Even though we thought we had settled the question, maybe not.

An acquaintance who voted out-of-country in the election of Baghdad’s Council of Representatives delegation (M=68, he said) told me tonight that he had cast his vote for a list. He did not use his option to cast any preference vote. He also said that, in order to cast a preference vote, one first had to vote for a list. Then one could express a candidate preference using information from a directory of candidates-by-party. (I do not know any more about how that worked because I do not read Arabic, but Pauline at FairVote offers a specimen ballot.) All this suggests a flexible-list system like the one we assumed had applied to the governorate (i.e. provincial) elections in January 2009. The question then becomes (once again) what quota of preference votes a candidate needs for his/her position on his/her party’s list to change.

If only it were that simple. My acquaintance moreover told me that one had the option to rank up to all of the candidates on a party’s list. Having been an amateur STV junkie for many years, I can imagine ways in which ranking would factor into a flexible-list system. But what’s the use in speculating when the basic rules are so difficult to nail down? And here I thought I had painstakingly tracked down the important details.

What can be said of all this uncertainty? Assuming my acquaintance spoke truth…

First, I am conditionally skeptical of cross-national work treating institutional details like these as variables. I had put a lot of time into once-and-for-all answering the open/flexible question, drawing on resources unavailable even to most academics, and here came a real-life Iraqi voter with reasons for doubt. If my acquaintance spoke truth, I can only wonder about the extent to which measurement error (i.e. incorrectly coding a polity’s electoral rules) makes suspect what we think we have learned about institutions’ causes and effects.

Second, I am conditionally convinced of the unreliability of ostensibly reliable sources. The one I edit may not be exempt from this charge, despite my obsessive propensity - as my beleaguered assistant can attest - for filling in every missing value. The same goes for information put online (yet no longer available in some cases) by the United Nations Assistance Mission in Iraq and the Independent High Electoral Commission… assuming, of course, that my acquaintance spoke truth.

Third, my belief in the investigative efficiency of talking to people is reinforced. Where Google searches yield quantity, a few personal conversations yield quality. Conversations allow one to ask for clarification and to pose follow-up questions. Assuming my acquaintance spoke truth, I learned more about the present, national Iraqi electoral system in five minutes than I did in days of scouring the Internet.

On a more technical note, my acquaintance said he was disappointed that the system had not provided for panachage.