Recent news accounts confirm the need to scrutinize the process by which provisional ballots are evaluated for eligibility, in order to assess the extent to which the Help America Vote Act (HAVA) is working as Congress intended.
Collectively, these news accounts indicate that there are three main concerns about the way in which provisional ballots are being evaluated to see if they should count as valid votes:
first, provisional ballots are being classified as ineligible if the individual is not currently listed as a registered voter, but it is unclear whether election officials are checking previous registration lists to determine if the individual was "purged" or otherwise removed from the rolls;
second, it is unclear how election officials are treating provisional ballots where individuals submitted incomplete registration forms, and whether this category of provisional ballots is being handled consistently among different counties within the same state – are they being classified as eligible as long as the would-be voters supplied the lacking information at the time they submitted their provisional ballots, or instead are they being rejected as ineligible on the ground that their incomplete forms prevented them from achieving the status of registered voters?;
third, some provisional ballots are being classified as ineligible because the signature submitted with the provisional ballot does not match the corresponding signature that the county has on file for this registered voter, but it is unclear what standards (if any) county officials are using to engage in this handwriting analysis or whether they are finding any of these signature comparisons to be "close" or "questionable" calls.
1. The "purge" problem. The first of these concerns is important because it goes to the heart of why Congress adopted the new provisional ballot rules in HAVA: in 2000, particularly in Florida, individuals had been turned away from the polls without the ability to cast a ballot because they had been improperly purged from current voter registration lists. Congress wanted to make sure that individuals would be able to cast a ballot even if their names did not appear on current registration lists, and Congress wanted those provisional ballots to count as valid votes if, upon subsequent examination, officials determined that these individuals previously had been registered but had been erroneously removed from the rolls. Therefore, to comply with HAVA, it is necessary that the evaluation of provisional ballots encompass an examination, not just of current registration lists, but also previous registration lists (including especially "pre-purge" lists). Yet it is unclear from recent news accounts whether this core component of HAVA compliance is currently occurring.
2. Incomplete registration forms. The second concern is significant because, prior to November 2, a large number of provisional ballots – in Florida, Ohio, and other swing states around the country – was expected to be individuals who attempted to register for the first time this year, but whose attempts were unsuccessful according to state officials because their registration forms were incomplete. In Florida, for example, this issue surfaced in the context of registration forms for which the "citizenship box" was not checked, while in Ohio it arose over the failure to include a driver's license number, or last four digits of a Social Security number, on the registration form. Lawsuits were filed before Election Day in an effort to require state officials to place these individuals on the registration rolls, so that they would be eligible to cast regular rather than provisional ballots. These lawsuits were unsuccessful, leaving for later the question whether provisional ballots cast by these unregistered individuals would be considered eligible for counting as valid votes. At the time of these lawsuits, it was estimated that in both Florida and Ohio this category of individuals numbered in the 10,000s.
Now that the provisional ballots in these and other states are in the process of being evaluated for eligibility, it is unclear how many of them actually fall into this category and whether they are being accepted or rejected as valid votes, or even if there is consistency in their treatment from county to county within the same state. A major unresolved question concerning the interpretation of HAVA is whether or not Congress intended to require states to count provisional ballots as eligible votes when the only obstacle is an incomplete registration form and the missing information can be verified at the time the provisional ballot is evaluated. Resolution of this question, and assessing whether HAVA needs to be amended in this regard (and, if so, in what way) requires a better factual understanding of this category of provisional ballots and how they are being handled this year.
3. Signature comparisons. The third concern merits attention for reasons relating to Bush v. Gore. The constitutional defect identified in that case was the lack of specific standards for evaluating chads to see if they should count as votes. The absence of specific standards caused officials to make subjective judgments, ballot by ballot, with the consequence being differential treatment of ballots from county to county, or even by different officials within the same county.
Regrettably, the examination of a signature submitted with a provisional ballot, to see if it matches a corresponding signature on file at the local board of elections, seems like the same sort of standardless, subjective judgment. HAVA provides no specific standards for making this handwriting comparison judgment, and neither does state law (certainly not in Ohio, where some number of provisional ballots are reportedly being rejected on this basis). Perhaps the standardless, subjective examination of signatures can be considered distinguishable under the Equal Protection Clause from the standardless, subjective examination of chads – the argument perhaps being that the former is unavoidable, while the latter is not (although handwriting analysis experts might disagree). In any event, this legal issue warrants further investigation, as this lurking Equal Protection question invites litigation in a situation where the eligibility of these provisional ballots might make a difference in the outcome of an election.
In sum, we need to know more about the processes by which the evaluation of provisional ballots is taking place. HAVA is a new law, being tested for the first time. The nation needs to assess its effectiveness.
A provisional ballot, as Congress envisioned when enacting HAVA, was supposed to function as a kind of voter's insurance policy: an eligible voter gets to cast a conditional, back-stop ballot that will be counted if, despite a glitch in the system that causes the poll workers to think the voter ineligible, it is subsequently determined that the voter is indeed eligible. But if provisional ballots are not working as intended – if instead they provide a kind of false security because whatever glitch caused the pollworkers to think the voter ineligible won't be fixable in a few short days after the election – then it would seem necessary to rethink the desirability of provisional ballots as a feature of our electoral system.
It is way too early to reach a verdict on how well this new system of provisional ballots is working. But it is not to early to investigate this issue. In fact, soon it may be too late to gather useful information about the functioning of the process: as boards of elections complete their procedures for evaluating the eligibility of provisional ballots, it will become impossible for the press to observe and report on this process contemporaneously. Once complete, all subsequent investigation will be after-the-fact.
For this reason, it is to be hoped that the press will continue the kind of reporting highlighted at the outset of this commentary.