The respected U. Va. political science professor Larry Sabato's been getting headlines all week for his analysis leading to the assertion that, if the election were held today, the Republicans would pick up 47 seats in the House. Problem is, his numbers don't add up.
I refer you to Prof. Sabato's "Crystal Ball".
I appreciate the fact that Sabato provides a seat-by-seat listing of those he currently rates as "Likely Dem", "Lean Dem", "Tossup", "Lean Rep", and "Likely Rep", and frankly, I don't have much issue with the ratings, as regards the House races I'm following closely.
Here's the thing, though: of the ones leaning or likely Democratic, most are currently held by Democrats--only three are Republican. Of the ones leaning or likely Republican, many are Republican-held, 17 are not. Take those and you get net +14 Republican. Fair enough.
Of the 29 seat races he rates as toss-ups, 28 are held by Democrats, and only one is currently Republican. If a toss-up means 50-50, split them evenly (give the Republican seat back to them), and you get +28 Republican. You could argue that all 29 will go Republican, but then why would he call them "toss-ups"? Even that somewhat extreme, certainly not objective, forecast would be "only" +42 Republican--bad enough to flip the House, but still not +47. So where does that number come from?
I have to accuse Prof. Sabato of some intellectual dishonesty--either he doesn't believe his ratings of the races, or he is disregarding them in favor of something else. Perhaps surfing the "wave" of perceived discontent?