A shorthand explanation for midterm elections has long been that they are “referenda on the incumbent president.”

The theory postulates that the party of the White House has lost House seats in 36 out of 39 midterms since the Civil War and Senate seats in 19 out of 26 midterms since the direct election of senators began in 1914 because voters are typically unhappy halfway through a presidential term.

The reality, however, is that this oft-used phrase may better describe the effect than the cause.

Political scientists have a number of theories addressing why midterm elections turn out the way they do. The classic was “surge and decline,” which originated in the mid-1960’s with the late University of Michigan social psychologist Angus Campbell, best known for coauthoring the seminal book, The American Voter, with Phillip Converse, Warren Miller and Donald Stokes, a must-read for political science majors. In a nutshell, Campbell’s theory was that when a party wins the presidency, it generally meant that the composition of the electorate as well as the issue agenda favored the winning party, creating that outcome, but two years later, an electorate composed of a smaller pool of voters was less than optimal for that party. Hence, that party would “surge” when winning the presidency, but decline two years later.

Several elections, particularly the 1970’s, discredited that theory among many, as a revised version emerged from a very much alive, but I am pretty sure not related, James E. Campbell, a top-notch political scientist at the State University of New York at Buffalo (written when he was a professor at LSU!). One academic wrote a paper focused on the “Campbell vs Campbell” surge-and-decline debate. (For those who really want to dig into the science of midterm elections, it would be hard to do better than the work of Alan Abramowitz, James Campbell, and Gary Jacobson, all titans in the field of congressional elections.)

The assumption that midterms are referenda on presidents is fine when the White House occupant is terribly unpopular with lousy job-approval ratings, or (rarely) is exceedingly popular with high approvals. In the current era of evenly-divided parties and straight-ticket voting, with many automatically approving or disapproving of a president or candidate based solely on whether they wear a red or blue jersey, there are now quite a few times when president’s numbers fall somewhere in between, making the election outcome much less certain.

At this point, President Biden’s job approval ratings are averaging around 51 to 52 percent—Monmouth University’s 48 percent, and the Associated Press/NORC’s 59 percent bracket the averages, and Gallup’s 50 percent is just a touch low.

So, if a president’s approval numbers don’t shine as much light on the probable outcome, what does? One of the less admirable traits displayed by many Americans is that they are rarely motivated to vote by appreciation, admiration, satisfaction, affection, or any other positive feeling for a politician, especially an incumbent. Sadly, Americans are more likely to vote based on regret, outrage, hate, or fear.

Having advised over 500 political campaigns in 42 states, Nashville-based Democratic consultant John Rowley argues that science supports the view that “people will do something to avoid ‘fear of loss,’ that they will never do to get an ‘A’ on a paper or to “self-actualize.” In politics, negatives trump positives; it has long been believed that voters are much more likely to recall the information in a negative ad than in a positive one. Operationally, this tendency backs the theory that voters of a party that loses the presidency are more motivated to vote in the midterm election that follows than those on the side that won. They are angry and want revenge for the loss—or, these days, they think the previous election was stolen, enraging them more.

There is a private concern among many Democratic operatives who cannot imagine why Biden is so hated by the Republican base. As Rowley puts it, “I’m still stunned at how high the Republicans drove up Joe Biden’s negatives” and that “we had swing districts where his negatives were as high as Trump’s.”

In an email Monday, I explained to a Republican pollster that I was “trying to get into the heads of those who will, probably, or might vote next year, what are they thinking and at motivation levels for each party's base and early gut feelings of those in between (not a part of) those bases.” The response was: "The few in between (the two parties) are … exhausted. It’s becoming more and more clear by the day. And they’re very frustrated with Washington. We’ve started in-person [focus] groups again and it’s been very interesting to hear from the ‘in-between.’ They had hoped removing Trump from D.C. would remove some of the paralysis and in-fighting; they haven’t been pleased with what they’re seeing.”

The pollster, who prefers to remain anonymous, continued, “Those on the right—unfortunately—continue to go farther and farther to the right. While the policy message is 100 percent spot-on for right and center-right voters right now, the grassroots/OAN (the very conservative One America News Network) crowd refuses to move on from the stolen election and the need to fight to ‘save our country.’”

Some Democrats point to reapportionment or redistricting as items that could ruin the next election for them, while others cite efforts by Republican state elected officials passing laws to suppress the Democratic vote. While these fears are legitimate, the reality is that a party holding a Senate majority by virtue of a tie-breaking vice president and only five seats away from losing their House majority, has lots of things to fear.

Almost anything going wrong could doom Democrats, but nothing more than if there is a disparity in energy levels that works to their disadvantage. That could turn this election into one that lines up with the historic patterns in an even more decisive way.

This article was originally published for the National Journal on August 3, 2021.

Image credit: AP Photo/Susan Walsh

More from the Cook Political Report