Posted: June 17, 2008
Election-Day Emergencies: Are We Prepared?
Since 2000, the election administration community has worked hard to reduce the probability of technical problems and maladministration in conducting the vote. However, in jurisdictions where those efforts have been most successful, the low-hanging fruit has largely been picked and further effort in these areas will reap only diminishing returns. For that reason, it may be time to take a step back and look at whether there are some other areas that have not received much attention, but where great gains can be made with relatively little effort.
One area that warrants attention is emergency preparedness. The EL@M team recently completed an analysis of 16 states to see whether and to what extent they were prepared for a hurricane, flood, or other emergency occurring on or right before election day. Eight of the 16 states had no law telling election administrators what to do if this type of emergency occurred. Furthermore, with the exception of Ohio, these eight states had no publicly available document addressing the subject (the Ohio Secretary’s only recommendation is that local election officials should have a risk mitigation plan). Because it is relatively easy and could potentially have great returns, administrators in these states should make it a priority to come up with some sort of basic statewide plan in this area.
One factor to consider when making this plan is who should be in charge in the event of disaster. Florida, which probably has the most well-developed disaster plan of the 16 states we analyzed, puts the governor in charge. The governor may reschedule the election to occur another day. Other states put the chief election officer or the courts in charge. While I’m sure it is possible to have an intelligent debate about who should be in charge, the most important thing is to clearly assign responsibility and power to somebody. Our analysis suggests that most states have not yet reached that level.
Another task is to make up some rules about whether and under what circumstances the election can be rescheduled. How bad does the disaster have to be to justify calling the thing off and trying some other day? Clearly some disasters, though unfortunate, do not disturb order enough to justify the cost of rescheduling. Other disasters are so bad that it would be unconscionable not to reschedule, but where do you draw the line? It is less productive to obsess over where the “ideal” line should be than to draw the line somewhere, and give power to somebody reasonable to determine when that line has been crossed.
A final question to consider is, in a statewide election, whether to permit leaders to reschedule voting in some jurisdictions and not others. Rescheduling in only some areas allows leaders to tailor the remedy so that it applies to only those areas where disaster occurred. While this might sound like a good idea, it also creates some problems. The biggest problem is that voters in the rescheduled jurisdictions, before deciding whether and how to vote, get to have at least some idea of how voting came out in the other jurisdictions within the state and elsewhere. If they see that their votes are likely to change the result of the election, they will turn out in numbers that are unnaturally high. This might cause both candidates and voters whose voting was not rescheduled to complain the process was unfair. It might also trigger an election lawsuit alleging the same. An alternative that avoids some of these complications is to require that elections be rescheduled throughout the whole state, or not at all. But even this alternative creates a possibility for strategic behavior as, in a Presidential race, the voters of the rescheduled state will get to see how all the other states voted before deciding whether and how to vote (note that rescheduling a federal election might not even be possible—see here). Furthermore, because the cost and degree of complexity inherent in rescheduling the election statewide is higher than rescheduling in only the areas affected by disaster, there is a risk that leaders will refuse to reschedule an election even when they really should.
As our analysis makes clear, there is no perfect solution to these problems, but there are different costs and benefits to each approach. However, there is one approach where the costs always outweigh the benefits—the approach of not having any plan at all. That’s where a lot of states are right now and, to be fair, having no plan in this area has worked okay so far. Still, just because one has gotten away with playing Russian roulette in the past is no reason to continue doing it. Hurricanes, tornadoes, floods and other disasters may be uncommon, but they are a lot more common than the freakish twists of fate that triggered Bush v. Gore. We should prioritize accordingly.