The Iraqi election results were released yesterday, more than a month after the election itself, and with early objections from Sunni groups unaddressed (at least publicly). While the Shia bloc did not obtain an outright majority, and will therefore have to govern in a coalition government or poach Assemblymembers from other parties, the Prime Minister will almost surely be a religiously leaning Shiite. Sunni slates won 55 seats; Kurdish parties 58. Two semi-surprises: former Prime Minister Iyad Allawi's secular coalition won only 25 seats. Allawi, you will remember, was the most clearly American-backed candidate.

The second semi-surprise is that Ahmed Chalabi, whom some administration figures proposed as the "strongman" Iraq needed after the US invasion, and his list received exactly ZERO seats in the new Iraqi assembly.

The New York Times has one of their neat interactive graphics here.

Why confusing?

To me, the most confusing part of this election is not the outcome, but rather the electoral regime itself. Poli Sci geeks know that one of the biggest problems with proportional representation is what to do with results that do not neatly round off, and what to do with the votes that went to very small parties that did not meet the threshold to enter parliament.

For instance, say you had 100 seats in the Assembly, and had a minimum threshold cutoff of 5 percent.

(This threshold was introduced in Germany after World War 2 to prevent the sorts of splintered parliaments that dominated the Weimar Republic. This requirement has been adopted most places to create centripetal forces in the election regime.)

If three parties each received 4 percent of the vote, together they would have received 12 seats in our hypothetical assembly. Separately, they received none. These extra, unallocated seats are called "hanging seats." What do you do with the seats that would have gone to those small parties? Usually, they are distributed among the larger parties according to some simple algorithm (i.e. the proportions they got). It's not too difficult; Europeans, at least, don't seem to mind this slightly undemocratic element of their systems.

However, in Iraq, the electoral regime compounds this problem by recreating it in each province. Instead of having national elections, and one reallocation of hanging seats, each province has its own proportional system and own reallocation.

To fix the quirks in this system, the Iraqi constitution invents "compensatory seats"-- seats reallocated by the Election Authority to parties that for one reason or another didn't receive seats approximating their electoral support.

In addition, to give the Assembly a national flavor and some stability, the constitution grants extra "national seats" based on Iraq-wide outcomes.

I've done some rough tabulations based on the New York Times and Washington Post vote totals, and compared the number of seats the parties got under the provincial system, and what they would have received under a purely national system. The biggest differences: the main Shia party received nine more seats that it would have; the main Kurdish party received eight fewer. Strangely, two parties that would have received no seats received one each.

So to sum up:

1. 18 provincial party lists, rather than one national one.
2. Hanging seats in each province, to be reallocated within each province.
3. Compensatory seats for parties that, through the quirks of this system, did not receive seats that matched their vote totals (a process that seems at great risk for corruption and manipulation).
4. A smaller number of "national seats" allocated on national results.

Results:

1. Smaller parties received representation greater than they would have.
2. Shia received more seats.
3. Kurds received fewer.

I have yet to read any explanation for this confusing system. I also am unaware of any regime that uses provincial proportional representation and lists to determine the makeup of a national assembly. It is obviously the product of the intense negotiations among the political parties in 2004. To me, the system privileges the political parties, which control the reallocation mechanisms and can therefore negotiate political power among themselves. What remains to be seen is whether the stability the parties bring to the system is worth the trade-off of democratic legitimacy. Why the US would allow a system that promoted the Iranian-backed Shia at the expense of the US-backed Kurds is another matter.