Pew Research Center

Turnout in this year’s U.S. House primaries is up, especially on the Democratic side

Voters at a polling station in Los Angeles for California’s primary election on June 5. (Xinhua/Zhao Hanrong via Getty Images)

Americans appear to be more engaged with this year’s midterm elections than they typically are. Not only do about half of registered voters report being more enthusiastic than usual about voting, up from 40% in 2014, but turnout has surged in the 31 states that already have held their congressional primaries – particularly among Democrats.

In those states, nearly 13.6 million people – or 10.1% of registered voters – have voted in Democratic primaries for the U.S. House of Representatives, according to a Pew Research Center analysis of state election returns. By this point in the 2014 midterm election cycle, fewer than 7.4 million people – or 6% of registered voters – had cast ballots in Democratic House primaries. (The same 31 states have held primaries as by this date in 2014.)

The total number of votes cast in Democratic House primaries so far this year is 84% higher than the total for the equivalent point in 2014. One reason: There have been a lot more contested primaries, which tend to attract more voters.

Republican turnout in House primaries also has increased, from a combined 8.6 million votes at this point in 2014 (7% of registered voters) to 10.7 million (7.9%) so far this year. But the increase is much smaller (24%) than on the Democratic side, and the total number of votes cast in Democratic House primaries is considerably higher. Overall turnout in U.S. Senate and gubernatorial primaries also is above 2014 levels.

In general, voter turnout falls in midterm elections relative to presidential election years, and primaries nearly always draw fewer voters than general elections. So even if the surge seen so far this year continues, final turnout rates for this year’s primaries likely will be low in absolute terms, even if they exceed 2014 levels. And based on past experience, partisan disparities in primary turnout don’t necessarily predict individual general-election outcomes.

Our analysis is based on official vote totals and voter registration figures from the states that have held primaries so far this year. (The next batch is in early August, making July an opportune time for a spot-check.) We included all valid votes for candidates, including write-in votes when reported, but excluded blank votes and other spoiled or void ballots. For comparability, we also excluded runoffs and special elections from the analysis.

The primaries held so far cover 308 House seats, which means there were potentially 616 contested Republican and Democratic primaries. In most cases, however, there’s only one candidate for the nomination (or, sometimes, none at all), so the actual number of primaries with at least two choices on the ballot is a lot smaller.

So far this year, 340 House primaries have been contested by at least two candidates, versus 251 in 2014. Most of that increase has been on the Democratic side, with 81 more contested Democratic House primaries this year (203) than in 2014 (122). By contrast, there have been only eight more contested Republican House primaries so far this year (137) than at this point in 2014 (129).

To date, more than 9.9 million people have voted in contested Democratic House primaries, more than twice as many as had voted in such races at this point in 2014 (fewer than 4.3 million). Turnout in contested Republican House primaries has risen too, but again less so than in the Democratic races: an increase of about 1.2 million votes between 2014 (5.7 million) and this year (just under 7 million).

The rules governing primaries can (and do) vary considerably from state to state, which can make it tricky to compare turnout across time, between states and among different offices.

Several states, such as Virginia, don’t hold primaries in uncontested races; some rely on party conventions to pick nominees, with primary elections as a backstop. In some states, parties limit their primaries to registered members; in other states, especially those that don’t register voters by party, primaries are open to anyone. California uses a “top two” system in which all candidates for a given office run in a single primary; the two gaining the most votes, regardless of party label, advance to the general election in November. This year, Maine used a “ranked choice” system in its primaries, in which voters ranked candidates in order of preference.

More often than you might think, one or the other major party might not even nominate someone for a particular office, depressing turnout while effectively ceding the general election to its main rival (and, in some cases, a batch of minor parties and write-in candidates). On the other hand, turnout in an uncontested race could be boosted by the presence of a different, contested race on the same ballot.

Turnout also has been higher in this year’s gubernatorial and Senate primaries, though the increases have been similar for both parties. (We analyzed those contests separately, even if they were on the same ballot as the House races, since some people may have voted in one or the other race but not all of them.) So far this year, around 16.8 million people have voted in 17 states’ regular Senate primaries, or 20.8% of those states’ registered voters. By this point in 2014, 9.7 million people had voted in 19 Senate primaries. (Comparisons are difficult because different sets of states voted in Senate races in 2014 and 2018, a consequence of senators’ staggered terms.)

Direct comparisons are easier in the 36 states that are choosing governors this year, because the same states did so four years ago. (Governors serve four-year terms in all states except Vermont and New Hampshire, where they serve for two years.)

So far, total turnout in the 20 states that have held gubernatorial primaries is 22.7 million (24.8% of these states’ registered voters), up from 14.9 million (18.4%) in 2014. (To be fair, there were no gubernatorial primaries in South Carolina four years ago, as both the Republican and Democratic nominees faced no competition, but that wouldn’t come close to explaining the gap: Only 608,451 people voted in this year’s gubernatorial primaries, both of which were contested.) A likelier reason is that there were a lot more incumbent governors running for re-election four years ago. Only four of the 20 gubernatorial contests held by this point in 2014 were open seats, compared with 12 this year.