By Daniel P. Tokaji

The 2008 election season is already upon us. Each day’s news contains more stories about the candidates’ attempts to raise money and woo voters. In addition, bills have been introduced in Congress dealing with such diverse matters as voter intimidation, electronic voting security, and a democracy index to measure the health of state election systems.

If significant changes are to be made in time for next year’s election, it’s vital that implementation start soon. Among the most important reforms to consider is in-person early voting.

As we look forward to next year’s election, the prospect of long lines at polling places is one of the most serious issues that should be considered. Long lines may discourage or even prevent some voters from participating. These lines are especially worrisome if they are concentrated in certain areas — such as major urban centers, rural counties, or near college campuses — since that presents a greater likelihood of skewing the electorate than would be the case if such problems were distributed statewide.

Ohio was among the states that had extremely long lines, in some but not all parts of the state, in the 2004 presidential election. Because Ohio is very likely to again be a battleground state in 2008, it’s worth paying special attention to its experience. Franklin County, where Columbus is located, had exceptionally long lines at some precincts. In Columbus, some voters reported waiting four to five hours. Another county that had problems was Knox County, home to Kenyon College, where some voters waited until the wee hours of the morning to cast their votes. This led a federal district court in Columbus to issue a temporary restraining order on election day, requiring that polls in these counties be kept open for voters waiting in line and that they be provided with some alternative way of voting.

Election-day court orders are sometimes necessary But the reality is that, by that time, it’s usually too late to do much about long lines. Some voters will be discouraged from coming out if lines are too long, while others will have other obligations that prevent them from waiting for hours. It would be far better to think now about what can be done to take pressure off of our polling places on election day.

Unfortunately, there are reasons to be concerned that the problem of long lines may actually get worse. There were reports of long lines in some places, including Cuyahoga and Miami Counties in Ohio, in the 2006 mid-term elections. Though it’s hard to know for sure, it appears very likely that turnout will again be high next year, at least in battleground states. Two other changes that have occurred in the last two years are also likely to increase pressure on the polls.

One change is the enactment in several states of laws requiring voters to identify themselves at the polls on election day. Although the Help America Vote Act of 2002 (HAVA) contained a voter ID requirement that took effect in 2004, this requirement applied only to a limited subset of voters, namely first-time voters who’d registered by mail. Since then, several states — including Florida and Ohio — have made changes to their laws to impose new identification requirements for all voters. It’s possible that these changes could create or exacerbate bottlenecks. Some of these requirements, particularly those in Ohio, are very confusing. And in Florida, voters who don’t have photo ID are obligated to cast provisional ballots. All of this can be expected to slow things down at the point voters check in.

The other major change that can be expected to have an effect in 2008 is the implementation of new voting technology. Although HAVA required states receiving federal funds to have new equipment in place by 2006, some of those voting next year will be using that equipment for the first time.

More importantly, it’s not clear counties purchased enough voting machines to accommodate all those who will be voting on election day. The fact that a polling place had enough voting machines in last year’s election doesn’t mean that there will be enough for next year’s election, given that turnout will probably be higher. One solution to this problem, of course, would be to buy new machines. This may not be a practical solution, however, given the expense involved and the limited time available to get new equipment in place. And just yesterday, the AP reported that the voting system in Cuyahoga County, Ohio’s most populous, is probably inadequate to handle the increased turnout expected next year.

It’s therefore essential to think carefully about what can be done to relieve pressure on polling places on November 4, 2008. One possibility is to expand opportunities to cast absentee ballots before election day. Most states now allow no-excuse absentee voting, in which anyone can cast an absentee ballot by mail, regardless of whether they have a reason for not going to the polls on election day.

There are, however, some serious concerns surrounding expanded mail voting. Foremost among these is the possibility of fraud and coercion, which is generally easier to accomplish with mail ballots than in-person voting. In fact, most of the evidence of modern-day voter fraud — anecdotal though it is — concerns absentee voting. There’s also the possibility that voters will be pressured by family members or caregivers to vote a particular way, especially in institutional settings. Mail ballots are also inaccessible to a significant number of disabled voters, including people with visual and manual dexterity impairments.

In addition, there are many ways in which voters can make mistakes in the absentee voter process, which result in their votes not being counted. One is by failing to comply with the application requirements for absentee voting, which are very complicated in some states. Here again, Ohio is an example, by virtue of it’s intricate rules regarding voter ID, which caused serious problems for some absentee voters in 2006. Even if they succeed in applying for an absentee ballot, voters can make mistakes in casting and returning it. Without the benefit of “notice” technology available at polling places, voters may mistakenly cast overvotes or undervotes. In addition, they may make mistakes in returning their ballot — like failing to include proper postage or required identification — that can result in their votes not being counted. The fact that voters don’t have an election official or poll worker to assist them may increase the risk of mistakes.

In-person early voting eliminates some of the inherent problems with mail voting. Typically, early voting takes place at a central location, such as a board of elections office or a local public library. Voters can appear at that location several days before the election in order to cast their votes. Unlike mail voting, election officials are available to assist voters who need help, either in complying with ID requirements or in understanding how to use the voting machine. In fact, local election authorities may be in a position to provide better assistance than is possible on election day, by having more knowledgeable paid staff rather than volunteer poll workers available for early voting. Voters also have the advantage of new technology that is accessible to people with disabilities, and provides all voters with notice and the opportunity to correct errors before they vote.

For all these reasons, in-person early voting is preferable to mail-in absentee voting, as a means by which to relieve election day pressure on the polls. This reform is included as part of the proposed Count Every Vote Act, section 352 of which would require that people be allowed to vote within the 15-day period before the election “in the same manner as voting is allowed on the date of such election.” However, this bill also includes some provisions that are likely to prove controversial, such as felon re-enfranchisement, no-excuse absentee voting, and contemporaneous paper record requirements.

Accordingly, supporters of in-person early voting should consider breaking off the early voting requirement as a separate, stand-alone bill. This is a reform that should be appealing to those across the political spectrum, since it has the potential to expand access without imperiling the integrity of our electoral system.

By Nathan Cemenska
Web Editor, Election Law @ Moritz
Moritz College of Law

Ohio Secretary of State Jennifer Brunner recently recommended that Ohio should begin experimenting with “all vote by mail” (AVBM) elections, in which the traditional polling place is replaced by what is essentially mandatory absentee voting for everyone. However, rather than moving to AVBM for all elections, Secretary Brunner proposes we use the system only for those elections in which there are no candidates, just ballot issues. In other words, where officials in many other states are considering an immediate transition to full AVBM, the Secretary proposes to study the issue and get some practical experience with it before making any big moves. Our legislature should support her proposal.

It should support it because there are a lot of issues with AVBM that need to be worked out before a full AVBM program could be responsibly implemented in Ohio. Specifically, while AVBM is relatively new and therefore difficult to judge, there is some reason to think it exposes ballots to a greater risk of fraud than traditional in-precinct or “hybrid” absentee/in-precinct voting systems. While this should not preclude experimentation, it weighs heavily against widespread use until we learn more about how it works. In the meantime, here is what Ohio might encounter if experimentation goes forward.

The main issue with AVBM is that it requires that all ballots be cast outside the polling place, whereas last November only about fifteen percent of Ohio ballots were cast in this way. The increase presents additional risks because, historically, ballots cast outside the polling place have proven to be convenient vehicles of fraud. Outside the scrutiny of the polling place, voters may be subject to unwanted influence from overzealous campaign workers or party officials. They may also be subject to types of influence that, sadly, are not necessarily unwanted, such as offers to buy votes. Mail ballots are also more likely to lead to multiple voting if corrupt players find ways to get their hands on extra ballots. Finally, mail ballots are sometimes intercepted and tampered with after casting. In one scheme, a friendly worker offers to “save the voter a stamp” by delivering the completed ballot on his or her behalf. Of course, no one can say whether the ballot arrives in its original condition, or at all. Also, AVBM systems might allow a postal worker or other insider to target and destroy ballots returned from a particular neighborhood he or she seeks to disenfranchise. In fact, according to a report by Paul Gronke, a prominent political scientist, Oregon officials would have “no way of knowing” whether ballots were intercepted and destroyed in this way (see page 4).

While our current no fault absentee voting system also presents these risks, it at least limits the number of mail ballots in proportion to the number of voters who want to vote absentee. In contrast, an AVBM system forces everyone to cast a mail ballot whether they want to or not. The result is that, if fraud is ever accomplished, it is likely to be more widespread under an AVBM system.

Despite these concerns, some legislators claim that all-mail elections will actually help reduce the risk of fraud. However, at least some of these statements seem to be referring to the risk of fraud accomplished by hacking DRE voting machines. While DRE hacking may be a potential issue, the number of people capable of understanding how it works seems fairly limited. In contrast, mail ballot fraud seems much more straightforward and likely to occur.

According to Gronke’s report, Oregon claims its system is more fraud-proof than other systems because of the rigor with which its officials verify signatures submitted on AVBM ballots against those contained on registration applications. Because AVBM spreads election activity over the entire ballot-submission period rather than trying to cram it all into one climactic day, the argument goes, administrators actually have the time to sit down and make sure the signatures match. Furthermore, they are required to have some training in signature verification, increasing the likelihood of preventing fraud. If all of this is true, then Oregon probably does have a better chance of catching forgeries than some other states, but that still does nothing to prevent those types of fraud that do not involve forgery. Furthermore, given the mind-numbing process that signature comparison must be, it is likely that some forgeries could slip through the cracks despite workers’ best efforts to focus their attention.

Then why would anybody want to move to an AVBM system? Part of the original justification was that it would increase voter turnout, but for whatever reason that hasn’t happened in Oregon (see page 2 of Gronke’s report). However, that doesn’t necessarily mean it wouldn’t increase turnout here in Ohio, which has different demographics. Of course, those very same demographics may expose new AVBM problems that Oregon hasn’t experienced. The Secretary’s incremental plan will help us get to the bottom of these interstate variations.

But besides the potential turnout benefit, AVBM requires a lot less money, effort, and staff than a hybrid system. Gronke’s report indicates that Oregon’s prior hybrid elections were both difficult and frustrating to administer, but that the AVBM system is much more manageable (see page 3). And officials nationwide are scrambling for alternatives that require fewer staff because they just can’t drum up enough poll workers to conduct proper in-person elections. Even where they obtain enough workers, they’re finding that many of those workers don’t have the background and training necessary to understand how to run a modern polling place.

With justifications like these, some would accuse officials of considering AVBM not because it’s better, but because they haven’t secured the money, resources and staff to do hybrid elections right. While in some cases there may be some truth to these accusations, there is more than a little bit of resentment behind them as well, and it is unwarranted. The reality that officials must deal with, and that those in the world of ideas can usually avoid, is that you’ve got to work with the materials you have. We all want what is best for our elections, and if traditional polling places are really doomed for whatever reason to suffer biblical plagues of long lines, late openings, confused workers, haywire touchscreens, security headaches, ballot shortages, fritzing scanners, blown VVPATs, counting delays, etc.—if traditional polling places are really doomed to suffer all these problems, then we have to let them go. But before we let them go, let’s make sure we’re doing it because they really are doomed, and not just because it’s easier.

And, in my mind, that is exactly what the Secretary’s plan would allow us to do. By starting off an AVBM “pilot” program of limited scope while continuing to tinker with our hybrid elections, we will be able to see whether and to what extent a transition to full AVBM is justified. For that reason, the legislature should pass the laws necessary to allow the Secretary to perform her proposed program of experimentation, and should do so quickly. Because if we don’t start learning how AVBM works now, forces may appear and make us implement a full AVBM system without having first had the benefit of the Secretary’s experiment. And that would be a rough transition.

By Heather Gerken
Professor of Law
Yale Law School

Many thanks to Ned Foley for letting me respond to his thoughtful post, about designing the “Democracy Index,” a national ranking of state election administration practices, which I proposed a few weeks ago in the Legal Times. Though I was tempted to reply to Ned’s helpful suggestions with a simple “amen,” it seems useful to sketch out an important issue flagged by Ned’s post and talk about some of the questions raised about the proposal during the last few weeks.

Picking the right metric. Ned and I agree substantially about the way the Index should work. It must focus on the nuts-and-bolts of election administration. It should be assiduously nonpartisan. It ought to encompass issues all voters care about (long lines, discarded ballots) while eschewing hot-button topics like felon disenfranchisement or campaign finance.

Ned is also right that the Index should measure how well the election system is run, not the general health of the state’s democracy. In practice, it will sometimes be tricky to separate those issues Consider, for instance, registration rates. Registration rates are plainly influenced by what Ned calls “social and cultural conditions,” and it seems unfair – or at least unproductive – to penalize a state for low registration rates caused by socioeconomic conditions beyond election officials’ control.

But low registrations rates may also be symptoms of a badly run system. Election administrators sometimes manipulate registration requirements to make it harder for voters to take part in the election. Think about Ohio’s Kenneth Blackwell’s failed effort to force voter registration cards to be printed on paper of a certain weight. I’d bet that most voters would think that requirement was a sign of mismanagement. Indeed, I suspect most voters would agree that states should make it easier, not harder, to register to vote provided there are adequate protections against fraud. The question, then, is whether we can devise output measures that penalize states for creating unnecessary barriers to registration without asking election administrators to fix problems beyond their capacity.

Three other questions about the Index have been posed to me during the last few weeks. The first is whether civil rights groups should back this proposal. The second is what to do if states fake their data. And the last is whether the Index would reduce the incentive for top-ranked states to improve their systems. Let me address each in turn.

The New Vote Denial. The first question – whether civil rights groups should be interested in backing this proposal – has already been answered by OSU’s own Dan Tokaji in his article, The New Vote Denial. As Dan points out, the “new vote denial” is the old vote denial. We are back to worrying about the use of basic administrative practices to deprive citizens – especially racial minorities – of the right to vote. For this reason, as Rick Pildes has observed, one of the best ways to protect minority voters is to enact uniform rules protecting the right to vote generally.

To be sure, the Index is different from the traditional weapons in the civil-rights arsenal. It would require advocates of reform to speak the language of corporate executives, not constitutional lawyers. The debate will not be about ideal practices, constitutional rights, or equal protection but hard numbers, accountability, and bottom-line results.

It is, however, essential that we cast the debate in those terms. One of the reasons that the new vote denial is so hard to combat is that it’s difficult to prove intentional discrimination or vote dilution in any given instance. We need a different kind of metric to challenge the new vote denial, and the Democracy Index provides that metric.

It would, of course, be a mistake to shift all reform efforts to the turf of “moneyball politics”; doing so would obscure the stakes of these debates. Nonetheless, as Spencer Overton has recently emphasized in his comprehensive article on voter i.d., data-driven, factual analysis ought to be part of every reformer’s vocabulary. Indeed, the Democracy Index seems like a promising way to bridge the political divide over election reform, as words like accountability and transparency are part of everyone’s vocabulary.

Faking the data. Another important question raised about the Democracy Index is whether states will manufacture data to improve their rankings. This worry is, of course, a happiness problem. If the Democracy Index were having such a powerful effect on politicians that they were tempted to cheat, we would already have come a long way. Nonetheless, every law professor is familiar with the outrageous lengths to which schools have gone to improve their ranking on the U.S. News and World Report ranking (University of Texas’s Brian Leiter has been one of the most trenchant observers of this problem, see, e.g., http://www.leiterrankings.com/usnews/guide.shtml).

Happily, there are a number of ways to verify state disclosures. The scholars who have designed the Environmental Performance Index, which ranks the environmental performance of nation-states, have succeeded in ensuring their data are reliable. The same can be done in the election context. For instance, many voters complained in 2004 about long lines at polling places. One way to test whether states are properly disclosing information about voters’ experiences would simply be to ask people during exit polls how long it took them to vote. We could also devise random sampling strategies that would give us a pretty good sense of which states were playing fast and loose with their data.

Resting on one’s laurels or keeping up with the Joneses? Finally, several friendly critics – who agree that the Democracy Index will encourage states at the bottom of the ranking to improve – worry that top-ranked states would lose any incentive to do better. This situation would still represent an improvement on the status quo – at least the Index would provide an impetus for change somewhere in the system. But it is also possible that a ranking system will encourage top-ranked states to compete among themselves. For instance, residents of reform-minded states like Massachusetts or California might care a great deal about how they rank against similarly progressive states even if they are ranked well above others. Consider, for instance, what took place when the first Environmental Performance Index was released, showing Sweden and Norway ranked first and second on the worldwide ranking. One might have thought that Norway would have celebrated its extraordinary achievement. Just the opposite occurred. Why? Norwegian leaders did not care that they were ranked ahead of 131 other states. What mattered to them? Sweden was number one.

Generating a race to the top is not the only way in which the Democracy Index might turn one of our system’s liabilities – decentralization – into something positive. For instance, an Index gives states the chance to experiment with different reform strategies while providing objective measures for holding them accountable. States can serve as the “laboratories of democracy” lauded by the Supreme Court. And we can test the results from these many experiments. That is because, consistent with the moneyball approach, the Democracy Index lets us figure out the real drivers of performance. Once we know who runs the best elections, we can figure out the formula behind those successes. For instance, is it money or training that guarantees a well-run system? Even if our end goal is uniform, national legislation, the Democracy Index should help Congress identify best practices, choose sensible standards for regulatory floors, and pinpoint local outliers.