ANALYSIS — Turnout in the 2022 midterm elections will be down. Guaranteed. But it’s not because of any voting access legislation percolating at the state and local levels. It’s because turnout is always down in midterm elections compared with presidential years.

The more pertinent question is: How low will turnout go?

At the outset, it’s important to remember that the last two elections set modern records for turnout.

In 2018, turnout was 50.1 percent, the highest for a midterm election since 1914, according to data from the United States Election Project at the University of Florida. Turnout in 2020 was 66.8 percent, which was the highest turnout for a presidential election since the 1900 race between President William McKinley and William Jennings Bryan, before women were even allowed to vote.

What’s been driving voters to the polls recently? There’s no question in my mind that one person deserves credit for the phenomenon: Donald J. Trump. He energized Republican voters for him and Democratic voters against him, leading to record-breaking turnout. Love him or hate him, Trump has been good for turnout.

While Trump is trying to maintain a public presence without Twitter and is interjecting himself into 2022 races, he won’t be on the ballot next year, which could decrease the fervor for and against him. Of course, he wasn’t technically on the ballot in 2018 either, but that election was predominantly about electing a Congress to support or oppose his policies. 

Going forward

So what will turnout look like in 2022?

It’s almost certain to be lower than 2020 because that’s what always happens in midterms. At least almost always. Turnout hasn’t gone up from a presidential cycle to the subsequent midterm since well before the Civil War. Turnout was 56.5 percent in 1836 and 70.8 percent two years later in the midterms. So declaring that turnout will be down in 2022 compared with 2020 is in the early running for the least provocative prediction of the cycle.

How much lower will turnout be? 

The average difference in turnout between a midterm and presidential election in recent history is 16 points. If that happens next year, it would put 2022 turnout back to the record midterm of 2018 at 50 percent. Obviously, that’s possible since it just happened, but with Trump out of the White House and potential voter complacency with the end of the pandemic and a growing economy, 2022 turnout could sink lower. In other words, less urgency could produce fewer voters.

In the 40 years before Trump was elected president, midterm turnout was very consistent at an average of 39 percent, with a high of 42 percent (in 1982) and a low of 36.7 percent (in 2014).  

If midterm turnout fell back to that pre-Trump average, it would be the largest difference between a presidential election and midterm since the 1940s. That seems unlikely, particularly since early fundraising totals this year indicate that donors remain very engaged in the process. ActBlue, the online contribution platform favored by Democrats, announced this week that 2.6 million donors gave $314 million during this year’s first quarter, up from fewer than 2 million giving $175 million in the first quarter of 2019. That’s one sign that voters haven’t yet tuned out en masse. 

Party splits

At this early stage, 2022 turnout seems likely to fall close to or just below 2018, based on the average historical difference between presidential and midterm elections. But a key factor is whether this decline in turnout would be equal between the two parties. 

For example, overall turnout was nearly identical in the 2006 (40.4 percent) and 2010 (41 percent) midterms, but the makeup of the electorate was significantly different and that led to significantly different outcomes.

In President George W. Bush’s second midterm, in 2006, 38 percent of the electorate self-identified as Democrats compared with 36 percent who said they were Republicans, according to the exit polls. (Party ID had been even in the preceding presidential election.) Self-described independent voters supported Democratic candidates for the House in 2006, 57 percent to 39 percent, and Democrats gained 31 House seats and made Nancy Pelosi speaker of the House for the first time. 

In 2010, in President Barack Obama’s first midterm, self-identified Republican and Democratic voters each made up a similar 36 percent of the electorate, according to the exit polls, after Democrats had a whopping 7-point edge on party ID in the preceding presidential race. Self-described independent voters supported GOP candidates for Congress in 2010, 59 percent to 41 percent, and Republicans gained 63 House seats and took over the majority. 

According to the 2020 exit polls, self-identified Democrats outnumbered Republicans by just a point (37 percent to 36 percent). And the composition of the 2022 electorate will be a key indicator as to which party has the advantage in the midterms. 

So even without states rolling back vote-by-mail programs adopted in response to COVID-19 or imposing new ID requirements, both parties face a challenge of turning out their voters next year. Republicans have yet to prove that the full Trump coalition will come to the polls when he’s not on the ballot, and Democrats have to keep their voters energized when Trump has been defeated and isn’t an omnipresent threat.

In short, yes, the midterms will all come down to turnout. (Another contender for the least provocative prediction of the cycle.)

Nathan L. Gonzales is an elections analyst with CQ Roll Call.