11/02/2006 04:04 pm ET | Updated May 25, 2011

What the Numbers Really Say

Although the Democrats hold a large advantage in generic ballot polls, there has been considerable uncertainty regarding whether the Democrats would win enough seats to control the U.S. House of Representatives. Doubts are often expressed about the accuracy of the generic ballot polls (polls which ask respondents only which party they plan to vote for, not specific candidates). How district lines are drawn raises further doubts about whether the Democrats could win a sufficient majority of the vote to win a majority of seats.

In our study [pdf] we estimate how the generic ballot "vote" translates into the actual national vote for Congress and then how this vote translates into the partisan division of seats in the House of Representatives. Based on generic ballot polls as of October 24, we found that Democrats are likely to gain 32 seats, making a Democratic takeover of the House (a gain of 15 seats or more) a near certainty.

A Democratic pickup of 32 seats might seem high to some readers, which begs the question, how are we interpreting poll data differently than other race watchers? To begin with, we estimate a regression equation predicting the House vote in the 15 most recent midterm elections, 1946-2002, from the average generic poll result during the last 30 days of each campaign. The generic polls turn out to be very good predictors, as our study shows. Based on the average of the generic polls as of October 24 (57.7% Democratic, 42.3% Republican) the forecast from this equation is a 55% to 45% Democratic advantage in the popular vote.

But would this mean that the Democrats also win the most seats? The Democrats winning 55% of the vote would represent a 6.4 percentage point swing from 2004, when they received 48.6%. If Democrats were to win exactly 6.4% more of the 2006 vote in every district than they won in 2004, they would win 228 seats. However, an average swing of 6.4% percentage points will be spread unevenly--sometimes more than 6.4% and sometimes less. Moreover, the prediction that the average vote swing will be 6.4% is itself subject to error.

To account for these possible variations, our study relied on a set of simulations which take into account district-level characteristics. The simulations suggest that a predicted national vote surge of 6.4 percentage points would yield the Democrats 235 seats, for a 32-seat gain. This is 7 seats more than we would get with uniform swing. (See Figure 1.)

In theory, the predictive value of generic polls (which, remember, ask only which party a respondent plans to vote for) should be reflected in the district specific polls asking about individual candidates. For a reality check, we compared our district level predictions from our simulations with the results of available district polls. The two sets of numbers matched nicely. Averaged across 32 Republican-held districts with October polling, the average district level poll margin is 51.7% Democratic, 48.7% Republican. For these same 32 districts, our average prediction is 50.3% Democratic, 49.7% Republican. Given these numbers, it may be that our simulations actually underestimate Democratic strength in the sampled districts.

Election watchers have many means toward predicting the election outcome. They may, for example, watch the candidate polls, follow the markets, analyze each race qualitatively or make a statistical prediction. The statistical prediction we make is based on an advanced model that takes into account both district-level data and historical data to forecast a partisan swing. If the generic poll results continue as they have been, our model predicts a very strong year for Democratic congressional candidates this year. Only November 7th will tell for sure.

Joseph Bafumi, Dartmouth College;

Robert S. Erikson, Columbia University;

Christopher Wlezien, Temple University