This is Labor Day weekend, the traditional beginning of the general election campaign. Of course, it's been going on for months, but a lot of the focus has been on primaries--which are still not complete. We've been making forecasts for nine months or more.
There are some very bold predictions being made in these unsettled days of great discontent; most of these bold predictions are favoring the Republicans' chances this year. These are people who are surfing what they see as a great wave of reaction to Democratic control of government, something driven by dissatisfaction of independents and readily seen in recent polling trends. Many of these admit there is no great wave of sentiment, coherence, or incisive argument on the Republican side--it's just anti-Democrat and anti-Obama, which, in our two-party system, can easily be enough.
It's Saturday, So I've Come to Bury Sabato, not Praise It
I have to point out something which I see as a surfing prediction from Larry Sabato, the respected political science professor and longtime electoral commentator at U Va. He has decided to take the spotlight, going public with a prediction that the Republicans stand to gain 47 seats in the House (they need to gain 39 to get control). To his credit, he has seat-by-seat ratings of those he does not consider safe; however, the math does not really add up. Of the seats he rates "Leans Dem" or "Likely Dem", most are currently held by Democrats but three are not. Of those rated "Leans Rep" or "Likely Rep", the majority are held by Republicans, but 17 are not. Of the seats rated "Toss-Up", 28 are Democratic-held, one held by Republicans.
If you credit his calls on the individual races--and for the ones I know about, I have no quarrel with his ratings--the Republicans have net +14, with 29 races as toss-ups. Give half of the toss-ups to the Republicans (plus the one currently Republican), and you get +28 for the Republicans--very close to the 25-30 seat gain I've been predicting all along. Even if you give every one of the toss-ups to the Republicans, you get +42--good enough to change control, but still not the +47 he trumpets.
I have to accuse him of a little intellectual dishonesty here: either he doesn't believe his own ratings on the individual races, or he decided to disregard them to get some headlines. He owes us a little better explanation, particularly after he leads his post by bragging about how he advises his students to stick to the facts and not go with feelings.
There are some good aspects to Sabato's post, particularly his contribution to the analysis of gubernatorial races, a key dimension of this year's election with legislative redistricting coming up. Quite rightly, he's highlighting a big Republican gain in statehouses and its major implications. However, as a New Mexican, I have to complain how he highlights all the races he's expecting to cause change of control, but doesn't mention by name our state, for which he is calling for a Republican takeover. I hope he's wrong there; I'm not sure he is, but I at least think we deserve to mentioned.
Finally, his ratings of Senate seats, which people have focused upon, are generally reasonable and consistent with current polls. He's got the Republicans gaining seven seats, but specifically is not discounting the possibility of a gain of 10, which would give them control of the Senate, too. He also points out an interesting fact, that in the six cases of the House flipping control since World War II, the Senate has also flipped each time, even though some of those changes were not anticipated.
I still go the other way: the House will not flip, nor the Senate. My belief in the Democrats' holding such seats as PA, CO, IL, CA, WA, NV, and WI (the first two rated by Sabato as leaning Republican, the others as toss-ups) is more based on Americans' coming to their senses and to the inferiority, or even outright imbecility, of some of the Republican candidates in key contested seats than it is hard poll numbers, but I do have this on my side: not a single general election vote has been cast yet, and polls do not decide any races.