Saturday, July 31, 2010

HOUSE UPDATE: Republicans Edge Up to a 48 Seat Gain

Based on polling data released over the last seven months, Republicans are expected to gain between 30 and 65 seats in the U.S. House of Representatives (with a 90 percent level of credibility). According to the model, the most likely outcome is about a 48 seat gain by Republicans, up five seats from the previous estimate. If the polling holds steady, Republicans seem to be poised to retake control of the U.S. House of Representatives with the largest electoral landslide since 1994. In fact, based on a Monte Carlo simulation, the Republicans have an 80 percent probability of retaking the House.

A statistical model that analyzes the results of 15 independent polling firms indicates that Republicans lead Democrats by a margin of 1.9 percent on the so-called generic ballot--a measure of national voter preference between the two major congressional caucuses. This figure is somewhat greater than the June estimate of 1.4 percent. The July estimate implies that, if the election were held today, Republicans would achieve approximately an 18 seat majority over the Democrats--a modest, but decisive advantage.

Over the next five months, I will update this model and estimate Republican electoral gains. The charts below will provide the change in voter preferences over time, along with the probability distribution for the number of seats gained by Republicans on election day. See my April 30, 2010 post for an explanation of my methodology.

NOTE: I tweaked my charts for the expected margin of majority and the probability distribution for the number of seats gained by Republicans.

Posted by at

2 comments:

  1. Greg RoseAugust 21, 2010 at 6:51 AM

    How does your statistical model and projections compare with the methodology followed over at Election Projection . com? They are only projecting a 35 seat gain by the Republicans in the House which would leave the Dems with a 7 seat majority.

    BTW I found your blog doing a Google search. Your projections are very encouraging to this voter.

    ReplyDelete
  2. schakjAugust 25, 2010 at 7:32 AM

    Greg, thanks for posting.

    Election Projection looks at each race individually. I think he incorporates pundit ratings, polls on specific races, and partisan affiliation. I calculate my projections using national general congressional ballot data. Because I do not use data from the specific congressional districts, my estimates are less precise than Election Projection.

    However, I aim to project the outcome on election day, while I think Election Projection is really predicting the outcome if the election were held today. I suspect that Election Projection's estimates will converge towards my own analysis as more polling is completed in the specific congressional districts. Generally, pundit analysis and contest-specific data lag national data.

    BTW, I am working on a way to list the Democratic seats that I think are most likely to flip.

    ReplyDelete

Subscribe to: Post Comments (Atom)