Guest Post: Metrics for Election Success
31 January 2005, 10:41 AM
A Gentle Reader who wishes to remain anonymous has contributed the following note. My own response will be posted, as is appropriate, in the comments. Here, I would just like to publicly thank the writer for contributing, and encourage any similarly inclined Gentle Reader to go ahead, anonymous or no.
How do you determine if an election is a success? You need both a metric and a way to evaluate the results.
In the United States, a successful election is defined as an election in which (1) candidates were allowed to communicate with potential voters, (2) those who wanted to vote were able to vote without unfair or undue obstacles, and (3) the vote count accurately reflects the intent of those who chose to vote. Federal and state election laws (and court decisions) exist primarily to ensure those three criteria are met, and the public generally accepts them. We do not measure the success of an election qua election on the basis of how many candidates there were, or how many viewpoints those candidates represented and how well those viewpoints corresponded to the viewpoints of potential voters, or how well candidates communicated with potential voters, or how many people voted, or how many people made positive choices rather than negative choices, or how informed those people's choices were, or how many people are satisfied with the outcome of the election. These additional concerns might be important to rendering the phrase "representative government" meaningful, but we accept the idea that market forces will take care of most of these additional concerns. We say that an election was successful even if there was only one candidate, even if that candidate didn't campaign, even if nobody really liked the only candidate, and even if most eligible voters didn't turn out to vote.
That's the metric. The evaluation method in the United States is correspondence with expected results as determined by polls, verified by media scrutiny and challenges in appropriate courts.
So what should be the metric and evaluation methods for Iraq?
The Bush administration would define the election as successful if there are multiple candidates/parties, ballots, and voters. It is hard for the Iraqi election to fail against such a metric, so there is no need for an evaluation method. While that's consistent with the administration's anti-scientific approach to the world, we might wish for other metrics and functional evaluation methods.
We must accept fairly low standards for a successful election in a country dealing with a violent foreign occupation, a brewing civil war, a long history of violently non-democratic minority rule, and tremendous obstacles to both campaigning and voting. But surely we could establish some rudimentary metric, made substantive by the results being meaningfully evaluated.
A modest proposal for a successful Iraqi election: there should be candidates who have different points of view, those points of view should reflect the varying points of view held by large numbers of Iraqis, those points of view should be communicated in attempts both to inform and persuade, voters should have some reason why they are choosing one candidate over another (i.e., there should be some voter intent), many of those who are eligible to vote should want to vote, and those who want to vote should be able to vote without unfair or undue obstacles.
The rate of voter participation is important because it demonstrates that there is some belief that the election is free, fair, and meaningful both in terms of expressing voter intent and in affecting the governance of the country. It is also a metric which can easily be evaluated, and at least among Iraqi expatriates, the rate of voter participation is appallingly low. News coverage from Iraq suggests that candidates were afraid to campaign (or even announce who they were), and that obstacles to voting were enormous and uneven across both geography and ethnicity. On the positive side, there were multiple candidates/parties which did vary and reflect varying points of view held by large numbers of Iraqis.
Faced with these realities, the world press is largely evaluating the success of the election by whether the proportions of the Iraqi vote corresponded to the proportions of the Iraqi population along geographic and ethnic lines. This evaluation is only appropriate if you assume that the election was intended to be a form of census. (Worse, the evaluation is only meaningful if you already have a reliable census to compare against, in which case there was no particular reason to hold an election which was both bloody and expensive.) That metric is hardly more appropriate than that put forth by the White House, though it at least allows the election to be called either a success or a failure.
Perhaps we should measure the success of the election by whether the machinery (both physical and societal) for holding elections was at least partially put in place. The new Iraqi government is not likely to be substantially less representative than the previous several, nor substantially more effective in maintaining civil calm. But someday peace may come to Iraq, and because of this election-like exercise the Iraqi people may have an easier time holding a free, fair, and meaningful election. Then it would be right to call the January 2005 exercise a success, even if it could not truly be an election. For the sake of giving purpose to the sacrifices made so this exercise could be held, I hope that turns out to be the case.