The Generic Ballot Model and the 2022 Midterm Election

Abstract

By the final weeks of the 2022 election campaign, there was a clear consensus among pundits and political analysts that Democrats were likely to experience a shellacking in the midterm elections, especially in the House of Representatives. Republican leaders and strategists were confident that a “red wave” or even a “red tsunami” was approaching. Even more objective observers such as Chuck Todd and Mark Murray of NBC News believed that a number of indicators were clearly pointing toward large GOP gains in the House, the most prominent being President Biden’s poor approval rating, which had been stuck in the low-forties for months. While many political observers expected Joe Biden’s poor approval rating to result in big Republican gains in the 2022 election, historically, presidential approval has not been a very accurate predictor of midterm seat swing. For the nineteen midterm elections between 1946 and 2018, the correlation of net presidential approval (approval-disapproval) with House seat swing was a rather modest .66 while the correlation with Senate seat swing was a very weak .36. Presidential approval explained only 44% of the variation in House seat swing and only 13% of the variation in Senate seat swing. One indicator that has been shown to produce more accurate forecasts of both House and Senate seat swing than presidential approval is the generic ballot—a question in which voters are asked which party they plan to vote for without providing names of individual House or Senate candidates. By combining the results of generic ballot polling with the number of House or Senate seats that the president’s