Sunday, February 13, 2005

Juan Cole's Hare chase

Juan Cole continues to be confused about the electoral system being used in Iraq. First, back in November, he argued that it was based on Israel's system, imposed by U.S. neocons:

the Americans, especially the neoconservatives, crafted a ridiculous electoral system based on that of Israel.

Israel does have a List PR system, like 35% of the rest of the world, but 1) this system was one of three proposed by the United Nations to the Interim Governing Council, which approved it by a vote of 21-4 -- and for the life of me I can't figure out how U.S. neocons manipulated that vote; and 2) the system for counting parties, using the Hare quota, is actually more fair to smaller parties than the D'Hondt system used in Israel.

Today, he tries to explain this system, calling upon the services of Andrew Arato:

AFP is convinced that the UIA may all by itself be given 140 seats, not 133, because of a "complex counting system" to be employed in seating delegates. Andrew Arato writes, "This would be so, because wasted votes for very small parties . . . would have to redivided. Say it is 8%. 48% of that is 4%. Even of 5 half is 2.5% that would put them over."

Clear as mud, right? If you're like me -- neither a Middle East expert nor a political scientist -- you might want to probe a bit more deeply than an AFP report cited by Cole, and try googling around for better information. For example, you'll find a far clearer explanation from the United Nations website about the process:

How many votes are needed to win a seat?

  • The number of votes required to gain a seat (the natural threshold) will be determined by the number of total valid votes cast . A maximum estimation of 14,270,000 valid votes would create an initial threshold of 51,891 votes to gain a seat; 10 million votes would require 36,363 votes; and, 5 million votes would require 18,181 votes.
  • The chosen electoral formula (the Hare formula) proceeds after the natural threshold calculation based on the “largest remainder”. In effect, this means that subsequent seats (for lists that pass the natural threshold) cost fewer votes.

The variety of electoral systems is discussed at great length over at wikipedia, but here's a quick and easy definition of the Hare system from Encyclopedia Britanica:

(or single transferable vote), election system that gives minorities representation on elective bodies in proportion to votes received; voters indicate first, second, or other choices; a quota of votes necessary for election is fixed; if all seats are not filled, surplus votes of successful candidates and those of weakest candidates are distributed.

Complex, yes.

Impossible to understand, no.

Cooked up by the Neocons, doubt it.

UPDATE, Monday morning: Juan Cole now admits that he doesn't understand the electoral system, and ends his post with more gobbledygook from Arato (who seems to be a smart fellow when it comes to other subjects). In an earlier post picked up by my RSS reader, Cole even contested whether the Hare system was being used -- because all press reports only talked about voting for one list. But now he's dropped that assertion, and leaves the reader with two of the wikipedia references I've noted for further reference.

If he hadn't backed off, I would have mentioned that only under a Hare system could the votes for the UIA jumped from 133 to 140. Otherwise, they would have picked up only one more vote at the most, under a simpler largest remainder system (like the one used in El Salvador).

Mea Culpa, Feb. 26: In response to Andrew Arato's comment here, I checked in with an academic friend who's an expert on electoral systems. Seems I was confused, but he also managed to explain what was going on here more clearly than either of Cole's posts:

The overrep of larger parties here does not have to do with Hare vs. D'hondt (or any other divisor system). Rather, the Iraqi rule holds that in order to be eligible to win any seats, a list must clear a threshold of one full Hare quota first (Total votes/275). That is, you can't win remainders seats without getting at least one by full quota. So all the votes won by all the lists that didn't get at least a quota (99 lists, I think) are wasted. The beneficiaries of that, of course, are the parties that cleared the TV/275 threshold -- that's where the seat bonus comes from.
At least I'm glad to see Arato didn't challenge my citation of Cole's ridiculous charge months ago that led off this post.


praktike said...

He needs an editor.

Eric Umansky said...

Well, one, I agree with Praktike. On another note, David: I'd like to talk with you about a story I'm doing. (And I couldn't find an email address for you.) Could you email me at thanks, eirc

Andrew Arato said...

Whoever you are, you are very confused. The Iraqi's voted list PR. They had only one vote. In the single transf. vote they would have had second, third etc. choices.

The crucial difference is between the Hare system and the Hare quota dealing with remainders in the PR... i.e. votes for parties that do not reach the natural threshold, and the extra votes for parties that do.

As you may have noticed, i calculated the exact number of seats the UIA got almost from the beginning. How did I do it? with the simplified calculation that was close to what the formula would produce. Obviously you paid no attention to waht I said.

I divided the total by 275. Got the natural threshold. Then estimated the % lost as remainders, and redivided it according to the original percentage that is not really correct, but is correct for the largest party --- the issue here.

So waht was wrong with this....all you say it is confusing, which it is not....