This comment is part of a colloquy initiated by John Fortier’s thoughts on the potential implications of Hurricane Sandy for this year’s election.
John raises important questions, both about how to handle the electoral implications of Hurricane Sandy specifically and also more broadly about how states and the federal government can better prepare for future emergencies that disrupt voting on Election Day.
It is noteworthy that these questions are not new. John himself had considered them previously, as has my colleague Steve Huefner. Moreover, John, Steve, and I all worked together in advance of 2008 to “war game” a hypothetical scenario in which a snow storm disrupted voting in Denver and surrounding suburbs, resulting in the differential extension of polling hours among affected localities, thereby raising potential Equal Protection questions. The distinguished three-judge panel that we assembled for this exercise rejected the Equal Protection claim presented by the specific facts of our hypothetical, essentially reasoning that the storm affected Denver worse than the suburbs and therefore justified extra voting hours in the city that was not available to voters in the suburbs. But the the panel’s anlaysis implied that the outcome under Equal Protection might be different if the specific facts of how a state government responds to a storm varied from our scenario in critical details (for examples, voters in two counties being treated differently even if they suffered the same storm-related disenfranchisement).
One disappointing lesson in the wake of all the scholarly work on this topic over the last decade is how little development in the law of “emergency electoral planning” there has actually been during this period. By and large, as John indicates (as have others over the last couple of days), states are no better prepared to handle an emergency that seriously disrupts Election Day than they were in 2004 or 2008. Is it too much to hope that, after this year’s election is over, we’ll use the experience of Sandy—whatever the level of disruption it actually turns out to cause—to put in place better risk management systems for the future.
In the meantime, given that Sandy is in fact upon us, it makes sense to focus specifically on what might best be done between now and November 6 to minimize the likelihood that effects from storm might disrupt voting on Election Day. Because Sandy is hitting a week early, rather than on Election Day itself, the most sensible risk management planning right now should be tailored to the specifics of this particular storm, rather than the separate situation in which a storm, or earthquake, or some other disaster, hits unexpectedly on Election Day itself.
From this vantage point, it seems that the most likely risk is that lingering power outages continue until Election Day, as well as the possibility that flooding may cause some neighborhoods to be uninhabitable on November 6. To be sure, early voting has already been disrupted, but that problem seems secondary at this point to making Election Day itself as widely available as possible, as well as potentially relying on additional absentee ballot opportunities for those who because of the storm will be unable to cast any kind of ballot—even an emergency backup paper one—at a polling place. There is already talk of potentially moving or extending Election Day beyond November 6 itself; but while that is an option, in my judgment it should be considered a last resort to be exercised only if other contingency plans cannot work.
In other words, if polling places are unable to open (or function properly) on November 6, one option might be to let those affected voters cast absentee ballots that, as long as they are postmarked by Election Day, are still eligible to be counted if they arrive at the local board of elections by November 16. That solution was used in a consent decree to protect overseas Florida voters, and other states (including Ohio) already have this as routine practice. Voters who have evacuated their homes may have access to a computer where they could download an absentee ballot from the internet, much as members of the military can, even though these voters displaced by the storm would be unable to get to their local polling place.
Moreover, backup generators may work to supply power to our increasingly electronic voting equipment: we use computers not just to cast or count ballots, but also track voter registrations. But election officials should prepare an adequate supply of paper ballots, just in case power that is temporarily restored flickers out again while voting is underway on Nov. 6. If necessary, because electronic poll books have failed, these paper ballots can be treated as provisional, so that voter eligibility is verified (in the same way that absentee ballots are) before the ballot is counted. In a close election, armies of attorneys will be ready to challenge every move made by election officials, and (as we have learned with touchscreen machines) the system is better protected if there is a paper record that can be audited in the election’s aftermath.
This combination of two types of paper alternatives—(1) emergency backup paper ballots in polling places capable of opening, and (2) emergency access to absentee ballots for voters who have been displaced from their homes, or who otherwise cannot get access to a polling place, because of the storm—is a far preferable option, if feasible, than extending voting beyond November 6 itself. Although Congress, in 3 U.S.C. § 5, has authorized states to do this for presidential elections, it could cause the undesirable situation in which one or two states, like Virginia and/or New Hampshire, are continuing to cast ballots—with the Electoral College outcome still in doubt—while the results are known in all other states. We would be in unchartered waters if a different state (for example, Ohio), which was less affected by the storm, decided to keep its polls open an extra day or two also, to give its citizens an extra opportunity to affect the Electoral College outcome.
Therefore, every effort should be made to develop contingency plans that will enable the casting of ballots to end, as expected, on November 6. Our electoral system was built with the understanding that it may take some days after November 6 to complete the counting of those ballots, but to keep the polls open several more days is much more unsettling to the system. It is not that we should never exercise that more extreme option. It is only that we should carefully assess whether it is truly necessary, after considering the feasibility of less extreme but potentially effective measures.