Update (2/7/08): It now appears that some 94,000 ballots cast in the American Independent Party primary and the Democratic primary may be inseparably commingled, making it impossible to go back and determine how many respective ballots should be counted for Clinton and Obama. Some other states deal with commingled ballots by applying various mathematical techniques of "apportionment," e.g., dividing the votes between the candidates according to the percentages they obtained in some agreed-upon sample pool. However, EL@M was unable to find any California case approving of this technique, so it is unclear whether these votes will count at all.
Flawed ballot design could result in the rejection of the votes of many L.A. County independent voters. Independent voters are given a special ballot with a bubble they have to fill in to indicate whether they intend the ballot to count in the Democratic or American Independent Party primary. If voters fail to fill in the bubble, their votes will not count, and some fear this will lead to disenfranchisement. The Obama camp has already threatened litigation.
The magnitude of this problem is difficult to determine, but it could affect thousands, if not tens of thousands, of votes. L.A. County has about 800,000 independent voters, and a high turnout is likely. Few choices for unaffiliated voters, Long Beach Press-Telegram, January 14, 2008. In the 2000 Presidential primary, for instance, turnout exceeded 21 percent (data for 2004 could not be located). Victories for McCain in lesser states not enough to stop Bush, Irish Times, March 8, 2000. California’s moved-up primary date will probably increase that number. Nevertheless, not all independent voters will be affected, even potentially. The Republican primary ballots are apparently fine and, while the Democratic ballots may be confusing, they will not confuse every voter.
Despite these mitigating factors, there is a substantial risk that one of the Democratic candidates will request a recount or file a lawsuit claiming that, but for the confusing ballot design, they would have fared better in L.A. County. Even where a candidate loses the state by a wide margin, there may be an incentive to do this. This is because California allocates the bulk of its delegates by congressional district, not by the at-large vote. L.A. County contains at least parts of 18 of those districts (see here), and litigation in any one of them could be a plausible way to pick up a few delegates.
The situation gets still more complicated because delegates in each district are not awarded on a winner-take-all basis, but instead are divided proportionally between all candidates who obtain more than 15% of the vote in the disputed district. This means that plaintiffs in an election lawsuit need not show they deserve a majority of the vote, but only have to show that the vote total should be increased just enough to tip one more delegate in their favor. This is a much more attainable goal than persuading a court that one obtained a traditional majority of votes.
Still, a candidate contesting the count in any of these districts would be up against significant challenges. One challenge would be to persuade the court not to get rid of the case by “blaming the victim.” In election contests based on confusing ballots, courts often say that the complaining candidate has an obligation to spot instances of poor ballot design and sue to correct them ahead of time; candidates who shirk this duty may be barred from later claiming that the design cost them the election. See, e.g., Kilbourne v. City of Carpinteria 56 Cal.App.3d 11, 17 (Cal.App. 1976). A second, but perhaps lesser, challenge would be to determine for whom the rejected votes were cast. From news accounts it seems like a court could do this simply by examining the rejected ballots and seeing whether voters filled in ovals for either of the Democratic candidates. Forensically, this would be a much easier task than determining the intent behind the traditional over- and under-votes that form the basis of similar lawsuits.
Contributed by Nathan Cemenska, Web Editor.