Chief Justice Roberts has been attacked from both the left and the right for failing to admit openly that his opinion in last week’s campaign finance case amounts to overruling the Court’s 2003 decision that involved the same McCain-Feingold provision. This comment, however, defends the new Chief Justice’s approach, given (1) a gap left open by the reasoning of the 2003 decision, and (2) the Chief’s basic orientation towards funding restrictions for campaign messages, which he was entitled to invoke to fill that gap.

The next comment in this space is scheduled for September 4.

If it is possible for conventional wisdom to develop within a single week, then the conventional wisdom about FEC v. WRTL (the Supreme Court’s “blockbuster campaign finance decision)” is that the new Chief Justice’s bluff got called by his colleagues. Chief Justice Roberts said he was not overruling one of the two major rulings in McConnell v. FEC, the 2003 landmark decision upholding the constitutionality of the McCain-Feingold reform law. But seven other Justices said he was, without fessing up to doing so, and the eighth—mild-mannered Justice Alito, who was the only one to join the Chief’s approach to the case—came fairly close to acknowledging as much.

What has attracted the most attention is Justice Scalia’s attack on the Chief’s avowal against overruling. “[I]ndefensible,” is how Justice Scalia, joined by Justice Kennedy and Thomas, described it. The well-settled relevant test for facial invalidation of a statute asks whether its overall scope is substantially overbroad in relation to its range of permissible applications. Because McConnell had rejected a claim of facial invalidity with respect to the particular provision of the McCain-Feingold law at issue (its prohibition on the use of corporate or union funds to pay for “electioneering communications”), it must mean that this provision was not substantially overbroad in its scope. Yet, as Justice Scalia observed, the position propounded by Chief Justice Roberts in the new case inevitably would render the applicable McCain-Feingold provision substantially overbroad. Therefore, the Chief’s approach “effectively overrules McConnell without saying so,” observed Justices Scalia, noting that six other Justices also saw the same incompatibility between McConnell and the Chief’s opinion.

Furthermore, Justice Scalia lashed out at what Chief Justice Roberts might have thought was the justification for holding back on overruling, even though he must have recognized the same incompatibility that all his other colleagues (including Justice Alito) did. Courts shouldn’t overrule precedents unless and until their validity is necessarily and squarely called into question, and a precedent about a statute’s facial invalidity is not directly at issue in a new case, like WRTL, that merely involves the constitutionality of the statute as applied to one particular set of facts. Justice Scalia, however, would not accept the invocation of that principle here. “This faux judicial restraint is judicial obfuscation,” he jabbed, and even folks who normally disagree with Justice Scalia have embraced his view on this point. Most prominently, Walter Dellinger has offered: “[I]t’s neither minimalist nor restrained to overrule cases while pretending that you are not.”

Yet in this debate I wish to take the Chief Justice’s side. In the area of campaign finance specifically, and most likely in constitutional law generally, it is better for the Court to leave precedents standing, unless it is absolutely impossible to do so.

McConnell’s Vulnerability

In siding with Chief Justice Roberts, I hasten to acknowledge that I agree with Justice Scalia’s incompatibility point. Moreover, I would not have joined the Chief Justice’s opinion. My own views on campaign finance law are closer to Justice Souter’s dissent than they are to the Chief’s opinion.

Still, the relevant question concerning the debate between the Chief and Justice Scalia, it seems to me, is whether for someone with the Chief Justice’s views the refusal to explicitly overrule McConnell was a responsible treatment of precedent. Although I think he could have—and should have—allowed himself to be guided by some straightforward inferences from McConnell, his unwillingness to do so is not necessarily a dishonest treatment of that precedent. Weaknesses in the reasoning of McConnell itself left it exposed to the way Chief Justice Roberts handled it. Indeed, it is quite possible that Justice O’Connor—had she still been on the Court in WRTL—might have joined Chief Justice Roberts’s opinion even though she had been a co-author of the McConnell majority.

McConnell, as mentioned, involved a challenge to the McCain-Feingold law as written. In rejecting that challenge, the McConnell Court could have ruled the statute valid in all its applications. The McConnell Court would have been justified in doing so: the combined need for both clarity in the statute’s scope and the avoidance of corporate or union spending on electioneering messages warranted enforcement to all messages that came within the statute’s bright-line coverage. As I had explained in a law journal article, “Narrow Tailoring” is Not the Opposite of “Overbreadth”: Defending BRCA’s Definition of “Electioneering Communication,” 2 Election L. J. 457 (2003), the Court’s pre-McConnell precedents supported this kind of across-the-board validation of the statute.

But the McConnell Court did not take this approach. Instead, McConnell left open the possibility that the statute could be ruled invalid as applied to particular cases. Moreover, in an earlier opinion in the same WRTL case, the Court unanimously confirmed that the statute was susceptible to as-applied challenges. As a result, when the case returned to the Court in the proceeding that yielded the current decision, it was the task for the Justices to decide in what circumstances the application of the statute would be invalid.

In confronting this task, the Court had essentially two alternatives: either the application of the law would be invalid only where no reasonable person would consider the particular message to advocate for or against a candidate’s election; or, flipping the burden, the law would be invalid unless no reasonable person would deny that the message advocates for or against a candidate’s election. For reasons I have explained previously, the former alternative is more in keeping with the spirit of McConnell. Nonetheless, McConnell did not specifically purport to foreclose the latter alternative, thus leaving it vulnerable to the narrow reading that the Chief Justice gave it.

What McConnell actually decided was that the constitutionally protected Freedom of Speech did not confine Congress’s prohibition on corporate or union funding of electioneering to an exceedingly narrow category of messages that contained certain “magic words” of electoral advocacy. (“Nor are we persuaded [based on precedent or principle] that the First Amendment erects a rigid barrier between . . . the presence or absence of magic words” is how the McConnell Court stated its ruling.) The McConnell Court, however, did not say how much farther beyond “magic words” Congress could go in imposing its prohibition on corporate or union spending on messages that might qualify as electioneering. Instead, the McConnell Court took the tack of saying only that the McCain-Feingold law as written was not entirely unenforceable, based on a prediction that—wherever that constitutional line would ultimately be drawn—most of messages subject to the McCain-Feingold law would fall on the unprotected side of the line (unprotected meaning, again, that Congress could impose its funding restriction on these messages).

That predictive approach proved to be the Achilles heel of McConnell, since it did not fix the constitutional dividing line. What is more, McConnell’s rejection of the “magic words” limitation was not the only relevant precedent on this point. In First National Bank of Boston v. Bellotti, 435 U.S. 765 (1978), the Court had ruled that corporations could not be prohibited from funding messages that advocate for or against the adoption of a referendum or other ballot issue. Reconciling McConnell with Bellotti required the Court to explain where the scope of Bellotti’s protection ended and the scope of messages permissibly subject to a source-of-funding restriction began. (Were any messages referring to a candidate, rather than just a political issue, outside the scope of Bellotti’s protection? Presumably not, or otherwise the McCain-Feingold law would automatically be valid in all its applications, and thus the McConnell Court could have—and should have—easily said so.) Thus, there was a gap left open between precedents of McConnell and Bellotti that the Court need to fill in WRTL, and Chief Justice Roberts decided to fill it one way rather than the other.

The Gap Between McConnell and Bellotti

As I have indicated, I think it would have been better if the Chief had filled the gap in the other way, which would have been truer to the resonances of McConnell. But I cannot say that the Chief was “out of bounds” to fill it as he did. Bellotti has resonances of its own, as do other background precedents supportive of the Chief’s position (Buckley v. Valeo being perhaps the most obvious one). Moreover, there are general principles of Free Speech law that the Chief could draw upon, as he did, to help decide which way to fill the gap.

Ultimately, in confronting the empty space in controlling case law between McConnell and Bellotti, a Justice would need to develop a view on a basic underlying question in this area of campaign finance law: is a restriction on the use of corporate or union money to pay for a political message tantamount to barring a particular institutional speaker—the corporation or union—from uttering that message? Although I would say “no” in response to this question and could cite some precedents other than McConnell itself to support this view (most notably, Michigan Chamber of Commerce v. Austin, 494 U.S. 652 (1990), on which McConnell heavily relied), it is notorious that these precedents are poorly reasoned, sitting uneasily with earlier precedents (like Bellotti and Buckley) and not explaining adequately why (or the extent to which) corporations and unions may be treated differently than other institutional speakers. In this environment of confusing and potentially conflicting relevant case law, it is certainly understandable that the new Chief Justice would adopt an orientation that corporations and unions should presumptively have the same Free Speech rights as other institutional speakers.

And once he adopts this perspective, it makes sense that he would invoke basic pro-speaker doctrines of First Amendment law in deciding which way to close the gap between McConnell and Bellotti. Recall that one alternative Chief Justice Roberts had was to say that a corporation or union may not use their money to fund a political message unless no reasonable person would think that the message advocates for or against a candidate’s election. The other (and opposite) alternative was that corporations or unions may use their money to fund political messages unless no reasonable person would dispute that the message advocates for or against a candidate’s election. If under the First Amendment corporations and unions are to be treated presumptively as other institutional speakers, it is not surprising that Chief Justice Roberts would embrace the latter alternative, saying “we give the benefit of the doubt to speech, not censorship.”

Gap-Filling Without Overruling

Having filled the gap between McConnell and Bellotti in this way, Chief Justice Roberts was not irresponsible in refusing to overrule McConnell. Overruling a precedent is not compatible with the gap-filling exercise that the Chief Justice undertook. And gap-filling is all he purported to engage in—and appropriately so. Even if he could have (and I would have) filled the gap in a way “closer” to McConnell and “further” from Bellotti, he was entitled to sail closer to Bellotti’s shores. That he did so simply means that the predictive judgment on which McConnell’s facial validation of the statute rests remains open for potential reconsideration, if the accumulation of evidence as the statute gets enforced undercuts McConnell’s pre-enforcement prediction (a point that both he and Justice Alito make in somewhat different ways).

To be sure, there remains the incompatibility between the way Chief Justice Roberts filled the gap and the “substantial overbreadth” test for judging the facial validity of the McCain-Feingold law, which the McConnell Court had applied. Shouldn’t the Chief have acknowledged this incompatibility and therefore confessed to overruling McConnell, at least as a practical matter? I don’t think so, although his footnote that attempts to explain why his opinion is consistent with the facial invalidation in McConnell is the weakest part of the Chief’s opinion. Still, he was essentially on solid ground in saying that the predictive judgment on which that facial invalidation rested is a separate matter from the line-drawing exercise for as-applied challenges that he was required to undertake. Although it certainly now looks like that predictive judgment will prove inaccurate in light of the constitutional line the Chief drew, all of us will have to wait and see just how often applications of the statute fall on the wrong side of the line. In the meantime, the statute remains standing so as to test that predictive judgment.

Justice Scalia objects to this “wait-and-see approach,” observing that the very point of the “substantial overbreadth” test is to permit would-be speakers to knock out a statute in its entirety, without having to worry about the constitutionality of its applicability to their particular messages, when it is readily apparent that the statute as written has a substantial number of invalid applications in relation to its overall scope. Justice Scalia would surely be right were the facial validity of the McCain-Feingold law a fresh matter, not previously resolved by a judicial decision. In other words, had the constitutional line between permissible and impermissible applications already been set in the way that Chief Justice Roberts drew it, then it would have been necessary to knock out the McCain-Feingold law as written when McConnell was initially decided. But the reverse is not true. Once McConnell was decided as it was, it and the statute whose facial validity it upholds are entitled to stay on the books until the necessity arises for reconsidering the previously settled question of its facial validity. Since the task of drawing the constitutional line for purposes of as-applied challenges most certainly does not necessitate revisiting the issue of facial validity, but in fact was occasioned by what the previous ruling on that issue left opened, it is not judicially irresponsible to leave aside the facial validity precedent in the course of addressing the unresolved line-drawing question about as-applied challenges.

As for the inevitable tension that now exists between McConnell and WRTL, it is worth observing that there would have been just as much uncertainty in the applicable law—if not more—had Chief Justice Roberts taken the extra step and formally overruled McConnell. Either way, the same constitutional dividing line would exist (again, the one which separates messages susceptible to a prohibition on corporate or union spending from messages that are constitutionally immune from this kind of prohibition). If there proves to be difficulty applying this constitutional standard to particular messages—to the point where corporations and unions cannot be reasonably confident whether their messages would be protected or unprotected from this kind of regulation—then the Court, having overruled the facial validity holding of McConnell in WRTL, would be forced to confront whether to overrule much more than that: both the portion of McConnell ruling that “magic words” was not a constitutional standard and the earlier precedent (Austin) that a prohibition on corporate spending is constitutional permissible when “magic words” are present. Thus, there would remain plenty of tension in the relevant case law even if WRTL had overruled the facial validity holding of McConnell. Leaving the facial validity holding intact permits the Court to resolve in more modest ways the tension exposed in future cases: if it turns out that the Chief’s constitutional dividing line is unworkable, the Court in principle could resolve the tension by rethinking whether to sustain the McCain-Feingold statute in all its applications, rather than ruling a prohibition on corporate and union funding unconstitutional in all circumstances. The latter approach would be more disruptive to the bulk of existing relevant precedents, but it would be harder to avoid if WRTL had already undone the facial validity holding of McConnell. (Now, however, in any such future case, additional arguments can be brought to bear on the need for the McCain-Feingold statute and its relationship to the appropriate scope of Bellotti.)

Moreover, were the McCain-Feingold statute facially invalid, there would be increased ambiguity about the relevant statutory standard as well as whatever ambiguity exists in the application of Chief Justice Roberts’s constitutional test. The McCain-Feingold law contains a backup provision in case its primary one is ever facially invalidated. But the backup provision does not exhibit the same bright-line clarity of the primary one. By leaving the primary provision standing, Chief Justice Roberts actually achieves more operational clarity in the law going forward than if he overruled McConnell on this point. Thus, insofar as Walter Dellinger asserts that “the whole point of adherence to stare decisis is to create stability and predictability in the law,” Chief Justice Roberts’s position in WRTL is consistent with that objective, more so than if he had jettisoned the facial validity ruling of McConnell.

Stare Decisis, Gap-Filling, and the Rule of Law

There is a larger point as well. If all Justices feel as free to overrule precedents they dislike as Justice Scalia does, then ultimately constitutional law self-destructs into a lawless enterprise. When they regain power, liberals on the Court will overrule distasteful conservative precedents with the same speed and vigor as Justice Scalia would jettison the liberal decisions that he abhors. Unless stare decisis imposes a real constraint on Justices, forcing them to abide by precedents they find repugnant, then the constitutional pronouncements of the Court will simply oscillate over the decades between liberal and conservative dictates, depending solely on which side holds the balance of power on the Court, without any sense that there exists a body of constitutional law that all the Justices are mutually bound to uphold.

I imagine that this kind of perpetual oscillation would be destructive in other areas of constitutional law: when liberals are power, the Constitution protects abortion rights, permits affirmative action, strictly separates church and state (and so forth), whereas when the conservatives gain the upper hand, the reverse is true on all these issues—and back and forth, as the balance of power on the Court teeter-totters from one appointment to the next. In any event, the perniciousness of this kind of lawlessness certainly would hold true in the area of campaign finance. As I have indicated, there are two basic approaches one can take to the constitutionality of restrictions on the use of corporate or union money for campaign messages: one, the conventionally liberal approach, treats corporate and union money as different from other source of funds for political messages; the other, the conventionally conservative approach, treats them the same. Both approaches have plausible pedigrees in our nation’s constitutional traditions, and neither side is likely to convince the other of the correctness of its basic perspective. Unless stare decisis has true constraining power, we can envision the constitutional law of campaign finance swinging back and forth between these two basic conceptions, as liberals and conservatives fight for control. It is not a pretty vision. Election law, in particular, depends on the premise that there are settled rules that govern the operation of the electoral process. Even if one political party or faction does not like those rules, at least they know what the rules are and can rely on the courts to enforce them, unless the rules are changed through established procedures. If the rules of campaign finance are entirely dependent on the personal perspectives of the five Justices who happen to control the Supreme Court at any particular moment, the sense of the electoral process being governed by a set of clear rules is severely jeopardized.

Thus, Chief Justice Roberts was right to let precedent stand and to confine himself to the gap-filling task created by that precedent. Mere gap-filling will undoubtedly perpetuate tensions that already exist in the body of applicable precedent. But those tensions exist in large part because of the two basic competing visions concerning campaign finance, as some precedents reflect the temporary dominance of one vision, while other precedents reflect the transient preeminence of the other. Those two competing visions will not disappear. Thus, it is better to confine the ongoing competition between these two visions to whatever gaps that have been left open by the Court’s previous decisions, rather than extend the competition over the entire field of campaign finance, including those questions previously decided by the Court. This gap-filling process may be messy, but at least it is a recognizable form of law. Wholesale repudiation of campaign-finance precedent each time one vision or the other becomes predominant is not.

The New Chief Justice and His Commitment to the Rule of Law

In his confirmation hearing, then-Judge Roberts stated: “Judges and Justices are servants of the law, not the other way around.” This point is absolutely true, and it requires a strong commitment to stare decisis to remain so. For precisely this reason, then-Judge Roberts added: “Judges have to have the humility to recognize that they operate within a system of precedent shaped by other judges equally striving to live up to the judicial role.” Chief Justice Roberts’s opinion in WRTL may not be in all respects a perfect example of this humility. Yet his decision to stick to the gap-filling exercise that case presented and to leave the McConnell precedent standing, on the whole, is consistent with his professed recognition of the judicial obligation to serve the law—even as that law comes from the Court’s own prior decisions.

Each state in the nation needs to develop a regular institutional process by which it easily updates, in light of the experience of other states, its own laws that will govern in the event that a close election in that state ends up in a legal dispute. The need for this routine rule-revising regime is an idea developed in a paper entitled The Analysis and Mitigation of Electoral Errors: Theory, Practice, and Policy, which I presented last month at the Conference on Elections and Democracy at Stanford Law School. The paper will be published by the Stanford Law & Policy Review, one of the conference’s co-sponsors.

The paper endeavors to contribute in several different ways to current discussions of how to improve voting administration. The first part of the paper continues the conversation on how one might measure the quality of a state’s voting administration, a topic energized by Heather Gerken’s proposal for the creation of a “Democracy Index,” which I have discussed previously in this space, as has she. The second part presents a preliminary analysis of one portion of a five-state study of voting administration that my Election Law @ Moritz colleagues and I are conducting with the support of the Joyce Foundation. The third part identifies some policy proposals suggested by this preliminary analysis.

In this commentary, I wish to highlight one of these policy proposals, which builds upon an idea originally propounded by Rick Hasen. The idea is that states periodically should audit the rules and procedures applicable in the event of a dispute over the outcome of an election. The purpose of these periodic audits is to discover ambiguities or other uncertainties with these rules and procedures, since these legal uncertainties would exacerbate the difficulties associated with any disputed election. To the extent that a state’s law unambiguously specifies how to resolve any particular dispute that might arise, the scope of disagreement is narrowed considerably, if not eliminated entirely. As a result, the party that ends up losing the dispute is much more likely to accept the winner’s legitimacy—an end result surely worth striving for in a democracy.

In echoing Rick’s suggestion, I want to emphasize two points. First, while a periodic audit of the full range of a state’s voting administration laws might be valuable, a particular institutional mechanism should exist to focus specifically on those laws that might affect the resolution of a post-balloting dispute over who is the winning candidate. To distinguish these focused audits from more general ones, we can call them Election Contest Law Audits (or “ECLAs,” for short), as “election contest” is the generally prevailing term for any legal dispute that arises over the outcome of an election, and thus the law applicable to the resolution of an election contest would be Election Contest Law.

This body of applicable law, to be sure, encompasses more than just the procedures for conducting the election contest itself. It necessarily also includes any substantive rules that govern the resolution of any particular disputed point serving as the basis for the legal contest. For example, because the outcome of an election contest might turn on the rules for what an eligible voter must do to become registered, these registration rules would form part of the state’s Election Contest Law even though they do not concern the procedures for conducting the contest itself. Consequently, as a practical matter, there is likely to be a considerable overlap between a state’s Election Contest Law and the full range of its laws concerning voting administration. Even so, the overlap will be far from complete, as many details of voting administration cannot serve as the basis for contesting the outcome of an election. For example, a state’s voting administration law may include provisions for the payment of poll workers, which would be irrelevant to the identification of the winner.

In any event, the important conceptual point here is that an ECLA would look specifically for ways in which uncertainty in a state’s law might complicate the resolution of an election contest. The ECLA would not review a state’s laws for the purpose of making voting more convenient or making voting administration less expensive or achieving other efficiencies or improvements. While these other objectives are certainly worthwhile, they should be pursued in other ways. The purpose of the ECLA is the single-minded pursuit of reducing legal uncertainty that might arise in the context of an election contest.

Second, to maximize the effectiveness of an ECLA in each state, it is necessary to create a multistate clearinghouse of information, so that each state can learn from the election contests that have occurred in other states. This point is the extension of Rick’s initial suggestion that results from the five-state analysis presented in my Stanford paper. Simply put, each state has uncertainties in its own Election Contest Law that could be eliminated merely by adopting another state’s position on the particular issue.

Uncertainty exists in part because it is difficult to anticipate all the specific circumstances likely to arise in an election contest. Fortunately, election contests are sufficiently infrequent that no state “has seen it all before,” or indeed has seen very much of what specifically has occurred elsewhere. “The devil is in the details” is an aphorism especially apt for election contests, and it matters little that a state’s Election Contest Law has sound general principles but has not anticipated how to handle the myriad of devilish details that emerge in actual disputes. If each state undertook to clarify in advance how it would resolve all the specific issues that have arisen in election contests in other states, then the scope of uncertainty in each state would reduce dramatically.

In principle, it should be fairly straightforward to establish a multistate clearinghouse that could easily convey to each state the collective wisdom from all states, including updates in time for each state’s next periodic audit. These updates would digest the details of all new election contests from around the country since that state’s last periodic audit. In this way, each state’s ongoing reduction of uncertainty would be based on not only its own experience with disputed elections, but the recent experience of all other states.

Surely, one of the existing multistate organizations concerned with voting administration law—National Association of Secretaries of State (NASS), the National Association of State Election Directors (NASED), the National Conference of State Legislatures (NCSL), the National Center for State Courts (NCSC), or the U.S. Election Assistance Commission (EAC)—could take on the creation of this clearinghouse of information as a project. (If they have done so, I confess that I am unaware of it, and would welcome learning of its existence. As far as I know, the clearinghouse that the University of Florida has contracted with the EAC to provide is not focused on identifying, and reducing, uncertainties in each state’s Election Contest Law in the way that I have described.) There is no need for the clearinghouse to take a position on what the correct answer to a disputed issue should be. If states that already confronted a particular question diverge on how they answer it, the clearinghouse can simply identify the states on each side of the issue. What is important is simply that the clearinghouse alert all states that the issue is one that needs an answer. As a bonus, the clearinghouse can provide alternative answers for each state to consider, based on the experience of others states with the same issue. The utility of this clearinghouse would be proportionate to the quality of the organization by which it conveyed its essential information to the states. As I was reminded by the research undertaken for my Stanford paper, there is no dearth of available information concerning how states have resolved election contests that have occurred over the years. The challenge, however, lies in organizing and presenting this abundance of information in a way that is most accessible for each state when engaged in the focused task of spotting holes in its own set of contest-resolving rules.

For any state to take advantage of the service that this kind of clearinghouse would provide, the state would need the flexibility to make frequent changes to its own Election Contest Law after completion of each periodic audit. As I indicate in the Stanford paper, flexibility of this sort presumably would require that states authorize administrative agencies to update their Election Contest Law, rather than insisting that their legislatures retain this revision authority. Legislatures are simply too busy with too many different kinds of pressing public policy matters to devote the attention necessary to fill gaps in their Election Contest Law exposed by recent election contests that have occurred in other states.

States would do well, then, to consider how this sort of administrative agency should be structured and what mechanism might exist for its revisions of the state’s Election Contest Law to be put before the legislature for review. (One possibility is that the agency’s revision takes effect unless the legislature within 60 days acts to repudiate the revision.) States have a lot on their plates right now in terms of potential election law reforms: most are still undertaking the implementation of their HAVA-mandated databases, and some are reconsidering the voting equipment they purchased with HAVA funds just a few short years ago. Many states are also still struggling to implement effectively HAVA’s provisional voting requirements, and virtually all states are coping with the increasingly difficult challenges of recruiting and training highly competent poll workers. In this context, it might seem too much to ask states to develop an institutional reform that would lodge the authority to revise a substantial portion of their voting administration laws in an agency that is either entirely new or at least never entrusted with this degree of legislative power.

The overriding value of this regular auditing process, however, is too great to neglect. As important as all the other reforms are, they are in some respects superfluous in elections that are not close enough in their outcomes to be worth contesting. The rules matter most in a situation where they can make a difference as to which candidate will win a tight race. It is not surprising, then, that the public expresses the greatest concern about voting procedures when they potentially affect the result of a race that the public particularly cares about. Thus, there is an inherent efficiency in focusing the efforts of election law reform on reducing the uncertainties that might arise if a close election is disputed. If law reformers were able to identify and eliminate in advance of the next election every uncertainty that might form the basis of a challenge in a close vote, the reformers would have put in place an exemplary set of voting procedures, which should satisfy both public and expert opinion on the need for change.

Accordingly, among all the activities that states are currently undertaking to improve their systems for voting administration, a high priority should be placed on the development of the auditing process I have described. Its key features would be new institutional arrangements in each state authorized to update expeditiously the Election Contest Law there, in light of the flow of well-organized information from a multi-state clearinghouse that tracks disputed elections.

Jennifer Brunner, Ohio Secretary of State, is justified in seeking removal of all four members of the Cuyahoga County elections board, who ultimately hold responsibility for the egregious mismanagement of the voting process there in 2006 (which followed on the heels of serious failings that occurred in 2004). But Brunner needs quickly to put in place a new bipartisan board, two Democrats and two Republicans, who will be able to exercise all the normal responsibilities of the board in 2008, so that she herself—an elected Democrat—does not displace the board’s authority for making all the local decisions that will arise regarding the casting and counting of ballots in Cuyahoga County in the upcoming presidential election.

Brunner is right to be concerned about preparations for the presidential election in this most-troubled, and most-populous, county. Cuyahoga County’s inability to conduct elections properly—with serious implications for the nation’s ability to pick a President—go back at least as far as 1972, when a systemic failure to deliver voting equipment to polling locations caused, among other problems, 16 precincts never to open on the Tuesday designated for the presidential primary. Consequently, in addition to keeping the polls open until midnight countywide, a federal court ordered a whole new day of voting in those precincts on the following Tuesday.

Having largely forgotten that fiasco, the nation was reminded of Cuyahoga County’s perennial trouble in administering elections when two of its officials were criminally convicted of having rigged the recount of the presidential vote in the 2004 general election. The fact that their motive apparently was laziness, rather than partisanship, does not sufficiently allay concerns about the integrity of vote counting in that county.

The severe problems of the May 2006 primary are too numerous to detail here, but among the most serious are the loss of memory cards electronically recording vote totals from fourteen percent of the county’s precincts and spoiled paper records for ten percent of the county’s electronic votes. Even after all the attention devoted to fixing problems of this sort before the November 2006 general election, major problems persisted. According to a Plain Dealer report, almost 12,000 ballots were cast in Cuyahoga County on November 7 by individuals who did not sign the poll books and whose eligibility to vote apparently was never properly verified by precinct officials, despite the requirements of state law that mandate these procedures. The most likely explanation for this breach of the rules is that substantial confusion occurred at these polling places, and poll workers allowed the individuals in question to bypass the check-in line and go straight to voting machines. Just imagine the nightmare, however, if one of the prominent statewide races last year—for example, the one between Mark Dann and Betty Montgomery for Attorney General—had turned out close enough that these 12,000 ballots provided a basis for contesting the result.

No wonder, then, that Secretary of State Brunner is being aggressive in an effort to prepare Cuyahoga County for 2008. Aggressive steps are necessary, including replacement of the old board with entirely new leadership. Three of the four have appropriately resigned at her request, but the one remaining holdout, Bob Bennett—who also chairs the Republican Party statewide—refuses to leave without a fight. Trying to accuse Brunner of attempting to control the voting process for partisan advantage, Bennett himself looks like he is putting partisan (or perhaps personal) motives before the public interest. (On Capitol Square, the Ohio News Network program, he said that he did not want his reputation tarnished at the end of his career. But prolonging this battle can hardly improve his reputation: the past mismanagement of elections in Cuyahoga County cannot be undone, and while no one thinks he is exclusively to blame for all the administrative errors that occurred there in 2004 and 2006, he surely shares responsibility as a member of the board, just as does the member of a corporate board that fails to prevent financial ineptitude perpetrated by the company’s management.) Rather than clinging to his board position, Bennett should step aside magnanimously—and immediately—so that a different Republican can take his seat, and the newly constituted board can begin tackling the difficult challenge of recruiting and training poll workers as well as all the other steps necessary to make the voting process work properly in May and November 2008.

Still, for her part, Brunner needs to do more to dispel any appearance of partisanship in her own actions. Among the moves she has made is announcing that she is putting the Cuyahoga County elections board under her “administrative oversight” that she says may last until December 31, 2008—after the presidential ballots are cast and counted, the Electoral College has met, and presumably all that is left is for Congress to officially pronounce the winner on January 6, 2009. Although Brunner claims that she is imposing this administrative oversight only to assist the new board in getting ready for 2008 (on the same Capitol Square program she said, “I’m not trying to run this Board. I’m not trying to tell this Board what to do.”), the fear is that she will make decisions for Cuyahoga County designed to favor the Democratic presidential candidate in next November’s voting.

Brunner should announce that, if Bennett resigns and local Republicans nominate a suitable replacement by May 1 (three weeks from today), then she will plan to terminate administrative oversight of the new board by December 31 of this year, rather than next. That timetable gives her, as Secretary of State, eight months to get the new bipartisan board ready to manage its own operations in 2008 in the same way that the local boards in other counties will. If serious problems occur in the May 2008 presidential primary, then Brunner can revisit whether some more drastic arrangement is needed to assure that the voting process works properly in November 2008. But even then the more drastic measure should have a built-in bipartisan structure, so that it is never possible to claim that Democrats control the process of casting and counting presidential ballots in Cuyahoga County.

The requirement of state law that county election boards be bipartisan in their membership is a salutary feature of election administration in Ohio, one that it is often mentioned when comparing our state’s electoral system favorably to others (for example, Florida’s). Some think that it would be even better to have the statewide authority for administering elections be vested in a similarly bipartisan board, rather than a single Secretary of State elected as a partisan candidate. These folks point to Brunner’s predecessor, Ken Blackwell, as Exhibit B to make their case (Exhibit A being Florida’s Katherine Harris). Whatever one’s position on that point, Ohio should not backslide by injecting more partisanship—or even just the appearance of partisanship—into the operation of voting procedures, including the counting and recounting of ballots, at the local level.

The 2008 presidential election in Ohio shows every sign of being an intensely fought campaign. We risk accusations and counter-accusations concerning alleged manipulations of process to give one side or the other an unfair or improper advantage. There should be sufficient statesmanship in this state for political leaders on both sides, while the November 2008 election is still 18 months away, to create a bipartisan mechanism to handle the controversies that inevitably will arise.

That statesmanship should start with Brunner and Bennett engaging in enough diplomacy so that an entirely new bipartisan board in Cuyahoga County can get to work right away and, having had at least half a year of Brunner’s tutelage, be in a position to operate the November 2008 election on the ground in that county as one would expect a local elections board to do.