Not Perfect, But “Reasonably Credible”

(an edited version of this essay was published in

the Wall Street Journal, October 25, 2004)


By Larry Diamond


It is one of many paradoxes facing Iraq today that national elections have become more urgently important as they have also become more difficult to imagine.  The interim government of Prime Minister Ayad Allawi is tough and capable.  But no government appointed by the U.S. and the U.N. will have the authority to take the difficult and risky steps—political and military—necessary to defuse and defeat the growing insurgency.  Only a government chosen by the people of Iraq, and seen to reflect their diverse interests, can rally Iraqis to fight effectively to restore order to their country.  This is why President Bush and Prime Minister Allawi were right last month to reaffirm their commitment to holding elections for a transitional parliament (which will also act as a constitutional assembly) by the internationally established deadline of January 31, 2004. 

Moving ahead rapidly with elections in Iraq carries risks.  Ill-timed and ill-prepared elections have repeatedly undermined post-conflict transitions and have even led to renewed fighting.  But the risks of a lengthy postponement of elections are even greater.  It is mainly the prospect of voting in January—delayed though that is from what many Iraqis demanded—that has bought reluctant acceptance of the appointed interim government by a majority of the population. Unless the legitimacy vacuum is filled soon, Iraq could slide into anarchy.

If the political transition is to move forward, Iraq must have elections soon, but it does not need—and cannot attain—perfect elections.  Iraq needs what U.N. special representative Lakhdar Brahimi called “reasonably credible elections.”  Inevitably, there will be at least localized violence and efforts to rig the voter and intimidate the voters.  In some parts of Iraq—Fallujah, Ramadi, and perhaps all of Al-Anbar province—it may not be possible to hold elections at all, even if a new counter-insurgency campaign is launched soon.  But most of Iraq must vote, and most Iraqis who vote must view the process as reasonably open and fair.

            If meaningful elections are to be held by January, the Iraqi interim government and its international supporters must urgently take several key steps.  They must rapidly develop a much stronger Iraqi capacity to provide order and security in the campaign and the polling.  They must improve the administrative capacity and reach of the Independent Electoral Commission of Iraq.  And they must train many thousands of Iraqi civil society activists to monitor the voting and vote counting.  These will be formidable tasks.

            One particularly urgent imperative, however, can be easily achieved: the electoral system must be changed.  The current plan is for Iraqis to vote for party lists through proportional representation in a single, nationwide district.  While this system is the easiest and simplest to administer, it has several crucial drawbacks.  Most crucially, it provides no means for deferring voting in some areas without disenfranchising them for the entire term of the transitional government.  If there is only a single nationwide district, then all the seats in parliament will be distributed according to the votes that different national lists obtain on election day.  Once each party or coalition gets a share of seats equal to its share of the national vote, there will be no way to award seats later to other districts when they are able to vote.  Neither will there be any mechanism for filling those seats temporarily.  Thus, if elections were to move forward without extensive voting in the Sunni triangle, Sunni Arabs would be cut out of the process of governing and constitution-making for the duration of the transition, leaving them little option but violent resistance.

            Allowing for elections to be held in different parts of the country at different times requires that Iraq be divided up into electoral districts of some kind.  There is not the time or the technical data to divide Iraq into 275 single-member districts.  But the country is already divided into 18 provinces, and existing census and food-ration data provide at least a rough estimate of each province’s share of the national population.  Each province (save Baghdad, which would have to be split up) could be established as a separate multi-member district.  Within each district, seats would be distributed to various lists, including independents, in proportion to their shares of the vote.

            The principal objection to this system—which figured prominently in the UN’s decision to recommend the national list system—is the lack of broadly accepted census data, which could leave some Iraqi communities (for example, Shiites and Kurds) aggrieved if they felt their provinces were shortchanged in seats.  However, it is possible to have a system of multimember districts in which some number of seats is held back to correct for imbalances arising from the way votes are allocated into parliamentary seats in the districts. In such a two-tier system, 80 percent of the seats (220) could be assigned to the multimember districts, and the remaining 20 percent (55 seats) would then be filled from national lists so that each list’s final share of parliamentary seats matched as closely as possible its share of the national vote.  This would not purely correct for unfairness in provincial seat allocations, but it would ensure that parties popular in a region got the share of seats equivalent to their overall national vote—and that is mainly what people care about.

When I was advising the Coalition Provisional Authority in Baghdad earlier this year, some Iraqi and international specialists (including myself) proposed precisely such an electoral system for Iraq.  Using available food ration data, we estimated that seventeen provinces would each have between 5 and 21 seats (if 220 seats were distributed to districts) and Baghdad would have 52.  We preferred to see the bigger districts, especially Basra (16 seats) and Nineveh (21) broken in two, but at this late date administrative simplicity is the more urgent imperative, and even a list of 21 candidates is a lot easier for voters to review than a list of 100—or 275.  If they vote in multimember districts, Iraqis will be able to pay attention to who, not just what, they are voting for.  Independents (who tend to be more moderate) will fare better, and voters will also be better able to contact representatives, review their performance, and hold them accountable.  If there is only nationwide voting, religion, ethnicity, and ideology will trump all other considerations.

            A system of multimember districts would allow voting to be suspended in one district without threatening the viability or legitimacy of the entire election.  The vacant seats could be held open until voting conditions permitted voting in the district, and the seats could even be filled temporarily through indirect election by as representative a set of social and political forces as could be identified. 

            By any calculus, bringing off timely democratic elections in Iraq is going to be formidably difficult.  Nationally, truly free and fair elections are unattainable any time soon.  But “reasonably credible” elections can still be accomplished, if Iraqis and their international supporters put in place now the best possible designs and preparations.


(Larry Diamond is a senior fellow at the Hoover Institution and a former senior advisor to the Coalition Provisional Authority in Baghdad).