Elections for the United States House of Representatives were held on November 7, 2006, with all of the 435 seats in the House up for election. This article discusses predictions for outcome of House races as a whole. Individual races that were notable are discussed in detail (a paragraph or so, each) at United States House elections, 2006; that article also has information on the results of the election.
The predictions below were those made just before the election was held.
Below is a table summarizing various non-partisan election analyses. Although the nomenclature varies slightly, the categories can be understood as follows:
"Safe" — the race was not expected to become competitive.
"Favored" — the race was not competitive, but might become competitive if new political factors intervened.
"Leans" — the race was competitive, but one candidate had a clear advantage.
"Tossup" — neither candidate had a clear advantage.
| Source | Safe Democratic | Democratic Favored | Leans Democratic | Tossup | Leans Republican | Republican Favored | Safe Republican | |
| CQPolitics.com as of Nov 6 | 182 Democratic Seats | 16 Democratic seats 1 Republican seat | 5 Democratic seats 9 Republican seats | 23 Republican seats | 20 Republican seats | 22 Republican seats | 157 Republican seats | |
| Cook Political Report as of Nov 6 | 182 Democratic Seats | 13 Democratic seats | 6 Democratic seats 5 Republican seats | 1 Democratic seat 38 Republican seats | 12 Republican seats | 13 Republican seats | 165 Republican seats | |
| Larry Sabato's Crystal Ball as of Nov 6 (← and → show trends) | 193 Democratic seats | 6 Democratic seats 1→ 3 Republican seats | 4 Democratic seats 1→ 26 Republican seats ←5 | (none) | 31 Republican seats ←10 - 3→ | 10 Republican seats | 162 Republican seats | |
| Majority Watch (final) as of Oct 30 | 198 seats | 24 seats | 18 seats | 2 seats | 10 seats | 6 seats | 177 seats | |
| Rothenberg Political Report (← and → show trends) as of Nov. 6 | 197 Democratic seats | 1 Democratic seat 3 Republican seats | 1 Democratic seats 6 Republican seats | 3 Democratic seats ←3 37 Republican seats ←8 10→ | 3 Republican seats | 8 Republican seats | 166 Republican seats | |
| Electoral-vote.com as of Nov. 6 | 203 Democratic seats 36 Republican seats | 1 Republican seat | 195 Republican seats | |||||
| Ed Fitzgerald's survey of polls as of Nov. 6 | 227 seats | 4 seats | 204 seats | |||||
Tradesports offers a contract on the Republicans retaining control of the House. On Tuesday, November 7 at 09:31 PM GMT it was indicating a probability of 16.1% that the Republicans will retain the House.http://tradesports.com/aav2/trading/contractInfo.jsp?conDetailID=291849&z=1162238556484
Iowa Electronic Markets offers four contracts covering all possible outcomes of the House and Senate elections.
As of 9:15 PM GMT, Tuesday, November 7, 2006, the two contracts predicting Republican control of the House were trading at $0.140-0.219 and $0.002-$0.010 respectively. Adding these values resulted in a probability of between 14.2% and 22.9% of Republicans retaining control of the House.http://iemweb.biz.uiowa.edu/quotes/Congress06_Quotes.html
Because of the large number of individual races covered by nationwide political services, these analyses often were a week or two behind the results obtained by local polling organization, but are still very useful.
The Cook Political Report, an independent, non-partisan election analysis newsletter, listed 89 seats - 68 held by Republicans and 21 by Democrats - as being potentially in play.[1]
As of November 6, 2006:
CQPolitics.com, an independent, non-partisan election analysis newsletter, at the end of November 6, 2006, made the following analysis:[2]
In early August, Sabato predicted a pro-Democratic shift of 12-15 seats in the House. In mid-October,[3] he revised his estimate, predicting a net gain for the Democrats of 18-22 seats. On October 26, he revised his estimate to 21–26,[4] November 2, he upped his estimate to 24–30,[5] and on the day before the election, he raised it to 25-33[6] with the following breakdown:
Electoral-vote.com is a site run by Andrew Tanenbaum, a professor of computer science at Vrije Universiteit. While it mostly covered Senate races in 2006, it also made algorithmic predictions for the House-based entirely on independent polling data. (Where no independent polls exist, the 2004 election results were used).
As of November 2, Electoral-vote.com was predicting 38 Democratic pickups and no Republican pickups.
His calculations predict the House to comprise: 240 Democrats, 193 Republicans, 1 Ties. The Democratic pickups are: AZ-05 AZ-08 CA-11 CO-04 CO-07 CT-02 CT-04 CT-05 FL-13 FL-16 FL-22 IA-01 IL-06 IL-10 IN-02 IN-08 IN-09 KY-03 KY-04 NC-08 NC-11 NH-02 NM-01 NY-19 NY-20 NY-24 NY-25 NY-29 OH-01 OH-02 OH-15 OH-18 PA-06 PA-07 PA-08 PA-10 TX-22 WI-08[7]
"Mystery Pollster" Mark Blumenthal and University of Wisconsin–Madison Professor Charles Franklin aggregated polling data for more than 60 competitive house races[8] and offered objective analysis on the methodology and the reliability of the numbers. On October 1, 2006, it began to post the most recent polls for these 60+ districts as soon as they were released to the public.
Majority Watch is a non-partisan poll from RT Strategies and Constituent Dynamics.[9] Their polling is done by automated telephone surveys of likely voters.
The Rothenberg Political Report is a non-partisan analysis of American politics and elections. Stuart Rothenberg is a regular columnist in Roll Call. On Nov 6 he predicted a Democratic gain of 30-36 seats, with the following breakdown:http://rothenbergpoliticalreport.blogspot.com/2006/11/2006-house-ratings.html
At politicalforecasting.com http://www.forecastingprinciples.com/Political/index.html, predictions for the 2006 House of Representatives elections were categorized by method (e.g., statistical models, prediction markets) and averaged within and across methods to arrive at the Pollyseat, the number of seats the Republicans are expected to lose on November 7. As of October 30, the Pollyseat value stood at 21. In other words, the Republicans were expected to lose control of the House of Representatives.
Ed Fitzgerald, at his blog, unfutz, compiled a summary of several dozen forecasters, with statistics and charts.
The map below shows the latest polling information as of November 6, 2006, for House races. No party polls included. Darker colors indicate stronger support. Where no polls were available, last year's election results were used. If two polls tied for most recent, the longer poll was used.