Watch Now


Viewpoint: Predicting midterms with historical data

History tells us that the party not in power will regain control in the next election

History is a good indicator of what will happen at the polls this November, with Republicans expected to regain at least one chamber of Congress. (Photo: Wikimedia Commons)

Despite being the cause of endless speculation and news coverage, midterm elections are one of the most predictable events in American politics. This holds particularly true in the House of Representatives, where the president’s party has lost seats in 17 of the last 19 midterm elections.

Most of the losses in the House are by double digits, with some climbing especially high, like the 54 seats lost in 1994 and the 63 seats in 2010. The two years with gains by the president’s party were relatively small – only 5 seats in 1998 and 8 seats in 2002 – when Bill Clinton benefited from a strong economy and George W. Bush’s popularity surged after the 9/11 terrorist attacks.

Since World War II, the president’s party has averaged 7.4 points worse in the popular vote for the House midterms than during the preceding presidential election year. Since Democrats won the House popular vote by 3 points in 2020, Republicans are theoretically on track to win by 4.4 points in 2022. To take control, Republicans would need to win 5 extra seats.

The power of this historical trend differs slightly across the two Congressional chambers. House members serve two-year terms, with every seat up for grabs in every midterm and presidential election year, while Senators serve six-year terms, with roughly one-third of Senators facing the ballot each election cycle. Therefore, in the Senate, big swings are less likely, and results are more affected by the dynamic of individual races. Yet still, the party of the president has lost Senate seats in 13 of the last 19 midterms.

This year, 35 of the 100 Senate seats are up for election. As of now, Democrats hold a tight 50-50 majority, with Vice President Kamala Harris able to break any ties. Therefore, Republicans only need one extra Senate seat to gain control. The historical data mentioned, coupled with Biden’s low approval rating and recent polling, suggest they will be able to do so.

Governor contests tell a similar story. Only once in the last 19 midterms has there been a net gain in governorships for the president’s party. This November, 36 states will hold governor elections, where 16 of the incumbents are Democrats and 20 are Republicans.


So, why does this phenomenon occur? There are two theories that rise above the rest: the political pendulum and demobilized turnout.

The political pendulum is the idea that political power in the United States tends to shift back and forth between the two parties, especially after one party dominates a national election. This helps generate what is referred to as the presidential or midterm penalty – demonstrated by the common pattern of sinking approval ratings in a president’s first two years. In layman’s terms: people want what they don’t have.

Demobilized turnout accounts for the tendency of voters who favor the president to be less motivated to vote in midterms because they feel they are already in-power, and the president is not at risk on the ballot. Out-of-power voters are more incentivized to mobilize and vote to reassert control in Congress against an unfavorable sitting president.

Time is slipping away for President Biden and Democrats to avoid this seemingly predestined fate. A memo by political strategist Doug Sosnik details that in the last four midterm elections, by June the public had decided who they would vote for in November – Donald Trump’s 39% job approval in early February 2018, Barrack Obama’s 41% approval in June 2014 and 45% approval in June 2010, and George W. Bush’s 38% approval in March 2006, were all consistent with their performance at the midterms.

As of May 31, President Biden’s approval rating is at 40%. It has remained below 50% since August, when the United States completed its withdrawal from Afghanistan to end an over-20-year war. While most of the public favored miliary evacuation, the majority criticized the administration’s handling of the effort.

The withdrawal, combined with rising inflation and legislative failures from Sen. Joe Manchin’s continual repudiation, have left Democrats hurting in the polls. If they have any hope of reversing course, Democrats will need to do a better job at touting wins, like infrastructure and international cooperation during the Ukraine crisis, while recapturing the debate around Roe v. Wade, where they have the majority of public support.

Republicans, on the other hand, will try to ride this trend into Election Day. Their strategy has relied on emphasizing inflation and cultural shifts, while stirring party identity.

It appears to be working, as turnout is up among Republicans in the first 10 states that held primaries when compared to the 2018 midterm elections. In fact, Republicans currently have a 17% advantage in voter enthusiasm. The last time the gap was this wide was in 2010 when Democrats lost more than 60 seats in the House.

The history lesson here is that Republicans are a good bet to win in November to take control of Congress.

However, the effects may not be long-lasting. The presidential penalty usually reverses itself by the next election cycle – most presidents are reelected, and their party’s down-ballot performance improves.

If not acted on by incredible effort, the political pendulum continues to swing.