The Other McCain

"One should either write ruthlessly what one believes to be the truth, or else shut up." — Arthur Koestler

Final Warning: Polls Are Not Elections

Posted on | November 2, 2010 | 25 Comments

PANAMA CITY, Fla. — We now have the final Cook Political Report forecast (“The midterm maelstrom pulling House Democrats under shows no signs of abating”) predicting a pickup of 50-60 House seats for Republicans. That would be a truly historic result, but it’s actually a note of sober caution compared to Jay Cost’s “Hulk smash” forecast:

A [Gallup final poll predicting GOP] victory of 15 points suggests Republican gains well in excess of my previous estimate of 61 seats. The Abramowitz model suggests a pickup of about 76 seats . . .

With all due respect to Jay Cost: Fuck you and the Abramowitz model you rode in on.

Polls don’t win elections, and regression analysis sure as hell doesn’t win elections. Politics is not a science, and trying to reduce elections to trends, polls and mathematical formulae is one of those situations where when the only tool you’ve got is a hammer, every problem looks like a nail.

If you’re a thousand miles away from a district and don’t actually know anything about what the candidates and their campaigns are doing, it is tempting to look at poll numbers, examine past voting trends, and start making assumptions about what the result will be. But when we yield to that temptation, we ignore the Hayekian insight: Information is diffused throughout society in such a way that no one — not even the best-informed “expert” — can know everything.

So Jay Cost doesn’t know everything, Charlie Cook doesn’t know everything and Michael Barone doesn’t know everything, either. Yet their status as political experts requires them to make predictions and we mere mortals . . . well, we don’t know nothing about winning no elections.

Anyway . . .  Allahpundit is optimistic, which is so abnormal it should worry us all. Jim Geraghty sees a net gain of 70 seats for the GOP (76 pickups minus six losses, including Ben Quayle going down in AZ-3). Nate Silver errs on the side of caution:

Our forecasting model, which is based on a consensus of indicators including generic ballot polling, polling of local districts, expert forecasts, and fund-raising data, now predicts an average Republican gain of 54 seats . . .
Of the projected Republican gains, many are tenuous. If we allocate all 435 seats to the leader projected by our model — no matter how slim the margin — Republicans would net a gain of 59 seats. In 15 of these 59 seats, however, the Republican is projected to win by fewer than 2 points. It is likely that Republicans will lose at least some of these — which is why the model forecasts an average gain of 54-55 seats, rather than 59, when looking at the seats on a probabilistic basis. If they lost all of them, however, their gains would be merely 44 seats, which would put the Democrats within striking distance of retaining the House.

See what I mean? Because there has been little polling in many House races, Silver is forced to use such inputs as fund-raising totals and “expert forecasts” — which is to say, he’s relying on second-hand data. And as if Republicans need a reason to be scared to death, Silver notes a very realistic possibility: Despite everything, Democrat manage to hold on t0 the majority.

To quote one expert: Don’t get cocky!

While I’m optimistic that Republicans will take the House majority, I’m trying to temper GOP overconfidence in my latest American Spectator column:

Parked beside the Steve Southerland campaign headquarters on Baldwin Road is a big RV with an orange sign on the side designating it the “I’ve Had Enough Express.” And indeed, if polls can be trusted, voters in Florida’s 2nd District have finally had enough seven-term Democrat Rep. Allen Boyd.
“We’re feeling very good about how things are going,” said Southerland campaign communications director Matt McCullough, as he sat in an office inside the yellow brick house that was once the home of the GOP candidate’s grandfather. Polls showing Southerland with a double-digit lead over Boyd, but the campaign is “not taking anything for granted,” McCullough said. “We’re still out there working as hard as we did from Day One.”
Despite such necessary cautions against overconfidence, Republicans here clearly expect to be celebrating a Southerland victory Tuesday night at the Boardwalk Beach Resort. If they want to learn whether their party wins a majority in the House of Representatives, however, they’d better be prepared to stay up past midnight. . . .

Please read the whole thing. I don’t know everything, but what I know, I know. And I know I’m going to take a nap today, because I expect this thing to go to the wee hours of the morning.

I’m leaving for Miami in about an hour. Please hit the tip jar.

Update (Smitty): Welcome InstaPundit readers! Allow me to direct your attention to a post supporting my local candidate.


25 Responses to “Final Warning: Polls Are Not Elections”

  1. Jason
    November 2nd, 2010 @ 1:10 pm

    Was it really necessary to tell Jay Cost “Fuck you”? Nothing he said warranted such a classless and hostile comment.

  2. John
    November 2nd, 2010 @ 1:54 pm

    You do realize that the media is trying to overproject Republican gains so that when they don’t materialize, they can claim the GOP failed and that the country obviously still loves the Obama agenda, yes?

  3. The Snarkinator
    November 2nd, 2010 @ 2:43 pm

    Go out and vote heavy. We have to beat the margin of fraud.

    and @ Joe point #3: ask for $75, because it always costs more and the budget numbers don’t have to add up.