During the Obama years, Republicans won total control of the state government1 in Indiana, Michigan, Ohio and Wisconsin. Then, on Election Day in 2016, Hillary Clinton narrowly lost in Michigan, Pennsylvania and Wisconsin — states Democrats had won at the presidential level for more than two decades. She was easily defeated in Iowa and Ohio, which had tended to be close.2 Clinton barely won in Minnesota, another state where Democrats are usually strong. Post-election, there was a lot of doom and gloom about the Democratic Party’s prospects in the Midwest, with both nonpartisan analysts and even some party strategists suggesting the party needed a dramatic overhaul or risk losing in this region, which will be packed with white, working-class voters, for the foreseeable future.
Never mind all that now. Democrats are looking strong in the Midwest — it is by some measures the region where they are likely to make their biggest gains this November.
Let’s start with the House. Because all 435 House seats are on the ballot every two years, the House will provide the fullest picture of how the two parties are doing in 2018. According to the Classic version of FiveThirtyEight’s House forecast, Democrats are favored to pick up about 14 seats in the Midwest. That’s a big number on its own, but it’s even more significant when you think about that gain in terms of regions. Sure, the Democrats are likely to gain seats in the Northeast and Pacific, as you might expect, but they are making even bigger strides in the Midwest in terms of the number of seats they are expected to gain relative to the total number of seats available in the region.
To compare how strong Democrats look in one region versus another, we have to put regions with different numbers of seats on equal footing. To do that, we divided the number of seats Democrats are favored to win in each region by the total number of seats there. We then multiplied that by 100, which shows us how many seats the Democrats would be favored to pick up in each region if every region had 100 races. When all regions are normalized in this way, we can see that the Midwest is shaping up to be the party’s best region. (How you divide the states into regions is a matter of great and impassioned debate, of course; here we’re using an old-school FiveThirtyEight categorization. But yes, feel free to object to Pennsylvania being included in the Midwest if you like.)
|Region||Includes||Current||Proj.||Change||change per 100 races|
|KS, NE, ND, SD||0||2||+2||+23|
|PA, OH, MI, IN||16||24||+8||+15|
|UT, ID, MT, WY, AK||0||1||+1||+12|
|NY, NJ, MD, DC, DE||33||39||+6||+12|
|AZ, CO, NM, NV||12||15||+3||+12|
|IL, WI, MN, IA||20||24||+4||+11|
|FL, GA, NC, VA, SC||23||30||+7||+10|
|CA, WA, OR, HI||51||57||+6||+9|
|TX, AL, LA, MS||14||16||+2||+4|
|MO, TN, KY, AR, WV, OK||5||6||+1||+4|
|MA, CT, ME, NH, RI, VT||20||21||+1||+2|
The potential Democratic gains are scattered throughout the Midwest, from Pennsylvania (where they’re favored for four pickups) to Iowa and Kansas (two) to Michigan (one).3
It’s not just the House, either. Democratic incumbents are clear favorites to win their Senate races in Michigan, Minnesota (two races there, actually), Ohio, Pennsylvania and Wisconsin. They are modest favorites to hold on to seats in Indiana and Missouri — which were very red in 2016. Democrats are favored to win gubernatorial races in Illinois, Iowa, Michigan, Minnesota and Pennsylvania. The gubernatorial contests in Ohio and Wisconsin are toss-ups, and Democrats even have a chance in two very conservative states, Kansas and South Dakota. A win by Tony Evers in Wisconsin would be particularly significant for Democrats, who have failed in three different attempts to defeat Gov. Scott Walker.
What’s behind what appears to be a Midwestern revival for Democrats? Well, the most obvious answer is that Democrats didn’t have such a big problem in the Midwest in the first place — so it is entirely logical that they would do well there in 2018 because it is shaping up to be a good year for Democrats nationally.
There were a lot of stories as Obama left office about how Democrats had lost hundreds of seats at the congressional, state legislative and gubernatorial levels while he was president. But parties that hold the White House often lose seats in other places — Republicans had huge down-ballot losses when George W. Bush was president, for example, and it looks like they may have similar declines under President Trump.
The pattern seems to be that the best thing for a party that wants to make gains at the state and congressional level is to lose a presidential election first. In particular, midterm elections are typically rough for the president’s party, even if that president is popular — and Trump is fairly unpopular.
So we should have expected Democrats to bounce back overall after 2016 — and that’s exactly what’s happening. It’s not that Democrats are likely to make gains only in the Midwest, after all. Our Classic model shows them picking up at least one seat in every region in the country.
But why are the biggest gains likely to come in the Midwest? First of all, 2018 is confirming what previous election cycles had suggested — this is an area that swings a lot between the two parties. Democrats made major gains in the Midwest in 2006, then faltered in the Obama years and look to be having a revival now. And because Democrats lost in many key races in the Midwest in 2010 and 2014, there was some low-hanging fruit for Democrats to pick off in a year like this. (Like the governor’s mansion in traditionally-Democratic Illinois, for example.) Contrast the Midwest with the South, where Republicans have steadily gained ground for two decades and are likely to lose relatively few seats even in what appears to be a Democratic wave year.
Secondly and relatedly, national polls suggest that white voters without college degrees favor Republicans in 2018, but the margin between the two parties is likely to narrow compared to 2016, when Clinton lost that bloc by more than 30 percentage points. That shift has outsized influence in the Midwest, which has higher populations of white voters without college degrees than many other parts of the country. So the Democrats’ problem with white working-class voters may not be as severe as it looked on Election Day 2016 — which perhaps had more to do with the conditions in that election than the party overall. What we are seeing in 2018 suggests that working-class whites are not a single national bloc, but still vote much differently by state and region. Working-class whites in Southern states like Georgia and Texas are overwhelmingly opposed to Democratic candidates in key races this year, but they are less GOP-leaning in Midwest states like Ohio and Wisconsin.
I don’t think the Democratic strength in these races in the Midwest is particularly tied to the candidates who are running or any kind of major ideological or tactical shift within the party. Wisconsin Sen. Tammy Baldwin is a populist, Bernie Sanders-style Democrat and has embraced a Medicare-for-all health care proposal. Minnesota Sen. Amy Klobuchar is more of an Obama-Clinton Democrat in terms of ideology and has not embraced Medicare-for-all. Both are likely to win. Ohio Sen. Sherrod Brown is often described as being particularly good at connecting with voters, but Wisconsin gubernatorial challenger Evers is considered, even by some fellow Democrats, to be kind of dull and uninspiring — and he could win too. There aren’t a lot of nonwhite Democrats running in key races in this region, but several women are running — so Democrats have a slate of candidates in the Midwest that is broadly similar to the party overall in terms of demographics.
It helps that Michigan Sen. Debbie Stabenow and a few other Democrats in this region are incumbents, but Michigan’s Democratic gubernatorial candidate, Gretchen Whitmer, has a sizable lead too — and she has never before been elected to statewide office.
Two caveats to all of this. First, the Midwest had some of the biggest polling misses in 2016. So maybe, for example, Ohio’s Democratic nominee for governor, Richard Cordray, whose race is considered a toss-up, ends up losing by 5 points. I still think, even if the polls are inaccurately tilted towards Democrats, they will end up having a better night in this region than they did in 2016. Secondly, I’ve cast this story as Democrats doing better in the Midwest than their worst fears in the days immediately following Clinton’s loss. Those fears now look overwrought, but Democrats could still have cause for concern after Election Day 2018. There’s a real chance Democrats narrowly lose the governors races in Iowa, Ohio and Wisconsin next month. If the party can’t win in these states when there’s an unpopular Republican in the White House, when can they win?
Aaron Bycoffe contributed research to this article.