Ace reporter and former debate partner Scott continues his series on the Israeli elections:
Exit polls currently show Kadima with a slight lead in the overall vote count while Likud is a razor’s edge behind. It’s difficult to get accurate information about the Israeli election results thus far, but I can tell you one thing – I don’t trust the exit polls. As a general rule, exit polls are just bad statistics. They are frequently not randomized (as appears to be the case here, more on that in a second), always have a higher intrinsic margin of error (approaching unusable), and often aren’t weighted by the proper metrics. In our own election, the final results were far different from the ones predicted by the exit polls (at least in terms of actual numbers, if not the predicted winner).
Most of the exit polls I’ve seen quoted are from densely populated urban centers like Tel Aviv and Jerusalem. This makes sense, as it’s easier to develop a large pool of respondents where there are more people voting. But it’s my understanding that these urban centers skew toward the center-left and left (much like our own urban centers). That’s not to say that Jerusalem is as liberal as Washington D.C., far from it. However, I think that the political climate in these cities is far more sympathetic to Kadima and Labor. What I’m getting at is that the available exit polls do not represent random samples, and I’ve seen no descriptions of their methodology, leaving me to conclude that it’s more likely than not that compensation for this geographic bias through mathematical weighting is not happening.
While recent polls showed Kadima closing the gap with Likud, none of them gave Kadima a lead, and I still think Likud is going to take the day. Over at the American Scene, Noah Millman has a smart post up on the various coalition possibilities taking shape. He concludes that a unity government (between Likud, Kadima, and/or Labor) is the most likely outcome. I disagree. Assuming Likud wins, they have no incentive to add an opposing party to their coalition when Lieberman could give them a majority without adding a potentially paralyzing political ally. Moreover, any Netanyahu-led unity government would be ideologically incoherent, and therefore more likely to collapse.
If Kadima wins, the story is much sadder. They can either try to pull some of the smaller hardline parties (like Shas) into a ragtag coalition of disparate partners (like the last coalition, only more unstable), or they can try the unity government route. Assuming Kadima wins, I think a unity government is more likely (Millman thinks the exit polls are right and Kadima will win, which probably explains his conclusions), if only because it allows the Left to blame Likud when the government inevitably collapses. Either way, a Kadima win heralds a weak government that may not last very long. Which, in the end, also strengthens the hardliners. Damned if you do…