From Lost Senate Races to a Majority, What 2014 Learned From 2010 and 2012

There is a particular kind of political grief that afflicts a party which loses the race it was supposed to win. It is worse than losing a hard race. A hard loss is a fact. A loss you could have avoided is a seminar, and the seminar lasts two years.

Republicans attended two such seminars in a row. In 2010 and again in 2012, the conditions favored them in the Senate, and they came up short both times. Then in 2014 they gained nine seats and took the majority. Let me tell you what the class learned, because the lessons are better than the usual ones, and because they have nothing to do with luck.

Two elections that should have been better

Start with the setup in 2010. The president's party was in the middle of a rough midterm, as the president's party usually is. Republicans gained six Senate seats, which is a serious number, and still finished in the minority. The Senate then stood at 53 seats for the Democratic caucus and 47 for the Republicans, and the majority leader of the Senate, Harry Reid of Nevada, had himself been regarded as vulnerable.

He won. In Nevada, the Republican nominee, Sharron Angle, ran a spirited primary campaign and a general election in which her own statements gave the Democrats their material. In Delaware, Representative Michael Castle, a former governor and a nine-term congressman, was widely seen as the strongest Republican candidate for an open seat; he lost the Republican primary to Christine O'Donnell, and the general election went to the Democrat, Chris Coons, by a wide margin. In Colorado, the Republican nominee, Ken Buck, lost narrowly to the appointed incumbent Michael Bennet. I am not criticizing any private citizen. These were public candidates who made public choices, and the point is a strategic one: three seats that were reachable in a favorable year went the other way.

Then came 2012, and here the map should have made things easy. Democrats were defending 23 of the 33 seats on the ballot, and Republicans only 10. A party with that kind of map expects to gain. Republicans lost two seats on balance and went from 47 to 45.

Look at the states. Mitt Romney carried Missouri, Indiana, Montana and North Dakota in the presidential race that year, and the Republican Senate candidate lost in all four. Here is the plain version: a party can carry a state at the top of the ticket and lose the Senate seat in that same state, and that is almost never an accident of the electorate. It is almost always the candidate.

In Missouri, the Republican nominee, Representative Todd Akin, made remarks in a television interview in August 2012 on a painful subject, and the remarks dominated the rest of the campaign. National party leaders asked him to withdraw, and he declined. Senator Claire McCaskill, regarded as one of the most vulnerable Democrats in the country, won re-election. In Indiana, a long-serving Republican senator, Richard Lugar, lost his primary to the state treasurer, Richard Mourdock, and then Mourdock lost the general election after a debate answer on the same subject that overshadowed the final weeks. The Democrat, Joe Donnelly, won a seat that most analysts had considered safe.

The coverage of those races tended to treat the candidates' remarks as freak events, bolts from the blue. I would argue they were something more useful for the party to understand: the consequences of a nomination process that did not test candidates for general-election readiness.

The Constitution's design for the Senate

Before we go on, a note on why the Senate works differently from the House, because it explains why candidate quality matters so much here.

Article I, Section 3 gave the Senate six-year terms and arranged them in three classes, so that one-third of the seats come up every two years. The Founders designed the Senate to be the steadier body: in Federalist 62 and 63 they argued that a smaller chamber with longer terms would resist the passions of the moment. The 17th Amendment, ratified in 1913, changed how senators are chosen. It provides: "The Senate of the United States shall be composed of two Senators from each State, elected by the people thereof, for six years; and each Senator shall have one vote."

Notice two consequences. First, because only a third of the seats are on the ballot, every cycle is a different map, and a party's fortunes can depend on which states happen to be up. Second, because the people elect senators directly, the primary is an election that decides which of a party's candidates the people of the state will see in November. A party that has a nominating process that does not produce candidates who can win statewide has, in effect, handed the general election to the opponent in the spring.

2014: what was done differently

The 2014 map was favorable. Democrats were defending 21 of the 36 seats up, many of them in states Romney had carried in 2012, among them Arkansas, Louisiana, Montana, South Dakota and West Virginia. But the map had been favorable in 2012 too, and the party had lost ground. So the question is what else changed.

I would put the answer in four numbered points.

1. The party recruited experienced candidates early. In state after state, the Republican nominee was someone who had already won election to something. Cory Gardner in Colorado, Tom Cotton in Arkansas, Shelley Moore Capito in West Virginia, Steve Daines in Montana and Bill Cassidy in Louisiana were sitting members of the House. Mike Rounds in South Dakota was a former governor, Thom Tillis in North Carolina was the speaker of the state's House of Representatives, and Dan Sullivan in Alaska was a former state natural resources commissioner. Joni Ernst in Iowa was a state senator and a veteran of the Iowa National Guard. A candidate who has won before has been vetted by voters, has a record that can be examined in advance, and has learned how to handle a hostile interview.
2. The primaries were run with the general election in mind. The national party committee and allied groups worked to ensure that the strongest candidate was the nominee, and the party's incumbents took their primary challenges seriously. Senate Republican leader Mitch McConnell won his May primary in Kentucky and went into the fall with a united party behind him.
3. The message was disciplined. Republican candidates ran on the economy, on the troubled launch of the health-care law's website in October 2013 and what it said about the government's competence, on spending, and on the argument that the Senate majority had been a barrier to bills the House had passed. They did not have to explain things that were not on the ballot.
4. The candidates stayed out of the story. The 2014 Republican nominees did not generate the kind of headline that had swallowed their predecessors. That was not luck. It was preparation, and some of it was media training, which is a polite name for teaching a candidate how to answer the question that has been asked rather than the one he wishes had been asked.

The result

On election night in 2014, Republicans won Arkansas, Colorado, Iowa, Montana, North Carolina, South Dakota, West Virginia and Alaska, and in December they won the Louisiana runoff. That makes nine seats gained. They also held Georgia, Kentucky and Kansas, where the incumbent Republican senator, Pat Roberts, faced a serious independent challenge and survived. The new Senate stood at 54 Republicans and 46 in the Democratic caucus, giving the Republicans the majority for the first time since they lost it in the 2006 election.

Add that up. Nine seats gained, 45 plus 9 equals 54, and in a 100-member chamber, 51 is a majority. They cleared it with three seats to spare. That is a margin large enough to survive an illness or a retirement, which mattered more than anyone thought at the time.

I would also note, with some satisfaction, what the coverage had said earlier in the cycle. A good deal of the commentary in 2013 argued that the party had a "problem" in Senate races because its primary voters were too ideological to nominate candidates who could win, and that it would again fall short. The coverage was not entirely wrong, since that was what had happened twice. What it missed was that a party can learn. A party that has gone through the seminar twice may come back with the notes.

The strongest objection

Now for the point-by-point section. The best argument against my reading is this: 2014 was a sixth-year midterm for a president whose party had held the White House for six years, and voters tend to punish the president's party in such elections. The map was lopsidedly favorable. Any Republican with a pulse would have won most of those races, and the story about candidate quality and primary discipline is a flattering narrative imposed after the fact.

It is a good objection, and it is half right. The president's party does tend to lose seats in midterms, and a sixth-year election is often the worst for the party in the White House. I take that up in Why the President's Party Almost Always Loses the Midterm, and the Exceptions. And the map mattered: if Democrats had been defending 10 seats, the Republicans would not have gained nine.

But look at what the objection cannot explain. It cannot explain 2012, when the map was even better. It cannot explain 2010, which was itself a favorable midterm. In both years, the favorable conditions were there and the Republicans did not win the Senate. If favorable conditions were sufficient, the party would have had a majority by 2012, and it did not. So conditions were necessary, and the candidates were what made them sufficient.

The proof lies in the contrast. In 2014 the party held every seat it was defending, including a scare in Kansas, and it won in states where a weak candidate would have thrown the seat away. A pulse was not enough. In 2012 the party had a pulse in Missouri and Indiana, two states it carried for president, and lost both Senate races. The map did not change between the candidates. The candidates did.

The lesson: comebacks are won a year early

If there is one sentence to take from the story, it is that a comeback is won in the primaries a year before November. By the time the general election begins, the party has already made most of its decisions: who the candidate is, what the candidate has said, and how the candidate has been tested. The general election is the audit.

Here is the practical checklist for a party, or for any citizen who cares about the outcome.

1. Start recruiting two years out. The best candidates are people who have already won something, and they need time to decide to run.
2. Take primaries seriously. The primary is the election that matters most to the party's fortunes, because it determines which candidate the general electorate sees.
3. Test candidates on the hard questions before the voters do. If a candidate cannot answer the question that will be asked in October, it is better to find out in March.
4. Pick the argument in advance and stay on it. Voters can hold three ideas in their heads at a time, and the campaign that gives them one wins the other two.
5. Respect the voters of the primary and the voters of the general. Primary voters choose the candidate; general election voters choose the senator. The nominee has to be acceptable to both.

None of that requires abandoning principle. A party can nominate principled candidates who are also prepared, and the 2014 class is proof. Principle without preparation is a hobby. Preparation without principle is a job. A party that wants to win and to govern needs both.

What the majority did next, and what it teaches

A majority is not an end. The Senate that began in January 2015 faced the realities every majority faces: a president with a veto pen, a filibuster that required 60 votes for most legislation, and the sort of procedural machinery explained in The Filibuster After the Midterm, Why Sixty Votes Now Protect the Outnumbered. It also controlled the nominations process, where a single senator's hold can slow a nominee, a power laid out in The Senate Hold Explained, How One Senator Can Slow a Nominee for Months. The Republican majority passed bills the president vetoed, including legislation approving the Keystone XL pipeline, which he vetoed in February 2015.

That is a lesson in itself. A majority wins arguments and seats, and then it learns what it can and cannot do under divided government. The comparison with other comebacks is in Five Conservative Comebacks Side by Side, and the Ingredient They All Shared, and the wider pattern of what the minority can control and cannot is in Comebacks Start the Morning After, Lessons From the Defeats of 1964, 1974, 2008. The numbers that decide control of the Senate are in The Numbers That Run Congress, From 218 to 60 to 290 to 38 and Back.

And since single-issue voters often decide primaries, a party has to understand the coalition it is asking to support its nominee; the piece called Why Single-Issue Voters Hold a Coalition Together, and When They Break It takes that up.

The case in brief

1. Republicans fell short of the Senate in 2010 and 2012 with favorable conditions, including a 2012 map on which Democrats defended 23 seats to the Republicans' 10, because several nominees lost races that were winnable.
2. In 2012 Mitt Romney carried Missouri, Indiana, Montana and North Dakota, and the Republican Senate candidate lost all four, which points to candidate quality and not to the electorate.
3. In 2014 the party recruited experienced candidates, ran primaries with the general election in mind and kept its message disciplined, and it gained nine seats for a 54-46 majority.
4. Favorable conditions were necessary and not sufficient, because the conditions were also favorable in 2012 and the party did not win.
5. A comeback is won in the primaries a year before November, and the general election is the audit of decisions already made.

What you can do: find the date of your state's next Senate primary on your secretary of state's website, mark it on your calendar, and read the candidates' records before you vote in it.