The Early Outlook for the 2012 Congressional Elections: A Forecasting Perspective

Facebooktwitterredditpinterestmail
In today’s Crystal Ball, Alan Abramowitz — whose election models are among the best in the business — provides an early look at what they tell us about the race for the House and the Senate. We suspect that these models are a little pessimistic for Democrats at this early point, although we agree with Alan’s conclusion, which is that Republicans are in strong position to keep control of the House and that they should at the very least cut into the Democrats’ 53-47 edge in the Senate. Forecasting models are an important tool in helping to predict elections, but they are not the only tool the Crystal Ball uses. A holistic approach — using polling, models, and race-by-race evaluations — gives a fuller picture, and this is how we traditionally make our projections as Election Day approaches.

A relatively simple model incorporating four predictors — the number of seats the Republican Party holds going into an election, whether it is a presidential or midterm election year, the net approval rating of the incumbent president, and the standing of the parties on the generic congressional ballot — can be used to forecast the seat swing in House and Senate elections well before Election Day.  The model accurately predicted the Democratic takeover of the House of Representatives and Senate in 2006 and the Republican takeover of the House in 2010.

After estimating the models based on the results of all House and Senate elections since the end of World War II, the forecasting equations for the 2012 House and Senate elections are as follows:

CRHS = (1.35*GENBALLOT) + (.21*PRESAPP) – (.36*PRHS) – (19.6*MIDTERM) + 86.1

CRSS = (.18*GENBALLOT) + (.05*PRESAPP) – (.81*PRSS) – (2.9*MIDTERM) + 14.8,

Where CRHS is change in Republican House seats, CRSS is change in Republican Senate seats, GENBALLOT is the average current Republican margin on the generic congressional ballot, PRESAPP is net presidential approval coded according to the party of the president, PRHS is previous Republican House seats, PRSS is previous Republican Senate seats, and MIDTERM is a variable distinguishing Republican and Democratic midterm elections from presidential elections that is coded +1 for midterms under a Republican president, 0 for presidential elections and -1 for midterms under a Democratic president.

The model is considerably more accurate for House elections than for Senate elections, which is not surprising given the much smaller number and greater competitiveness of Senate elections.  However, the two equations are actually very similar.  In both models, the generic ballot and presidential approval have significant effects on seat swing, and midterm elections produce a consistent swing against the president’s party.  In addition, in both models, and especially in the Senate model, the more seats Republicans are defending, the greater the Republican seat loss.  In other words, exposure predicts losses.

Based on the most recent polling results, which show a near tie on the generic ballot and a net presidential approval (approval minus disapproval) of close to zero, the House forecasting model predicts a very small Democratic seat gain (2-3 seats) in the House but not nearly the 25 seats Democrats would need to take back control of the House.  On the other hand, the Senate forecasting model gives Republicans a good chance to regain control of the Senate with an expected pickup of 6-7 seats.  That is due almost entirely to the fact that Republicans are defending only 10 Senate seats this year while Democrats are defending 23 seats.

The Senate forecast especially should be interpreted cautiously because the Senate model has a fairly large error term due to the small number of seats in each election.  And of course, it is still early and both the generic ballot and the presidential approval variables could change over the next few months.  However, both have been fairly stable in recent weeks.  Based on these results, it would be surprising if Republicans did not hold onto their majority in the House in 2012 and gain at least a few Senate seats.