Charlie Cook is usually known for his calm, somewhat conservative (in the non-political sense) predictions for elections. He waited quite a while before joining the chorus that said Republicans would retake control of the House in this midterm cycle, and has remained diffident to the GOP’s chances in the Senate until recently. Today, Cook gives his final predictions for the midterms, which hit the disaster level for Democrats:
In its final forecast for the election cycle, The Cook Political Report expects a gain for Republicans of 50 to 60 seats in the House, with six to eight seats in the Senate. Below are the final outlooks and latest ratings changes.
His reasoning for the House is rather self-evident. Democrats have trailed all year in polling, not just in the generic polls but also in specific races. More than a quarter of the Democrats in Congress have trailed their GOP challengers at some point over the last few weeks, which means about 64 of them are in deep trouble. Half of them have failed to hit the 50% mark in surveys. Somehow, though, Cook believes that Democrats have 181 safe seats even with those numbers in play, while Republican incumbents can count on 204 certain holds.
Cook’s prediction on the Senate seems a little overly cautious:
The Cook Political Report is adjusting its current outlook to reflect a net gain for Republicans of 6 to 8 seats, down from 7 to 9 seats. While it is becoming increasingly likely that Republicans will hold all 18 of its own seats, Democrats’ prospects in three of their 19 seats have improved in recent days. Sens. Barbara Boxer in California and Patty Murray in Washington now appear to be headed for re-election, albeit by small margins. In the special election in West Virginia, Democratic Gov. Joe Manchin now holds an advantage. Currently there are 57 Democrats, two independents that caucus with Democrats, and 41 Republican Senators. Post-election, Republicans could hold between 47 and 49 seats to 51 to 53 seats for Democrats. This new outlook means that the odds of Republicans winning a majority in the Senate are now non-existent.
I wouldn’t be too quick to discount a couple of these races. Manchin has taken a late lead in polling, but he has seriously underperformed the entire cycle for a reason. Palin just spent some time supporting John Raese in the last couple of days, and that could have a big impact. Murray also doesn’t look like a lock for re-election in Washington — in fact, the trends there look like Rossi has picked up enough momentum to beat her. Boxer has an edge in California but has lost five points in a week. Linda McMahon trails in CT, but Rasmussen’s latest poll shows her halving Blumenthal’s lead in a week (more on that later). If a national wave pushes these three to the GOP, Republicans will come pretty close to control of the Senate, or at least force Joe Biden to cast tiebreaking votes.
As Nate Silver suggested last night, there is good reason to think that the Republican wave this year has been underestimated:
Throughout this election season, I’ve tried to stress that there is a great deal of uncertainty in the outcome. Not necessarily uncertainty in individualraces: people probably overestimate that. But uncertainty, rather, in where the House and the Senate will finish over all. People probably underestimate how strongly polling and forecasting errors are correlated from district to district. If Republicans tend to overperform expectations in some races, they will probably also overperform in many, most, or maybe even almost all races. The same holds true for Democrats. (The most recent time something like this occurred was 1998, when polls underestimated the standing of Democrats by 4-5 points nationwide and in almost all individual races.)
If a situation like the one I described above transpires, it’s going to catch a lot of people by surprise. It really shouldn’t; it’s well within the realm of possibility.
He gives five potential reasons for missing the amplitude, and pay close attention to #5:
In addition to wrongly excluding some Republican “unlikely voters” (see Point No. 2), it’s also conceivable that some likely voter models based on past voting histories are overrating the propensity of Democrats to vote. The reason could be that some of them are based on past voting history, and a common question is whether the voter had participated in the last two elections.
But the last two elections — 2006 and 2008 — were good ones for Democrats, one in which there was little if any “enthusiasm gap,” or it may even have favored Democrats. This is, in fact, quite atypical: Republicans usually do have a turnout advantage, especially in midterm elections. Their demographics are older and whiter, and whites aged 50 and up are the most reliable voters. If likely voter models are benchmarked to 2006 and 2008 patterns, therefore, they could underestimate the turnout gap, giving too much credit to Democrats who voted in 2006 or 2008 but who don’t ordinarily. Sean Trende at Real Clear Politics makes a nice version of this argument.
The results Cook predicts would represent a stunning victory on any terms. It would put Republicans in control of the agenda at least in the House and act as a brake on Obamanomics and wild spending, assuming that the GOP has learned its lesson from 2006 and 2008.
My predictions: I’m guessing that the GOP picks up 65 seats in the House, and nine Senate seats.
Update: AJ Strata has more optimism than I do, or at least more than I will allow myself to indulge …. (h/t: Adam Baldwin)