Rep. Jerry L. Carl earned 140,000 votes and Democrats got zero. Democrats still would have lost a hypothetical matchup in this district — but their vote total would have been much higher than zero.

The cumulative result of all these calculations: Republicans would have gone from winning the House popular vote by about three percentage points to winning by around two. That’s not much of a shift.

A caveat: These estimates are more of a first cut than a final word. As time goes on, I plan to include more data on candidate quality, which party fared better in state-level races and other potentially relevant factors. And the final uncounted votes might shave a bit more off the GOP lead.

But the basic facts of the 2022 election will almost certainly remain unchanged. Republicans won more House seats than Democrats, and they won more votes. The verdict of the voters doesn’t appear to have been distorted by uncontested races.

Update: As of Monday night, Arizona’s official statewide canvass shows Gosar with 192,796 votes and two write-in Democrats with a combined 4,389 votes. These votes do not change any of the statistical models or projections described above.

Technical notes: To estimate turnout, I used a regression model that predicted total two-party 2022 vote using 2020 total vote, the percentage of White voters in each district and the competitiveness of the district (as measured by absolute value of the 2020 presidential margin). The 2020 vote total was far and away the most powerful predictor. To estimate vote share, I used a regression model that predicted two-party share using 2020 results and incumbency.