The Charger Blog

University of New Haven Political Scientist Examines What Will Tip the Scales in U.S. Midterm Elections

Chris Haynes, assistant professor and coordinator of the University’s political science program, projects the Republicans maintain a majority in the Senate and the Democrats take control of the House.

November 6, 2018

By Chris Haynes, Assistant Professor of Political Science


Headshot of Chris Haynes
Chris Haynes, Assistant Professor

The election is upon us and pundits and political scientists like me are asked constantly, “How do you think the election will turn out?” In all truth, I have to admit that I do not know. But we have some data and past precedence to help make some educated predictions.

Today’s midterm elections pits the traditional political gravity of a humming economy against the polarizing nature of President Donald Trump. Political scientists such as me who are steeped in past election data would dare not violate the one rule of elections and voting: that good economies favor incumbents. Would we?

While this election, like our President, is certainly like nothing we have seen before, there are a few clues that might justify a semi-believable set of hypotheses.

While other indicators such as excitement to vote and early voting numbers seem relatively even, there is one important statistic that is not: presidential approval.

After the economy, one of the best metrics we can use to predict elections is the approval rating of the incumbent president. Unpopular presidents can often weigh down their own party and are often cited as the reason why late breaking undecided vote for the opposition party.

"And to put it bluntly, the current incumbent is not popular. That said, Trump’s unpopularity is not evenly distributed across the country. This matters because our midterm elections are just that, many regional and local elections."
What Does This All Mean?

So in order to make sense of this data, we must examine not only national approval ratings, but also regional and state ones. What the data shows is that President Trump is unpopular in the Midwest, Southwest, West, and Northeast and popular in the South, Southeast, and the Plains states. President Trump is unpopular in urban and suburban/exurban areas of the country while wildly popular in rural areas.

What does this all mean? Well, it means that Democrats could fare well in toss up races in states and districts that tend to disapprove of President Trump. Based on what we know about the races in the Senate that are projected to be within the margin of error in the most recent polls, we might expect Missouri, Montana, Tennessee, and Texas to go Republican, and New Jersey and Nevada to go Democratic. Overall, I predict that after today’s election the balance of power in the Senate will remain unchanged.

"On the other hand, the situation is very different in the House, where most of the competitive races are those from suburban or exurban districts filled with the type of voters who have tended to reject President Trump’s fiery brand of politics."

Districts such as California 45 and 48, Kansas 2, Colorado 6, and Virginia 7 and 10 should all go Democratic. And while there are a few districts like Pennsylvania 14 and Minnesota 8 that will almost certainly go Republican, they, based on the data, appear to be the exception.

In the end, I expect President Trump’s unpopularity with some traditional Republican constituencies such as white, college-educated women to cause the House to flip to the Democrats with a net change of +33 seats.

Finally, similar to what it taking place in the House, much of the gubernatorial map is being fought in states that have become unfriendly to President Trump. Nowhere is this more evident than in the Midwest, where I expect six Republican seats in states where Trump has a net negative approval rating to flip to the Democrats.

Predicting Connecticut's Gubernatorial Race

Closer to the Connecticut home front, I expect Democrat Ned Lamont to squeak by Republican Bob Stefanowski. Even on the heals of an unpopular Democratic governor, the national narrative and President Trump’s ubiquity will be too much for the Republicans to overcome in this Democratic-friendly state. Overall, I project Democrats should pick up a net +8 governor seats.

In all, midterm 2018 will return a mixed decision with the Republicans holding the Senate and the Democrats almost gaining parity in the gubernatorial landscape and winning control of the House Speaker’s gavel.

The one common theme across the country is that our current President will play a central role in the midterm elections. Would you expect anything less?