2014: The Senate Map Was the Strategy

Republicans gained nine United States Senate seats on 11/04/2014, finishing with 54 seats against 44 Democrats and two independents, and took the Senate majority for the first time since the 2006 election. The Senate Historical Office records the chamber moving from 53 Democrats, 45 Republicans and two independents to 44, 54 and two. In the same election Republicans reached 247 seats in the United States House of Representatives, their largest House majority since the 71st Congress of 1929 to 1931. The House Historian’s party division table records both figures.

The standard account credits the map. The evidence says the map was necessary and was not sufficient. Senate terms run six years, so the seats contested in 2014 had last been contested in 2008, and that exposure was knowable in 2009. Republicans had comparable structural openings in 2010 and converted fewer of them, because five winnable races went to nominees who came through contested primaries and then lost. What changed by 2014 was the selection process rather than the arithmetic of the calendar.

That position is falsifiable. If the party had been converting its structural opportunities at a steady rate across cycles, the map would carry the explanation on its own. It was not. Republicans gained six seats in 2010 without a majority and finished 2012 two seats worse off, with the same kind of map in front of them.

What the 2014 Senate map looked like before anyone campaigned

Thirty-six United States Senate seats were on the 2014 ballot, 21 held by Democrats and 15 by Republicans, and seven of the Democratic-held seats sat in states Mitt Romney had carried in 2012. Those counts are set out in a 2014 Senate overview. Only one Republican-held seat was in a state Barack Obama had carried. The 2014 Senate elections record notes the comparison. Republicans needed a net gain of six seats for control. The same overview states the threshold

The 2014 Senate map before anyone campaigned

Which party held the 36 seats on the 2014 ballot, and how many sat in states the other party’s presidential nominee had carried in 2012.

A map is mostly an inheritance. The seven exposed Democratic seats were in Alaska, Arkansas, Louisiana, Montana, North Carolina, South Dakota and West Virginia, and no decision made in 2013 or 2014 put them there. What a party committee actually chooses is narrower: which of the inherited opportunities to treat as must-win, how early to commit, and whether to spend into the two or three seats outside the obvious set that would provide margin for error.

Senate seats before and after the 2014 election

Party control of the United States Senate at the end of the 113th Congress and after the elections of 11/04/2014.

Republicans made that second choice and it is the difference between six seats and nine. Colorado and Iowa were both states Obama had carried twice, and both were won, by Cory Gardner by 1.95 points and Joni Ernst by 8.3. A party that had confined itself to the seven Romney-state seats would have taken the majority and had no cushion in it.

Candidate discipline after two cycles of nominee problems

Republicans lost five Senate races in 2010 and 2012 that the party had expected to contest, each with a nominee who had emerged from a contested primary. In 2010 Harry Reid held Nevada against Sharron Angle by 5.7 points, Chris Coons won Delaware against Christine O’Donnell by 16.6, and Michael Bennet held Colorado against Ken Buck by 1.8. The 2010 margins are on the record. In 2012 Claire McCaskill beat Todd Akin in Missouri by 15.8 points and Joe Donnelly beat Richard Mourdock in Indiana by 5.8. The 2012 margins are on the record. Republicans gained six Senate seats in 2010 without reaching a majority and finished the 2012 cycle two seats worse off than they began it.

Five Senate races Republicans lost in 2010 and 2012

Democratic winning margin in five seats Republicans had expected to contest, where the Republican nominee emerged from a contested primary.

Set those five races against the cycles they sat in and the map stops carrying the argument on its own. Republicans gained a net of six Senate seats in 2010 and a net of nine in 2014. Adding back Nevada, Colorado and Delaware would have made 2010 a nine-seat cycle, and adding back Missouri and Indiana would have made 2012 a break-even one. That is arithmetic rather than a counterfactual, and it is enough to show where the variance lived.

Republican Senate seat change by cycle, with the nominee losses added back

Net change in Republican US Senate seats in each of three cycles, next to an arithmetic illustration that adds the seats lost with nominees who emerged from contested primaries.

The response was a program rather than a slogan. Time reported that the National Republican Senatorial Committee, chaired by Senator Jerry Moran with Senator Rob Portman as finance vice chairman, Rob Collins as executive director and Ward Baker as political director, began running three-day media sessions in March 2013, before most candidates had announced. Time documented the program on 11/04/2014. The committee spent about 250,000 dollars in 2013 researching both Democratic and Republican candidates, down to flags on candidates’ own social media accounts. Media training was made mandatory, for the first time, for candidates who wanted the committee’s financial support. On 10/01/2013 trackers met 16 potential Senate candidates at Reagan National Airport in Washington and the committee played the footage of their reactions back to them. Some candidates went through training sessions 15 to 20 times.

The counterfactual is not clean and the honest version says so. The political scientist John Sides observed that no 2014 gaffe was as large as Akin’s and that the polls showed no comparable swing after any 2014 remark, so there is no way to measure how many seats the program saved. Sides made that point in the same Time account. What is observable is the process and the absence of a race-ending episode in nine pickups, three of which were decided by under two and a half points.

How the party fixed its recruitment process

Eight of the nine Republicans who took Democratic-held Senate seats in 2014 had already won elective office. Tom Cotton in Arkansas, Bill Cassidy in Louisiana, Steve Daines in Montana, Shelley Moore Capito in West Virginia and Cory Gardner in Colorado were sitting members of the United States House. Mike Rounds in South Dakota had served two terms as governor. Thom Tillis was Speaker of the North Carolina House. Joni Ernst was an Iowa state senator. Dan Sullivan in Alaska had served as the state’s natural resources commissioner.

Margins and results from the 2014 US Senate election returns; the 2012 presidential winner in each state follows Democracy in Action’s 2014 Senate overview. Louisiana was decided in a runoff on 12/06/2014.

State	Republican winner	Office held before the election	2012 presidential winner
Alaska	Dan Sullivan	State natural resources commissioner	Romney
Arkansas	Tom Cotton	US Representative, 4th district	Romney
Colorado	Cory Gardner	US Representative, 4th district	Obama
Iowa	Joni Ernst	Iowa state senator	Obama
Louisiana	Bill Cassidy	US Representative, 6th district	Romney
Montana	Steve Daines	US Representative, at large	Romney
North Carolina	Thom Tillis	Speaker, North Carolina House	Romney
South Dakota	Mike Rounds	Governor of South Dakota, 2003 to 2011	Romney
West Virginia	Shelley Moore Capito	US Representative, 2nd district	Romney

The nine Senate seats Republicans gained in 2014

Winning margin in each Democratic-held seat that changed hands, including the Louisiana runoff of 12/06/2014.

Prior office is a crude instrument and it is a useful one, because it carries two things a committee can otherwise only guess at. The first is a vetting record: a candidate who has already run a contested race has been through opposition research once, and the material that would have ended a campaign has usually surfaced. The second is demonstrated performance with a real electorate rather than a projected one. Neither guarantees a win. Both narrow the distribution of outcomes, which is the entire job of a party committee operating across nine states at once.

The cost is real and the 2014 committee paid it openly. Making training a condition of funding meant the committee was choosing among candidates before voters did. In 2013 the committee told the Alaska Republican Joe Miller that his negatives were too high to win a general election, the conservative writer Erick Erickson answered that conservatives should stop giving incumbents free rides, and the Republican strategist Rick Wilson said the committee was making itself look like an insider protection racket. BuzzFeed News reported the exchange in August 2013. Early money carried the same character: by 06/10/2014 the committee had booked 10.2 million dollars of advertising across Alaska, Arkansas, Colorado and North Carolina, and three of those four races were ultimately decided by under two and a half points. Roll Call listed the committee’s reservations on 06/10/2014. A party that manages nomination risk is trading some legitimacy with its own base for a narrower spread of general election outcomes. That is a defensible trade and it is still a trade.

What turnout did in 2014

Turnout in the 2014 midterm was 36.4 percent of the voting-eligible population, the lowest rate in a United States midterm since 1942, when it was 33.9 percent. The United States Elections Project figures were reported by PBS NewsHour. Both the favorable map and the low turnout pushed in the same direction, and no honest accounting separates them cleanly.

The distinction that matters operationally is about timing of knowledge. The map was fixed in 2008 and legible by 2009. The turnout rate was not knowable until the votes were counted. A party that organizes around the first and treats the second as a range it has to survive is planning where the information actually is.

What a campaign can take from the 2014 midterm

Three things carry over to any multi-race program, whether it is a Senate map or a slate of County Commission seats. Start from the inherited structure, because exposure is usually knowable years ahead and it sets the ceiling on what tactics can achieve. Treat nomination risk as a measurable quantity with a budget attached, because a nominee who cannot survive four weeks of scrutiny is the single most expensive thing that can happen to a winnable seat.

Then spend past the obvious set. Republicans in 2014 needed six and contested enough to win nine, which is the difference between a majority that depends on every close race and one that can absorb a loss.

Frequently asked questions

How many Senate seats did Republicans gain in the 2014 midterm?

Republicans gained a net of nine United States Senate seats on 11/04/2014, moving the chamber from 53 Democrats, 45 Republicans and two independents to 44 Democrats, 54 Republicans and two independents. They needed a net gain of six for control, so the final margin was three seats wider than the minimum.

Why was the 2014 Senate map favorable to Republicans?

Thirty-six Senate seats were on the 2014 ballot, 21 of them held by Democrats and 15 by Republicans, and seven Democratic-held seats were in states Mitt Romney carried in 2012 while only one Republican-held seat was in a state Barack Obama carried. Those seats had last been contested in 2008, a strong Democratic year, so the exposure was set six years in advance.

What did Republicans change about Senate recruitment after 2012?

The National Republican Senatorial Committee made media training a condition of its financial support, began three-day media sessions in March 2013 before most candidates had announced, and spent about 250,000 dollars in 2013 researching both parties' candidates. On 10/01/2013 it had trackers meet 16 potential candidates at Reagan National Airport and played the footage back to them.

What was voter turnout in the 2014 midterm election?

Turnout in the 2014 United States midterm was 36.4 percent of the voting-eligible population, according to the United States Elections Project. That was the lowest midterm turnout since 1942, when the rate was 33.9 percent.