- The Legal Workshop - http://legalworkshop.org -
Resolving Election Error: The Dynamic Assessment of Materiality
Posted By Justin Levitt On October 30, 2012 @ 6:18 pm In Constitutional Law, Due Process & Equal Protection, Law & Politics/Social Science, Law Review Article, William and Mary Law Review | No Comments
Error will drive the result of at least one election in 2012. This is not a difficult prediction: though the Bush v. Gore election brought the issue to the national spotlight, photo-finish and error-laden elections recur in each cycle. As a practical matter, the prospect of error is unavoidable. Election codes are enormously detailed. Every detail creates the possibility of mistake. And in the aggregate, all of these mistakes create the near certainty of an outcome-determinative mistake somewhere. Yet despite the fact that errors are inevitable, neither courts nor scholars have developed consistent principles for resolving the errors when they occur.
Below, and at more length in the Article on which this piece is based,1 I argue that the resolution of an election error should turn on its materiality. That is, deviations from procedure should generally preclude the counting of a vote only when the deviation is material to determining a voter’s eligibility or ballot preference. Moreover, I introduce the powerful insight that unlike materiality in other contexts, the materiality of a voting error can, and should, be reassessed over time. Though both powerful and simple, most jurisdictions do not currently employ this principle. A dynamic reassessment of materiality could transform the way that disputed elections are resolved, in 2012 and beyond.
Elections serve several functions. First, elections allow us to select responsive representatives according to the preferences of the represented community, as best we can ascertain them. Second, elections allow each eligible member of the community to participate in the representatives’ selection; this is why a sampled poll, however accurate, is an inadequate substitute for a community-wide vote. Elections as we know them also fulfill a third purpose: the comparative cost-efficiency of inviting electors to cast ballots centrally, rather than surveying them one by one.
Election procedures help to fulfill the three goals above. And when deviations from these procedures inevitably occur, a principle to address those errors should track the reasons for holding the election in the first instance. That is, the resolution should promote the most accurate assessment of eligible community members’ preferences, with the broadest participation by eligible members of the community, at the least cost.
Under these conditions, the most straightforward decision rules for resolving error are also the least adequate. Imagine, for example, that every procedural deviation were forgiven: any ballot would count no matter the context. With no means to enforce eligibility, election results would be wildly unreliable.
More often proposed, but just as flawed, is the converse: zero tolerance for error, with no ballot counted upon any departure from prescribed procedures. Note that “any” means “any”: a voter entering her name in cursive where block capitals are called for, or an election official failing to initial an absentee ballot form. Procedural violation, by any actor for any reason, leads uniformly to default. This sort of hard line fails for the same reason as its softer counterpart above: it does not inspire confidence in the election’s results. Given the number of regulations and the number of departures from those regulations that are inevitable in any large-scale election, a true zero-tolerance rule would exclude too many ballots of eligible electors to constitute a reliable representation of the community’s aggregate preferences.
The prevailing method in practice is to muddle along under a third approach. This approach embraces the intuition that some errors should not be “charged” to the voter in counting her ballot. Decision makers usually frame this approach in terms of “substantial compliance” with election regulation, distinguishing between “major” and “minor” errors.2 But little grounds such a rule beyond the preferences of the decision maker. “Minor” errors are not self-defining, nor is “substantial” compliance with a regulation. As a result, judicial decisions to disregard “minor” errors appear, in practice, dangerously ad hoc.
To the distinction between “major” errors and “minor” ones, some jurisdictions add a dollop of “fault.” Errors that are the voter’s fault become preclusive; those that are not are forgiven. When such assessments of blame occur, they are rudimentary, with little analysis of comparative contribution to the error. But even if it were possible to assign blame clearly, the frame of fault implies that voters should be punished for errors for which they are primarily responsible, by refusing to count their ballots. As I have explored elsewhere, such punishment serves little purpose.3 Voters who stray from election procedures rarely get enough feedback to avoid problems in the future. Particularly when deviations reflect mistake and not malfeasance, refusing to count an otherwise valid ballot offers little meaningful systemic benefit.
Rather than the magnitude of an error or the blameworthiness of a perpetrator, I suggest a different guidestar for resolving mistakes: materiality. Elections exist to discern the political preference of eligible members of the political community; the materiality inquiry takes its lead from the same mandate. Error should preclude the counting of a ballot only if it is material to determining a voter’s eligibility or ballot preference. When error is immaterial—when it leaves no reasonable decision maker with substantial doubt or uncertainty about the voter’s eligibility or preference—then it should not stand in the way.
Some of these decisions will be easy. A regulation may require a voter to state his date of birth in the form “XX/XX/XXXX”; writing “November” instead of “11” is an error, but an immaterial one, because it leaves no doubt about his eligibility.
Now consider a more complex scenario. Imagine a seventy-five-year-old elector, physically appearing before an official who has no doubt about her identity or other qualifications. She may err by offering the current date instead of her birthday on some prerequisite election form. Yet the unambiguous visual evidence renders that error immaterial in assessing her qualifications. Reliable evidence that a flaw is immaterial can come from sources other than the immediate vicinity of the flaw itself.
That realization leads to the insight at the heart of my proposal: the materiality of an election error can and should depend on information that becomes available to decision makers at different points in time. At the moment that an error is first evaluated, it is material if it creates real and reasonable doubt as to whether the individual in question is qualified. But that which is material at time t—say, an individual’s failure to affirm his citizenship on a voter registration form—may become immaterial at time t + 1 if it later becomes clear that the voter is a citizen. At the time when there is no longer a doubt about the individual’s qualifications, the earlier error becomes immaterial. I suggest that the original error should therefore not provide cause to deprive the individual newly known to be eligible of her valid vote.
Some applications of this concept, though not identified as such, are embedded in current practice. The easiest example is an election official’s discovery of a mistake. A data entry clerk may hit an errant key while typing information from a registration form; that error may lead to legitimate questions about the individual’s identity or residence. The instant that the data entry error is discovered as such, it becomes instantly immaterial, because the individual’s qualifications are no longer in doubt. The same is true if the applicant, not the election official, caused the error; once the error is discovered as an error, and accurate information is provided instead, the original flaw is no longer material.
The Help America Vote Act of 2002 (HAVA) provides a more sophisticated example. HAVA requires states to test the identity of new voters registering by mail.4 The statute contemplates that states will first attempt to match information on a voter registration form to information held in motor vehicle or Social Security systems. If the information matches, the applicant’s identity is confirmed. Errors, however, sometimes cause a mismatch.
At that moment, time t, with the error in the record undiscovered, the mismatch may be material to determining the voter’s qualifications. Yet under HAVA, the error is not preclusive. Instead, mismatched voters may show a piece of documentary identification, at time t + 1, up to and including at the polls. Once the voter provides her documentary identification, her identity is no longer in doubt. And because the original error in the mismatch, whatever its source, has become immaterial to determining her qualifications, it no longer interferes with the elector’s opportunity to cast a valid ballot.
Although both of the examples above are drawn from existing practice, the dynamic nature of the materiality decision has never before been articulated as such in the election context. Indeed, present standard operating procedure is to assess election errors only when they occur, ignoring the potential to reassess with more complete information. This is a lost opportunity.
In an election cycle, initial answers need not be final answers. Perhaps the voter contacts a registrar with curative information. Perhaps an error is uncorrected until Election Day, when the voter is able to submit a segregated ballot with curative information. Or perhaps she is summoned before a court overseeing a postelection proceeding. All of these represent opportunities to reassess the materiality of errors occurring earlier in the cycle.
Officials need not create these opportunities in order to utilize them as they arise. The materiality principle does not demand incremental procedures to seek information bearing on the validity of a vote. Instead, it merely changes the standard by which votes are evaluated when there is cause to undertake an evaluation, converting the focus of the inquiry from the severity of the initial error to the substantive qualifications or preferences of the individual in question. Errors that leave reasonable doubt about a voter’s eligibility or preference remain preclusive. Those that do not, yield.
There are several legitimate limits on the use of the materiality principle articulated above. In most circumstances, voters have no opportunistic incentive to violate election procedures relying on the chance that a breach will eventually be found immaterial. A voter will tend to follow the extant rules to the best of her ability, because it will always be easier to attempt to comply with the rules in the first instance than to try later to dispel the doubt that a material error creates.
In some cases, however, the nature of a regulation creates an incentive for strategic actors to “err” in order to gain a meaningful tactical advantage. Deadlines represent one such category. Just as homework and tax forms are regularly submitted at the last possible moment, voters act primarily on the brink of election deadlines. It is not difficult to imagine that a rule forgiving “immaterial” lapsed deadlines would lead to intentional tardiness. Where noncompliance offers strategic benefit, the materiality principle may be misplaced.
Errors in marking the ballot itself also present an unusual case. It is one thing to evaluate ballot-face errors for materiality using the evidence on the face of the ballot: that is, evaluating whether a reasonable decision maker could substantially question the voter’s intended choice. This static incarnation of the materiality principle presents little opportunity for strategic manipulation. Not so for dynamic reassessment. The temptation for a voter casting an ambiguous ballot to fabricate her prior intent in a postelection proceeding, or for a voter to vote in a strategically ambiguous fashion in order to preserve postelection options, presents a substantial opportunistic risk.
These are exceptional circumstances. Where there is no meaningful risk of opportunistic behavior, evaluating errors using a dynamic reassessment of materiality represents an improvement over alternative decision rules. And the value of the concept truly shines in a photo-finish election. When the margin of error exceeds the margin of victory, there are five basic options: (1) hold a new election; (2) discard all votes connected to any procedural violation, risking an inaccurate assessment of the electorate’s preferences and public dissatisfaction with the justice of a process replete with technical error; (3) count some votes related to a violation of established procedure, based perhaps on a subjective and flexible assessment of fault or magnitude—with the same risk of an inaccurate result and a process perceived as unjust; (4) count the votes of citizens known to be eligible when errors are, or have become, immaterial, in an effort to come as close as possible to an accurate understanding of the electorate’s preferences without any risk of diluting the electoral pool; or (5) acknowledge the futility of any attempt to accurately determine the plurality choice when an election is sufficiently close, and decide the victor by lot.5
In some ways, this last option presents the most serious challenge to the materiality principle. I acknowledge the substantial appeal in delivering to a random procedure those scenarios in which the margin of irreducible error exceeds the margin of victory. But I also suggest that once we have entrusted a selection process to a popular election, there exists a moral imperative to attempt to reduce apparent error whenever possible. Among comparatively equal-cost alternatives, the preferred option should be the one that yields the most accurate result. The materiality principle may be the least bad of the available alternatives.
The materiality principle is presently reflected in some aspects of election law, but should be adopted far more widely. As mentioned above, HAVA’s identification provision incorporates a dynamic assessment of materiality. And more broadly, the Civil Rights Act of 1964 contains a general materiality principle that most advocates and observers have overlooked.6 In the Article on which this piece is based, I review in much greater detail this materiality provision, which prohibits disenfranchisement based on certain immaterial errors. The scope of the Civil Rights Act’s provision is not coextensive with the materiality principle that I propose: the federal statute is limited, for example, to errors or omissions on records or papers involved in the voting process. But it is a welcome start.
Some state statutes also expressly reflect materiality principles, for some ballots in some circumstances. For example, several states, following a federal statute applying to overseas voters,7 have adopted the materiality principle with respect to voters’ misspellings of the name of a write-in candidate.8 And at least eight states have enacted broader materiality provisions, beyond the write-in context.9 Others should follow.
* * *
Despite concern about the aggregate decline of competitive elections, photo-finish races will not disappear. When an election heads into “overtime,” it tests our capacity to deliver a result that is both inclusive and reliable. In 2000, that capacity broke down in the face of numerous errors by voters and officials, with serious and lingering consequences. Nearly a decade later, our conceptual approach to election-related errors remains inadequate.
Above, I suggest one component of a solution, dependent upon the robust application of a basic materiality principle. Particularly by recognizing the dynamic nature of materiality, it is possible to resolve election-related flaws in a manner better tailored both to the rationale for election procedures and to the more foundational reasons why we vote.
Justin Levitt is an Associate Professor of Law, Loyola Law School, Los Angeles.
A version of this article will appear in the October 2012 issue of the William & Mary Law Review.
Article printed from The Legal Workshop: http://legalworkshop.org
URL to article: http://legalworkshop.org/2012/10/30/resolving-election-error-the-dynamic-assessment-of-materiality
Copyright © 2009 The Legal Workshop. All rights reserved.