After a number of unexpected upsets in key 2018 primaries, you might be wondering how many surprises the November elections have in store. The short answer: very few.
Every two years, FairVote releases its Monopoly Politics report, an in-depth analysis of U.S. House of Representatives elections and our system’s problems of polarization, partisanship and lack of competition. The interactive component of the latest Monopoly Politics features two kinds of up-to-date projections for the 2018 House races: “high confidence” projections, for the very safest seats, and “full” projections, which project all seats based on competition level and expected vote share according to the national climate. Our model indicates that more than 54 percent of voters would need to favor Democrats in November for the Democrats to gain a narrow majority in the House of Representatives - a very high bar. But how confident are we in our projections?
FairVote’s high confidence projections have been over 99.9 percent accurate across three election cycles, projecting not only who will win each seat but also by how much. In 2016 our full projections were only off by two seats overall: incorrectly projecting close wins for seven Democrats and five Republicans. FairVote’s full projections also came quite close to the mark in terms of candidates’ shares of the overall vote in prior House Elections.
In 2014, FairVote’s vote share projections had a 92.4 percent fit with the election’s outcomes. (Click on the tabs in the interactive graphic below.)
In 2016 House elections, FairVote’s model had a 93.9 percent fit with actual results.
Comparing our 2016 projections with pure partisanship or previous district performance (the controls) shows these are more than just lucky guesses.
There were, of course, some instances, in which FairVote got it wrong.
As their name implies, swing districts are hard to call. Some districts might be very responsive to waves favoring big gains for one party. A weak campaign or a very high quality challenger can impact the outcome significantly. Nevertheless, the structure of our electoral system means that almost every incumbent who wins his or her primary is assured a seat in the 116th Congress.
FairVote’s methodology projects with “high confidence” only the very safest seats. Under our current system, we can predict 379 of the 435 House seats — or more than 87 percent of the total — with high confidence. Ideally, in a well-functioning democracy, the outcomes of hundreds of elections should not be so easy to predict with a high level of confidence. This lack of meaningful competition has left many voters feeling unrepresented and ignored. However, a more fair system is possible. Our Fair Representation Act Report, and corresponding page, explain how we can revitalize our democracy through electoral reform.
To see our projections and make some of your own check out our interactive spreadsheet. The spreadsheet also details the data behind our projections, such as national and state-by-state statistics and even the levels of competition for each seat.
Illustration by Mikhaila Markham