It will take weeks to fully digest the surprising results of last week’s election. This is true at all levels of the ballot, where Democrats vastly underperformed. And not just in the presidential outcome: Democrats were held to a stunningly low two-seat pick up in US Senate races, and the mildly disappointing result in US House contests, where Democrats scored a net gain of six or seven seats—depending upon one uncalled contest in California.
Classifying the House results as mildly disappointing is based on mid-October signals. Available information suggested that Donald Trump was building momentum in non-suburban districts, which were an essential component of a national Democratic strategy to gain up to 15 seats. The final NCEC prediction projected a 10 to 12 net gain in the House, far short of our post-Labor Day assessment that Democrats might gain 20 or more House seats.
Despite what appeared to be an impenetrable electoral college advantage in favor of the Democrats, recent elections have shown that its reliability was overstated. For the second time in the last five national elections, the Democratic presidential nominee won the popular vote and lost the electoral college. When all votes are tallied, Hillary Clinton will likely win the popular vote by more than 1 million votes, almost doubling Al Gore’s 2000 plurality over George W. Bush.
In the 14 swing states (Arizona, Colorado, Florida, Georgia, Iowa, Maine, Michigan, Minnesota, Nevada, New Hampshire, North Carolina, Ohio, Pennsylvania and Wisconsin) Trump won 142 electoral votes compared to 32 for Clinton. Contrary to the national result, Trump won the popular vote by slightly more than 2 points in these 14 states.
Conversely, Clinton will likely win the popular vote by 2.5 points when the vote is counted in the remaining 36 states and the District of Columbia. In California, New York and Illinois alone, Clinton’s popular vote margin will exceed 5 million votes. Comparatively, Trump won Texas by about 810,000 votes, and no other state by more than 600,000. If there is a clarion call for the end of the Electoral College, these statistics make a strong case.
Trump’s triumph in the eastern-to-midwestern corridor from Pennsylvania to Indiana, including Michigan and Wisconsin, was fueled by two elements that we will examine in a detailed analysis later in the week: 1) the shockingly low Democratic turnout in urban centers like Detroit and Milwaukee, and 2) a larger than expected shift away from the Democratic Party in middle income, formerly industrial counties across several states. In Pennsylvania this spanned from counties such as Luzerne, Lackawanna and Berks, all of which surround Philadelphia, to Beaver County in the western part of the state. The shift continued in Trumbull, Mahoning and Stark Counties in Ohio, Macomb County in Michigan (home of the “Reagan Democrats”) and Kenosha County, south of Milwaukee in Wisconsin.
Until the last week of the campaign, Democrats were favored to gain four or more seats in the US Senate, bolstered by what we assumed would be strong coattails from Hillary Clinton. Had she carried Pennsylvania and Wisconsin—two states holding crucial US Senate contests—we could have tested this hypothesis. However, when looking at the numbers, it’s conceivable that even if Clinton had won those traditionally Democratic states, the Democrats still would have lost the Senate. Here is why:
The NCEC tracked ten competitive US Senate races: eight held by Republicans, plus Democratic Colorado and Nevada. Clinton, despite losing six of the ten states, outperformed the Senate candidates in eight of them! Had Democrats not unseated Republican Kelly Ayotte by a scant 0.1 percent of the vote in New Hampshire, the only Democratic gain would have been Illinois—a state that Clinton won by more than 900,000 votes.
The following chart, comparing Hillary Clinton’s margin to that of the Democratic senate candidate (sorted by the difference), illuminates this result:
|State||Incumbent Party||Clinton Margin||Dem Senate Margin||Difference|
In the House, Democrats faced the near-impossible task of gaining the 30 seats required for a majority, barring a complete implosion caused by Trump. The net gain of six or seven is disappointing but explainable.The difference between the national Democratic margin for President and US House was under 3 percentage points in both 2012 and 2016.
|Year||Presidential Margin||House Vote Margin||Difference|
|2016||0.8% (Estimated)||3.1% (Estimated)||2.3% (Estimated)|
In 2016, Democrats captured either nine or ten Republican-held districts, while losing three Democratic-held districts. Overall, six or seven incumbent Republicans lost, compared to only one Democrat—Brad Ashford in Nebraska’s 2nd District.
Defeated Republican Incumbents
|District||Defeated Republican Incumbent|
|*CA-49 (San Diego County)||Darrell Issa|
|FL-07 (Orlando-Seminole County)||John Mica|
|FL-13 (St. Petersburg)||David Jolly|
|IL-10 (Chicago Suburbs)||Robert Dold|
|NH-01 (Manchester)||Frank Guinta|
|NV-04 (Las Vegas Suburbs)||Cresent Hardy|
|NJ-05 (New York City Suburbs)||Scott Garrett|
|* Still to be decided|
None of the defeated Republican incumbents represented rural districts, a trend that has held in five consecutive elections starting in 2008.
The Democratic strategy to amass significant gains required success in both rural/small town and suburban districts. But Democrats were only able to win seats in largely urban or suburban districts. Even among these districts there were missed opportunities, such as Pennsylvania’s 8th District, a suburban Philadelphia district with the demographics of a blue collar middle-income constituency (namely Bucks County). There was a total failure to gain traction in rural and small-town districts.
As the election neared, several rural or partially rural Republican districts appeared to be vulnerable, but ultimately went to the Republicans.
Missed Opportunities in Rural Districts
|District||Republican Margin of Victory|
The critical factor that shattered any prospect of double-digit Democratic gains was the disappointing failure in suburban districts. Examples are Colorado’s 6th District (Denver suburbs), Virginia’s 10th (Washington DC suburbs), Florida’s 18th (Palm Beach, Martin and St. Lucie Counties), Kansas’ 3rd (Kansas City suburbs), Minnesota’s 2nd and 3rd (Minneapolis suburbs and exurbs), New York’s 1st and 24th (Long Island and Syracuse).
Missed Opportunities in Suburban Districts
|District||Republican Margin of Victory|
|CO-06 (Denver Suburbs)||9.9%|
|FL-18 (Palm Beach, Martin and St. Lucie Counties)||10.6%|
|FL-26 (Miami Suburbs)||11.8%|
|KS-03 (Kansas City Suburbs)||10.7%|
|MN-02 (Minneapolis Suburbs)||1.8%|
|MN-03 (Minneapolis Suburbs)||13.8%|
|NY-01 (Long Island)||18%|
Another critical factor that hurt the Democrats was a repeat of low turnout among Hispanic voters in marginal California districts, particularly in California’s 10th and 21st Districts (Fresno and Central Valley areas). Similarly, low Hispanic turnout was seen in other marginal districts, such as Texas’ 23rd District, where Pete Gallego failed to regain a seat that he won in 2012. In Texas, Hispanic turnout was low in crucial border counties, and the natural Democratic advantage was offset by suburban San Antonio Republican voters.
A Light at the End of the Tunnel
The Republicans now control both the White House and the US Capitol. Apart from George W. Bush and Franklin Roosevelt, no incumbent president has presided over midterm election gains two years after assuming the presidency. Some recent examples: Democrats lost 63 seats in 2010 (Obama’s first term) and Republicans lost 26 seats in 1980 (Reagan’s first term).
Additional observations will become clear as we dig deeper into the numbers. Stay tuned.