Weak Spots in Democrats’ Strong Midterm Results Point to Challenges in 2020

Democrats had a great showing in the 2018 midterm elections. But even in such a strong year, they sometimes struggled to match their traditional support in electorally significant areas — with serious implications for 2020.

Their triumph was a somewhat narrow one, concentrated in well-educated, affluent communities. Overall, the distribution of Democratic and Republican support was reminiscent of the 2016 election, when Democrats won the popular vote but lost the Electoral College.

Yes, Democrats can muscle their way through those disadvantages with a big enough win, like their 7-point advantage in the House popular vote. But white voters without a degree are overrepresented in the most important Midwestern battleground states. The most straightforward alternative for Democrats goes through Florida, which probably gave Republicans their most promising results last week.

Florida was probably the biggest disappointment for Democrats last week, and Miami-Dade County the biggest single cause for it.

Hillary Clinton won Miami-Dade by 29 points; the incumbent Democratic senator, Bill Nelson, won it by only 21 points.

Democratic weakness in the county may well prove to be decisive in the Senate race and the governor’s race, once the recounts are done. The county includes 1 in 10 of the state’s voters, and, in general, Democrats got what they needed across most of the rest of the state. They won traditionally Republican Duval County (Jacksonville) and Seminole County (northern suburbs of Orlando; Sanford). They won Pinellas County (St. Petersburg) by a comfortable margin, and ran generally even or ahead of Clinton elsewhere in the state.

Democrats didn’t have a particularly good showing in Orange or Osceola counties, where growth among the Puerto Rican population was supposed to propel the party to statewide success. Low turnout seems a likely culprit, and lower turnout among nonwhites is not unusual in midterm elections.

But Democratic weakness in Miami-Dade County can’t be dismissed so easily.

There are many potential explanations for Republican strength in Miami-Dade. Gov. Rick Scott, running for senator, campaigned hard for Cuban voters, who play an outsize role in the county. The Republican candidate for lieutenant governor was Cuban-American as well.

There’s also a strong generational gap among Cuban voters that may have worked to the GOP’s advantage in a midterm election, when lower-turnout younger voters are more likely to stay home.

But there’s another possibility that should concern Democrats: the prospect that the president is more popular in Miami-Dade County than one might guess.

Upshot/Siena polls of Florida’s 26th and 27th districts showed President Donald Trump’s net approval rating was about 10 points better than his margin in the 2016 election. The polls nailed the results of the two congressional races there.

In 2016, Trump’s weakness in Miami-Dade was a big reason the state was so close, despite his gains among white working-class voters in the state. If Democrats can’t count on a big year in 2020 from voters in the county, Florida might slip down the list of battleground states.

Miami-Dade wasn’t the only place where Democrats were weak in predominantly nonwhite areas. In particular, they struggled to match Clinton’s already tepid margins in rural nonwhite communities.

In the Georgia governor’s race, Stacey Abrams often ran behind Clinton and the previous Democratic candidate for governor, Jason Carter, across much of the state’s rural and sometimes mostly black areas.

This could have been a result of Republican strength among white voters in the region: In Hancock County, the state’s most Democratic, predominantly black rural county, Republican nominee Brian Kemp won more raw votes than Trump did, 870 to 843. Abrams fell short of matching Clinton, 2,645 to 2,701.

In the Texas Senate race, Beto O’Rourke also struggled to match Clinton’s performance across largely Hispanic South Texas. He ran behind Clinton in Hidalgo County, which includes McAllen, Texas, and in Webb County, which includes Laredo.

The Arizona Democratic Senate candidate, Kyrsten Sinema, ran well behind Clinton’s performance in two additional heavily Hispanic border counties, Yuma and Santa Cruz.

The county-level data isn’t granular enough for us to be sure about whether strong white turnout, weak nonwhite turnout or nonwhite support of the Republican candidates were responsible for Democratic weakness in these areas. But the answer could make a big difference in 2020.

Democrats had a solid night in the Midwest, winning Senate races in Ohio, Michigan, Wisconsin and Minnesota, and flipping the governor’s mansions in Wisconsin, Michigan and Kansas.

The statewide results give the impression of a Democratic resurgence in the region. But they won without surging back where Trump made big gains in 2016.

Democrats struggled, to a surprising extent given the national environment, in the old mining and industrial towns where the party used to dominate. Instead, they won with big margins in the suburbs and with a healthier performance in rural areas and agricultural communities where the Democrats weren’t so dominant in the 20th century.

Although Democrats proved they could win back Midwestern states without their old strength in industrial-era Democratic bastions, this is a tougher route to victory in states where white voters without a degree represent an above average share of voters.

Republican John James won Monroe County, Michigan, in the industrial suburbs south of Detroit, by 11 points over the incumbent Democratic senator, Debbie Stabenow. Barack Obama won the county twice.

Overall, Stabenow’s 6.5-point win was smaller than the Democratic margin in the national vote. The Democratic showing in Florida and Arizona was also weaker than was the party’s overall national performance.

To win the presidency, Democrats will probably need at least one of Florida, Arizona or Michigan, or else they’ll most likely need to win a state where they lost more decisively in 2016 — like North Carolina, Georgia or Texas. Democrats fell short, or seemed on track to fall short, in prominent races in those three states last week. Democratic weakness in old industrial towns was a pattern outside Michigan, too.

Sherrod Brown, though successfully defending his Senate seat in Ohio, couldn’t win Sandusky County, where Obama won twice. In the Ohio governor’s race, Democrat Richard Cordray won Mahoning County (Youngstown) by only 12 points and lost the election. By comparison, other than Clinton, all Democratic presidential candidates since 1980 have done better in the county.

Democrats lost the open race in Minnesota’s 8th District, one of the most reliably Democratic areas of the 20th century. And on the West Coast, Sen. Maria Cantwell, a Democrat, couldn’t win coastal, blue-collar Grays Harbor, Washington, which voted Democratic in every presidential election from 1932 until 2016.

The pattern, however, did have some variations. Democrats had some decent results in Wisconsin and Pennsylvania. Unlike in Ohio and Michigan, they combined their newfound strength in well-educated suburbs with more of their traditional voters. Democrats routinely outperformed Clinton and Obama in traditionally Democratic parts of western Pennsylvania.

The overall picture is fairly clear: The states that voted for Obama and switched to Trump are still winnable for Democrats, but the party will find it hard to return to its prior levels in the communities that swung most to the current president.