2006: How a Wave Gets Built, Seat by Seat

Democrats gained 31 seats in the United States House of Representatives and six seats in the United States Senate on 11/07/2006, taking control of both chambers for the first time since the 1994 election. The seat totals are recorded by the Office of the Historian of the US House. Republicans lost 30 House seats on the same night and won no seat that a Democrat had held, and no Democratic House incumbent was defeated in the general election. Both of those facts are in the 2006 House returns.

The standard account credits the national environment. The evidence points at supply. A national swing converts into seats only where a credible candidate had already filed and had enough money to be heard, and the 2006 record shows the Democratic committee’s largest bets losing while two seats it had passed over were won. The decisive choices were about which districts had a funded campaign in March, and they were made before the environment was visible.

That position is falsifiable. If committee money had tracked the seats that changed hands, the environment account would be sufficient on its own. The available figures show it did not.

What Democrats gained in the 2006 midterm

Democrats finished the 2006 election with 233 US House seats to 202 for Republicans, up from 202 Democrats, 232 Republicans and one independent at the opening of the previous Congress. Those party divisions come from the House Historian. Democrats won the national House popular vote by eight points, 52.3 percent to 44.3 percent, and defeated 22 sitting Republican members. The certified national totals were 42,338,795 Democratic to 35,857,334 Republican.

House seats before and after the 2006 election

Party divisions at the opening of the 109th and 110th Congresses, as recorded by the Office of the Historian of the US House.

The Senate was the harder half. Democrats needed every one of the six Republican-held seats they contested seriously, and they got all six: Pennsylvania, Ohio, Rhode Island, Missouri, Montana and Virginia. The chamber finished at 49 Democrats, 49 Republicans and two independents who caucused with the Democrats, which produced a working majority of 51. The Senate Historical Office records that division for the 110th Congress. Two of the six were decided by less than a point. Jim Webb carried Virginia by 9,329 votes out of more than 2.3 million cast, and Jon Tester carried Montana by 0.9 points. The certified Virginia result was 1,175,606 to 1,166,277.

The six Senate seats Democrats took in 2006

Winning margin in each Republican-held US Senate seat that changed hands on 11/07/2006.

Sources: Office of the Historian, US House of Representatives; 2006 US House election returns; 2006 US Senate election returns.

Measure	Before 11/07/2006	After 11/07/2006
US House seats	232 R, 202 D, 1 I	202 R, 233 D
US Senate seats	55 R, 44 D, 1 I	49 R, 49 D, 2 I caucusing with Democrats
National House popular vote		52.3 percent D to 44.3 percent R
Republican House incumbents defeated		22
Democratic House incumbents defeated		0
Senate pickups decided by under one point		2 (Virginia, Montana)

How recruitment changed the 2006 map

The Democratic Congressional Campaign Committee supported 56 House campaigns through its Red to Blue program in the 2006 cycle, against 27 in 2004, and raised 22.6 million dollars for them, which was 15.1 million dollars more than the program had raised two years earlier. Roll Call reported the program’s totals for both cycles. The program also supplied strategic advice and staffing to the campaigns it took on, so the choice of who entered it was a choice about which districts got a professional operation rather than a volunteer one.

How many House challengers the Democratic committee funded

Candidates supported through the Democratic Congressional Campaign Committee’s Red to Blue program in two consecutive cycles.

Doubling the number of funded challengers is a different act from doubling a television budget. Money that arrives in March buys a campaign manager, a finance director and a year of voter contact. Money that arrives in October buys advertising against an opponent who has already defined you. The 2006 committee made more of the first kind of bet, in more places, earlier, which is the part of a wave that a party actually controls.

Candidate fit in conservative districts

Two of the clearest Democratic pickups of 2006 came from candidates whose personal profiles matched districts their party had not carried in a decade. Brad Ellsworth had been the elected sheriff of Vanderburgh County, Indiana, since 1999 and ran as a conservative Democrat; he beat the six-term Republican John Hostettler in Indiana’s 8th congressional district with 131,019 votes to 83,704, or 61 percent to 39, the largest defeat suffered by any House incumbent that year. The Indiana 8th district result and his tenure as sheriff are on the record. Heath Shuler, a former National Football League quarterback, ran on what he called mountain values and campaigned against abortion rights, same-sex marriage and gun control. He beat Charles Taylor, an eight-term incumbent and an Appropriations subcommittee chairman, with 124,972 votes to 107,342 while Taylor spent 4.4 million dollars including 2.5 million of his own against Shuler’s 1.8 million. North Carolina’s 11th district result, the spending figures and Shuler’s stated positions are documented.

Two conservative districts and the candidates who fit them

Results in Indiana’s 8th and North Carolina’s 11th congressional districts on 11/07/2006.

Recruitment of this kind is a measurement problem wearing a human face. The useful question about a potential candidate is narrow: does the distribution of opinion in this specific district have a shape that this specific person fits, on the two or three issues that voters there actually use to sort candidates? That question has an empirical answer and a campaign can get it before a filing deadline rather than after.

The approach carried a real cost and it is worth naming. The Blue Dog Coalition grew from 35 members in the Congress elected in 2004 to 50 in the Congress elected in 2006, and to 54 after 2008. The coalition’s membership counts and its role on the health care bill are documented. In 2009 Blue Dogs on the House Energy and Commerce Committee held up the health care bill until after the summer recess, and their opposition to a public insurance option is why that provision came out of the final law. A party that recruits for district fit is buying seats and selling cohesion, and the exchange rate is not always favorable.

Contesting seats the national ratings had written off

Several 2006 pickups came in districts that neither national party committee had treated as priorities, and the committee’s own biggest investments went elsewhere and lost. Nancy Boyda beat the five-term Republican Jim Ryun in Kansas’s 2nd congressional district with 114,139 votes to 106,329, or 50.6 percent to 47.1, in a district described afterward as low on both national parties’ political radars. The official Kansas canvass gives those totals; a widely mirrored encyclopedia entry reports 111,759 to 104,128 with the same percentages, and the official canvass is used here. In New Hampshire’s 1st district, Carol Shea-Porter beat the Republican incumbent Jeb Bradley with 100,691 votes to 95,527, having first won the primary 54 percent to 34 against a better-funded, party-backed opponent to whom Rahm Emanuel had personally contributed 5,000 dollars. The New Hampshire 1st district general and primary results are both recorded. The primary contribution was reported by Rick Perlstein.

Two seats the national map had passed over

Results in Kansas’s 2nd and New Hampshire’s 1st congressional districts on 11/07/2006.

Set those against where the money went. Rick Perlstein reported that the committee’s preferred candidates in Kentucky’s 2nd and 4th districts lost by nine and twelve points, and that about 3 million dollars went into Tammy Duckworth’s campaign in Illinois’s 6th district, which lost by four. Perlstein set out the spending and the outcomes in The New Republic. Independent expenditure records for that district filed with the Federal Election Commission and compiled by ProPublica show 1.7 million dollars of independent spending from all committees combined, with the Democratic committee’s own total reported inconsistently within the same page. The ProPublica compilation of those filings does not reconcile internally. The precise figure is uncertain and the direction is not: the committee’s largest commitment in that district lost, and two seats it had not prioritized were won.

Where the party committee’s judgment landed in 2006

Five House races and the Democratic committee’s posture toward each. Margins are the two-candidate result; the funding descriptions follow contemporaneous reporting.

That is an argument about error rather than about judgment. A target list is a forecast and a forecast has a distribution around it. In a year with a large national swing the error is asymmetric, because ratings built on the previous cycle’s results understate how far the tail moves, so the seats a list is most likely to miss are the ones on its edge. Fielding a credible, funded candidate in those districts is cheap relative to a late television buy and it is the only way to be wrong in the direction that still wins seats.

The counter-example belongs here too. Boyda declined the committee’s help in her 2008 re-election campaign, saying Kansas voters should control Kansas campaigns, and she lost the seat. Her 2008 decision and its aftermath are on the record. Contesting broadly is an argument about the party’s list. It is not an argument that committee money does nothing.

What a campaign can take from the 2006 midterm

Three decisions in 2006 are repeatable in any race, from County Commission to United States Senate. The first is to settle the candidate question against measured local opinion rather than against a national ideological template, early enough that the answer can change who files. The second is to treat the target list as an estimate with error, write down how confident you are in each line, and fund the uncertain edge rather than only the center.

The third is about sequencing. The committee that doubled its funded challengers did it in the spring, which is when a challenger can still build an organization. A campaign that waits for a September poll to decide whether a seat is in play has already spent the window where that decision was cheap.

Frequently asked questions

How many seats did Democrats gain in the 2006 midterm election?

Democrats gained 31 seats in the United States House of Representatives and six in the United States Senate on 11/07/2006. The House finished at 233 Democrats to 202 Republicans, and the Senate finished at 49 Democrats, 49 Republicans and two independents who caucused with the Democrats, giving them a majority of 51.

Which Senate seats did Democrats win in 2006?

Democrats took six Republican-held US Senate seats in 2006: Pennsylvania (Bob Casey Jr. over Rick Santorum), Ohio (Sherrod Brown over Mike DeWine), Rhode Island (Sheldon Whitehouse over Lincoln Chafee), Missouri (Claire McCaskill over Jim Talent), Montana (Jon Tester over Conrad Burns) and Virginia (Jim Webb over George Allen). Webb's margin was 9,329 votes and Tester's was 0.9 points.

What was the DCCC Red to Blue program in 2006?

Red to Blue was the Democratic Congressional Campaign Committee's program for directing donor money, strategic advice and staffing to House challengers. In the 2006 cycle it raised 22.6 million dollars for 56 campaigns, compared with 27 candidates supported in the 2004 cycle, and the 2006 total was 15.1 million dollars higher than 2004.

Why did Democrats recruit conservative candidates in 2006?

Democrats recruited candidates whose records matched the districts they were running in, because many of the available Republican-held seats were conservative. Brad Ellsworth was a county sheriff who opposed abortion rights and gun control and won Indiana's 8th district by 22 points. The approach won seats and also produced a caucus that was harder to hold together on legislation.