It is undeniable that Tuesday night was a good one—perhaps even a great one—for Republicans. In contrast to 2012, when polls showed a bias against Democrats and Democrats ended up winning several more Senate races than models had predicted, 2014 polling showed a bias toward Democrats. Though several states, most notably Colorado, have trended sharply toward Democrats in post-Election Night counting and now resemble the polling consensus, almost all of Tuesday night’s surprises were in favor of Republicans. So, let’s take a look at where we were wrong and what it all means.
Senate
In the Senate, we made three incorrect predictions (four in the unlikely but still possible event that Sen. Mark Begich (D) pulls a comeback with absentee ballots in Alaska). In North Carolina, it was fairly straightforward: we thought Sen. Kay Hagan a slight favorite to win by a small margin—perhaps one or two points—and the national environment dragged her to a 1.7pt defeat at the hands of Republican Thom Tillis.
Our two more substantive errors were in the Democrats’ offensive opportunities: we thought that Democrat Michelle Nunn would hold Republican David Perdue below 50% and force a runoff in Georgia: Mr. Perdue ended up winning 53%. Part of this mistake dealt with Libertarian Amanda Swafford, who we thought would draw 3-4% of the vote but fell below 2%. This was part of a broader underperformance by Democrats in red states. We flipped a coin to give Independent Greg Orman the edge in Kansas, but Republican Sen. Pat Roberts pulled a ten-point victory, more than double his margin in any public poll. And though we had largely foreclosed on the possibility that Democrats would win in Arkansas or Kentucky, Republicans’ margins in these states were very impressive: Sen. Mark Pryor (D-AR) lost by seventeen points, and incoming Majority Leader Mitch McConnell defeated Alison Lundergan Grimes by 15pt.
Outside red territory, there were two major underperformances by Democrats: Iowa Senate candidate Bruce Braley lost by 8pt in a perennial swing state—probably a testament to his lackluster campaign and lack of personal appeal—while the biggest shock of the night was the near-defeat of Sen. Mark Warner (D-VA), whose reelection was seen as a given but won by less than 1pt in the Old Dominion.
But if there was a bright spot for Democrats, it was their solidification of light blue states: senators in Oregon, New Mexico, Minnesota, and Michigan all won reelection by double digits. Especially in Michigan, which was polling competitively over the summer but saw Rep. Gary Peters (D) cruise to a 13pt victory in the seat vacated by retiring Sen. Carl Levin, and in Minnesota, where Sen. Al Franken won by only 400 votes in 2008 but won by a solid 10pt this election, Democrats should be pleased.
Governors
Governors’ races were definitely the bleakest part of the night for Democrats; unlike the Senate, which most Democrats anticipated losing, Team Blue had high hopes for flipping statehouses in Florida, Kansas, and Maine, plus a variety of peripheral targets including Wisconsin and Michigan. But the night was instead grim: Democrats only picked up Pennsylvania—a long foregone conclusion—while shedding statehouses in Massachusetts, Illinois, Arkansas, and, shockingly, Maryland. The open seat in Arkansas was lost before Election Day, but Asa Hutchinson’s margin of victory (13pt) is an indicator that Arkansas is entirely lost to the Democrats who have governed it since Reconstruction.
We called the victory of moderate health care executive Charlie Baker in Massachusetts—the Bay State has a history of electing such pragmatic Republicans, and we do not see it as indicative of a Republican wave. And while Gov. Dannel Malloy (D-CT) held on in his second bout with private equity mogul Tom Foley, even more unpopular Gov. Pat Quinn (D-IL) lost a similar race to Bruce Rauner. While we picked Mr. Quinn to narrowly prevail, we are not particularly surprised by Mr. Rauner’s victory. Similarly, we are not shocked that Gov. Rick Scott (R-FL), whose race we called “Leans Democratic” defeated Charlie Crist by 1pt. Only somewhat more surprising was the victory of Gov. Sam Brownback (R-KS), who won by 4pt despite a backlash to his ultra-conservative agenda; but considering his 31pt victory in 2010, Mr. Brownback was only saved by a combination of his ruby-red state and a nationwide Republican surge. All of these results are perfectly consistent with a normal six-year-itch election, giving a modest bump to Republican candidates across the country.
But there were three races that legitimately took us by surprise: Maryland, Maine, and Vermont. In Maryland, which we rated “Likely Democratic,” Republican Larry Hogan won a solid five-point victory over Democratic Lt. Gov. Anthony Brown. We can be partly forgiven for missing the results—public pollsters pulled out of the state in early October, missing Mr. Hogan’s late surge—but the result was a shockingly harsh referendum on the tenure of Democratic governor and long-shot presidential aspirant Martin O’Malley. Unlike Massachusetts, Maryland has no real tradition of electing Republican governors to check a liberal legislature, and the race can only be viewed as a repudiation of state tax policy—including the “stormwater remediation fee,” dubbed the “rain tax” by its critics. In Vermont, where Democratic Gov. Peter Shumlin’s wasn’t even on the national radar screen—he won his last reelection campaign by 20pt—underfunded Republican businessman Scott Milne came just 1pt from defeating the governor. And in Maine, bombastic and far-right governor Paul LePage, who was elected with a bare plurality of 38% in 2010, won with 48% despite left-leaning independent Eliot Cutler falling to just 8% in the polls. If you told us that Mr. Cutler would only get 8%, we would have predicted a solid victory for Democratic Rep. Mike Michaud. And since Mr. LePage hardly fits the pragmatic profile of Governors-elect Baker and Hogan, his win in a blue state is all the more shocking.
House
Though five races, all currently held by Democrats, remain to be called (AZ-02, CA-07, CA-16, CA-26, and NY-25) a shift of a few thousand votes in any of those races will hardly change our outlook on the House, which is threefold. Firstly, the last deep-red districts held by Democrats are gone, and there is no longer any asymmetrical ability for Democrats to hold territory that is hostile at the presidential level. Secondly, a small number of strongly Democratic districts (most notably IA-01, IL-10, NV-04, and NY-24) fell to Republicans, some by large margins. But ultimately, with these caveats, the night was good for Republicans, but not great. Rather, it is the combination of three advantages—incumbency, gerrymandering, and spatial efficiency—that conspired to turn a good night into the largest Republican majority since World War II. In addition, the large number of conservative southern Democrats—now extinct with the defeat of Rep. John Barrow (D-GA) and retirement of Rep. Mike McIntyre (D-NC)—means that while this is the largest Republican majority for some time, there have been several conservative majorities of larger size since then. For instance, the 1946 election that gave Republicans 246 seats—the number they currently hold, pending any further wins in the five remaining races—did it with only two Republicans in the 105 seats residing in the states of the former Confederacy. In Texas, Democrats won all twenty-one seats, in fourteen of which they were unopposed and in only three of which did the Republican exceed 10% of the vote.
There are too many districts that were lost or only narrowly won by President Obama where not-particularly-entrenched Democrats won: Ann Kirkpatrick (AZ-01), Kyrsten Sinema (AZ-09), Raul Ruiz (CA-36), Scott Peters (CA-52), Gwen Graham (FL-02), Patrick Murphy (FL-18), Rick Nolan (MN-08), and Sean Patrick Maloney (NY-18) among them. A good example is NH-01, where three elections in a row have been contested between Democrat Carol Shea-Porter and Republican Frank Guinta. In 2010, Mr. Guinta defeated Ms. Shea Porter, then the incumbent, by 12pt. In 2012, Ms. Shea-Porter returned and unseated Mr. Guinta by 4pt. And this year, Mr. Guinta retook the seat by 4pt. While redistricting makes it difficult to make these comparisons to 2010 across the board, we would argue that the combination of redistricting and incumbency effects makes the Republicans’ 242-seat win in 2010 markedly more impressive than the 246-250 seats they will end up with this year. And while we will not have a good figure for the House popular vote nationwide for several weeks as late ballots get counted across the country, it looks like the GOP will have a six-point edge, down from the seven-point edge Republicans took in 2010 without fifty-plus additional incumbents. This, again, plays into the idea that this was a good year for Republicans—but hardly a wave.
So what explains the wins for Republicans in NV-04, IA-01, and NY-24, all of which President Obama won by double digits in 2012? Some parts of the country certainly saw a wave: the final Democratic bastion in the South, Arkansas, was thoroughly overrun by Republicans. Nevada, led by popular Gov. Brian Sandoval (R), saw Republicans sweep into every major office and the state legislature, while Democrats underperformed throughout New York. But Democrats might actually gain House seats in California, held onto control of the Kentucky House despite Majority Leader McConnell’s romp on top of the ticket, retained the Iowa Senate despite wins by Gov. Terry Branstad (R) and Senator-elect Joni Ernst (R). While 2010 was an even wave for Republicans, washing out Democrats in right-leaning districts quite uniformly, 2014 was more schizophrenic. The pieces were in place for a Republican wave across the country, and it certainly took hold in places where the stage was set correctly: depressed turnout in Nevada due to the uncompetitive governor’s race, for instance, or vicious in-fighting in New York between Gov. Andrew Cuomo (D) and his left flank, the Working Families Party. But in places like California, Florida, and the Great Lakes, the Republican wave looked much more like a ripple.
Ballot Measures
When liberals went looking on Election Night for a ray of hope, ballot measures were there. Every state that had a rise in the minimum wage on the ballot—a quite conservative group of Alaska, Arkansas, Illinois, Nebraska, and South Dakota—passed it easily. Referendums in Oregon and Alaska to follow Washington and Colorado into the world of legalized recreational marijuana succeeded, as did one in the District of Columbia (a medical marijuana referendum in Florida received 58% of the vote, but failed due to a super-majority requirement). Personhood amendments failed in North Dakota and Colorado. Washington passed a universal background check referendum by a large majority. Alaskans put the kibosh on a controversial mining project at Pebble Bay. In all, perhaps the most prominent referendum that didn’t break leftward was a measure in Tennessee that sought to toughen abortion restrictions: it passed with just 52%. And finally, in the most important referendum of the night, Louisianans rejected a proposition to add members to the state’s Wildlife and Fisheries Commission.
This piece marks the end of Midterm Countdown. Thanks to all who have been reading, and we’ll be back soon! After all, the cycle never ends.