Report released September 30, 2002
Results reported November 8, 2002
2004 Projections posted November 8, 2002
CVD's model of projecting House winners based only on past federal election results and the senior of incumbents had a 99.9% accuracy over the past 3 elections. Out of more than 900 projections over the last 3 elections cycles based, the model made only one incorrect projection. We have applied the model to the 2002 elections and report our findings in a detailed report. Using data available as as November 6, we report on the accuracy of our projections.
Great effort went into gathering the data used in this report. Because the Center believes this information should be readily available to the public, we are releasing all of the data and the precise methodology used. We hope that members of the public will make use of the data and the model and experiment with adjusting variables and assumptions to gain a better understanding of how US House elections function.
Full Report: Monopoly Politics 2002: How "No Choice” Elections Rule in a Competitive House
Data and Model: This link contains spreadsheets for making projections for House races for any year from 1996-2000 as well as 2002, all of the data used in these projections, and detailed explanations of the model and all of its variables.
User's Guide, Algorithm and Glossary of Terms: See links on the left.
100% accuracy of 2002 projections and 2004 projections in more than 350 races are already available. (Please see Monopoly Politics 2004)
The September 30 press conference was broadcast live on CSPAN-1 and re-aired four times. It can be viewed at cspan.com. You can also view the press conference on CSPAN's Politics page by searching for the September 30 2002 listing "The Race for the U.S. House: Past, Present & Future."
For media inquiries, please contact:
Rob Richie, Executive Director