Tuesday, February 24, 2009

So the people didn't vote the way the Israel Democracy Institute thinks they should and suddenly electoral reform is all the rage.

There are a number of directions that reform can take and they need to be carefully distinguished. The one most often discussed is switching from proportional to first-past-the-post regional elections. I've already discussed that before. In short, the idea has some merit but plenty of risks. It's certainly not the most pressing reform.

Less discussed but simpler and more pressing is how the Knesset forms a government. One current favorite idea of the leftist elites is that the head of largest party should be named prime minister with no Knesset ratification necessary. Coincidentally, in the current situation that would make Tzipi Livni prime minister despite coalitional support approaching zero.

That fact alone should explain the abject stupidity of the idea but let me explain in broader terms. The effect of changes to a mechanism for coalition formation is not only at the last stage. The main effect is the one on the strategic considerations of the voter. When voters know that the largest part will get a huge bonus (in this case, its head becoming prime minister), they have a very big incentive to vote for a party that has a chance to be the biggest one. Such incentive for strategic voting already exists but the proposed change would exacerbate it. This would either destroy small parties (if their voters shift to the large parties) or it would turn them into spoiler parties (if their voters do not shift).

As it happens, I'm not a huge fan of small parties (they have a natural tendency to focus on narrow interests at the expense of broader ones), but it would be catastrophic, socially and politically, for them to be wiped out. Furthermore, if the voters for a given small party belong predominantly to a certain bloc, it would be a distortion of the voters' will for voters for that party to harm the chances of their preferred bloc to form the government. Supporters of this proposal understand full well that, since the right is divided fairly evenly among religious and secular parties, the proposal is advantageous for the left which is not inherently divided (except for the Arab parties).

From a purely objective point of view, the proposal also facilitates a situation in which the government and the Knesset are at complete loggerheads, potentially creating paralysis.

A much more reasonable proposal is that parties be allowed to form blocs prior to elections that would name a joint candidate for prime minister. All votes for any of those parties would then go towards the bloc's candidate for prime minister. The prime ministerial candidate of the largest bloc would automatically be named prime minister.

The strategic effect would be that, most likely, two blocs would result, since voters will only vote for a party in a bloc with a legitimate chance to be the largest one. The advantages of this method are manifold:
1. Voters for any party would know in advance where their party stood. This is simply more fair to voters.
2. The political power of the larger parties would increase since post-election horse-trading would be eliminated. It would also be harder for small parties to jump blocs after elections, having sold their voters on a particular they (For this advantage to be fully realized, the conditions under which a government could be brought down through no-confidence would need to be tightened a bit.)
3. The numerical power of small parties would remain stable, if not increase, since the incentive to jump to a large party is greatly diminished. (This point could make it a hard sell to the big parties, who might not be savvy enough to realize that point 2 is more important than point 3.)
4. Extremist parties that can't join any bloc would be severely weakened.
5. The chances of ending up with a prime minister who does not have broad Knesset support are small, since the government is guaranteed to have at least the support of the largest bloc. I concede that it is theoretically possible for there to be three blocs with the largest one having only 40-something seats, but this sounds like a stretch (and it wouldn't be as catastrophic as, say, 28 seats).

There's much more to say on this topic but I'd be happy to get comments on this teaser.

Wednesday, February 11, 2009

Once the soldiers' votes are counted, the messy calculations begin. If you want to know how the system of seat allocation works see my post on it from back in the previous election season.

Addendum: You can get more info in Hebrew on this wacky system here.
Note that the system favors large parties (or pairs of parties with remainder agreements), so Likud/YB are well-placed to benefit. In the past, parties have even picked up two seats as a result of the remainder calculations . (For example, Likud in 2003 went up from 36 to 38.)

Update (Thursday 12:00): The final tally not counting double-sealed ballots (soldiers, diplomats, patients and prisoners) has been posted here. Right now, Lieberman is teetering between 14 and 15 seats. Likud/YB will probably have to waste their first bonus from the remainders just to get that 15th seat secured. Unfortunately, once Lieberman's 15th phantom seat is accounted for, Kadima is closer to picking up an extra seat than Likud/Lieberman.

On the other hand, the soldiers' (et al.) votes are almost finished being counted and they are running in favor of Likud over Kadima (though, not by a huge margin) and Lieberman is very strong (though, less than Kadima).

The results of the exit polls have been announced and the situation is as follows. Kadima is projected to get about 30 seats and the Likud about 28 seats. Lieberman is in the area of 14-15 and Labor in the area of 13. Both Bayit Yehudi and Ichud Leumi seem to have gotten in, though barely. One of the two Arab parties might not get in. Overall, the right bloc is in the area of 63-64 seats.

So what will happen? First, the dry legal facts. the president must consult with the various parties and then make a determination which party has the best chance to form a coalition. That party gets the first chance to try. By all accounts, Likud has the best chance to form a coalition, but Peres is strongly tied to Kadima and might choose to give them the first shot.

So how will it play out? It's pretty straightforward. This is a bargaining game between Likud and Kadima. The most obvious coalition is one that includes both of them. The question is which gets to form the government and which joins it. The fact that Kadima is projected to be a bit larger (this is likely to change) seems like a key point in its favor, but in fact is not. The crucial issue in a bargaining game is the "disagreement point". The disagreement point is the result that ensues when the two sides fail to reach agreement. In this case, failing agreement, Likud has a plausible right-wing government, but Kadima has no plausible government. As a result, Likud has the upper hand in the negotiation with Kadima.

Several fine points should be noted. First of all, put yourself in Lieberman's position or in Shas's position. Given the fact that if Likud goes with Kadima, the negotiating position of these parties is reduced to close to zero, they have a very strong incentive to prefer a narrow government. This puts Likud in a strong position vis-a-vis them . For that reason, I predict that by tomorrow, we'll be seeing all these parties declaring their undying loyalty to a narrow Likud government. Second, from Likud's point of view, it would be better for Peres to give Kadima the first shot at forming a government. This is because once they fail to do so (and they will fail, if given the chance), Likud's advantage (the disagreement point) will be that much more clear to both sides. (Of course, if what you really want is a narrow right coalition, it might be better to just get on with it, but it's quite clear that that is not what Bibi wants. He wants only to wield the threat of such a government, not to realize the threat.)

One detail. As of now, Ichud Leumi has three seats. This might change once all the votes, including soldiers, are counted. If it holds, it means that Michael Ben-Ari is not in. While MBA and I share many views, the fact is that, fairly or not, there is just no chance of Ichud Leumi sitting in any coalition so long as he is in the faction. Without him, a narrow right coalition is that much more plausible.

A final note. One thing that held fairly constant in the polls for months was that Likud + Lieberman were worth 44 seats. Right now, they are projected at 42. My guess is that we'll see that number creep up to 44 as the real results come in. In fact, it is not unlikely that Likud will end up with more seats than Kadima.

Update (Wed. morning): A few brief observations. 1) It's hard to imagine anybody who voted for Kadima watching Livni's "victory" speech and not feeling buyer's remorse. 2) It seems that all of Kadima's surplus over what was expected came from Labor and Meretz. The tentative total of 44 for the three parties combined is what was anticipated. The Old Left is officially dead. (But voters on the left at least have the good sense to rally around the candidate with the best chance. If only the right had as much sense.) 3) Bibi will be the next Prime Minister and nobody knows it better than Peres. (The right has 65 and that number can only go up with the soldier votes and the remainder agreements.) That's why he'll work behind the scenes to engineer a deal that will give them equal power. It must be avoided. 4) The dumbest thing the right-wing parties can do is announce that they intend to drive a hard bargain with Bibi. The less likely a right-wing coalition is, the stronger Kadima is vis-a-vis Likud. We'll find out pretty quickly which of these parties is headed by a mathematician and which is headed by a Rov. 5) The next government will be Likud - Kadima - Lieberman - BY - one other dati party.

Tuesday, February 10, 2009

I've got time on my hands and really don't want to discuss elections. There'll be more than enough of that tomorrow.

In the last few years, the explosion of text available via the Internet has generated tons of research that can be applied to Torah. So far, very little of it has been exploited. Here is a short list of applications that are worth pursuing:

1. Using the Responsa Project as a base, for any given halachic topic, automatically construct a structured table consisting of the main sources relevant to the topic. Simple searches (as now done in the Responsa Project and similar searchable corpora) don't separate the main sources from those that mention a keyword in passing. They also don't find texts that are about the topic but don't use the query term. They also don't structure the results according to the interactions between sources. All this is doable provided that cross-references can be identified and properly exploited.

2. Given any anonymous Hebrew-Aramaic passage, use stylistic analysis to either identify the author (whether or not the passage is found in whatever corpus you're using) or to profile the author: what period did he live in, what region, who were his teachers.

3. Given multiple manuscripts of the same text, reconstruct the original text. This involves determining dependencies among the manuscripts and also using clever methods to determine the reliability of each manuscript even without knowing any ground truth against which to compare it.

4. Find a precise formal definition of kal ve-chomer that satisfactorily explains when it applies and when it does not.

5. Explicate the rabbinic theory of causality and indirect action. It should be able to explain when to apply the principle of grama benezikin patur as opposed to da-in dina de-garmi. Determine if the same theory is applicable to the laws of Shabbos.

6. Consider the set of all passages (loosely defined) in the Torah and determine the optimal clustering of the passages according to stylistic criteria. Measure the quality of the derived clusters to determine if they are sufficiently robust to qualify as organic units. (For the record, I don't see any theological issue here.)

If anybody is interested in any of these research problems, give me a shout and I'll go into greater detail.

Sunday, February 08, 2009

What will the results look like on Tuesday night?

The pollsters will tell you that a lot depends on the undecideds. This is nonsense. The undecideds are almost all undecided only about which of two very similar parties to choose. The real unknown is how many votes each of the parties south of Labor will get.

The Likud-Lieberman bloc holds a lead of about six seats over the Kadima-Labor bloc, modulo certain assumptions about how the votes will distribute among the smaller parties. These assumptions are based on nothing with any statistical significance. Shas can get anywhere between 7 and 14, Agudah can get anywhere between 4 and 10, etc. As with every election, the surprises will come down at the bottom and will strongly impact the overall results. Keep your eyes posted especially on Chadash and Ichud Leumi, both of which are likely to make the pollsters look foolish.