One of the catch-phrases during the 2004 election season was the "margin of litigation." Some have attributed the absence of a protracted post-election fight to the fact that the margin of victory in key states "exceeded the margin of litigation." As Congress and state legislatures consider whether to amend existing election laws, they would do well to consider the impact of proposed changes on the margin of litigation. Several provisions of Ohio's pending election bill, for example, would have the unfortunate effect of increasing the margin of litigation, making post-election lawsuits over the outcome more likely.
What Is the Margin of Litigation?
It is first helpful to clarify what is meant by the term "margin of litigation." While there may not be universal agreement among those who have used this term, it may be understood as the number of votes below which a particular election might be contested in court – or, conversely, the number of votes above which an election is "safe" from legal challenge. Thus, if the margin of victory exceeds the margin of litigation, an election will probably not be contested in court; on the other hand, if the margin of victory is less than the margin of litigation, then lawsuits challenging the outcome can be expected.
As my colleague Steve Huefner noted in the wee hours of the morning after the November 2, 2004 election, it appeared for a moment that the margin of victory in Ohio – the pivotal state in 2004 – might be within the margin of litigation. That's largely because it wasn't known how many provisional votes were outstanding and yet to be counted.
By mid-morning on November 3, however, it had become clear that President Bush's margin of victory was large enough to survive any plausible legal challenge. As Professor Rick Hasen explains in his forthcoming article "Beyond the Margin of Litigation," President Bush's morning-after lead in Ohio was approximately 136,000 with around 153,000 provisional ballots still to be counted. Given that not all of those provisional ballots would be deemed valid, and that some of them would have been Bush votes, Senator Kerry's campaign "did the math" and concluded that there was no realistic chance of prevailing in a legal challenge to the outcome. On the other hand, Professor Hasen explains, a margin of victory under 36,000 votes would likely have been within the margin of litigation, leading to a possible replay of the 2000 election controversy.
So what was the margin of litigation in Ohio's 2004 election? While no one can say with precision, my estimate is that it was no more than 90,000 votes. As I explained in this post, that's the most votes that Kerry could have hoped to pick up through post-election litigation, on the most optimistic (from his perspective) assumptions. Thus, the morning-after difference of 136,000 votes was substantially greater than the margin of litigation – and the ultimate margin of approximately 119,000 votes supports the conclusion that litigation wouldn't have altered the result.
By way of contrast, the ultimate margin of victory in the State of Washington's 2004 gubernatorial election was a slender 129 out of over 5.8 million cast. My colleague Ned Foley therefore observes that the margin of victory in that election "was well within the now-proverbial margin of litigation." This explains the protracted legal fight that took place in that state, and that continued long after the declared victor had been sworn into office.
Lessons for Election Reformers
One of the lessons that should emerge from these experiences is that, in considering proposed changes to our election system, it's essential to consider how they will affect the margin of litigation. A reform that makes post-election litigation less likely may be said to decrease the margin of litigation – it decreases the number of votes above which post-election lawsuits can be expected. By contrast, a change increases the margin of litigation, if it raises the number of votes that must separate the candidates in order to be safely out of the litigation danger zone.
All other things being equal, then, a reform that decreases the margin of litigation is a good thing. It promotes the value of finality, making it more likely that there will be a resolution on (or shortly after) the day of the election. On the other hand, again all other things being equal, changes that increase the margin of litigation are a bad thing, because they make it more likely that the outcome of an election will be fought out in court afterwards.
Of course, all other things aren't always equal. While finality is important, it's not the only value that our election system must serve. Two other values of importance are access (making sure that all eligible voters can vote independently and have their votes counted) and integrity (making sure that only eligible voters are allowed to vote, and that the process of vote tabulation is free of fraud and error). Improving our election system will sometimes require that difficult choices be made among these three values.
One of the factors that can increase the margin of litigation is the use of unreliable voting equipment like the "hanging chad" punch card. This equipment tends to result in ambiguously marked ballots that can't correctly be read by mechanical counters. Its use may therefore trigger calls for a manual recount and attendant litigation, as occurred in Florida's 2000 election.
Perhaps the most important factor in determining the margin of litigation is the number of provisional ballots cast. More than anything else, a large number of provisional ballots has the potential to call the immediate post-election vote totals into question. For example, if there had been 400,000 provisional ballots cast in Ohio's 2004 election rather than 153,000, then the margin of litigation would have been significantly higher – probably high enough that it would have led the Kerry campaign to take the fight to court, rather than conceding.
A Case Study: Ohio's Pending Election Bill
These considerations should inform the debate over proposals to revamp existing election laws. The Ohio legislature, for example, is currently considering a bill that would make numerous changes to the state's election system (Sub H.B. 3). Among the questions that should be asked, in determining whether to adopt such legislation, is whether it will increase or decrease the margin of litigation.
An examination of the Ohio bill reveals several provisions that can be expected to increase the margin of litigation – and therefore to make litigated elections more likely in the future. One provision would allow pre-election challenges to voter eligibility to be granted without a hearing. In addition, there's a new provision allowing for hearings on late-filed challenges to be deferred until after an election. In both of these circumstances, voters who are challenged would still be entitled to cast provisional ballots. The net effect, however, will likely be more voters casting provisional ballots rather than regular ballots on election day, thus increasing the margin of litigation – and with it the chances of the parties' lawyers fighting things out in court after the election.
While the bill eliminates partisan "challengers" on election day, a change that will probably decrease the margin of litigation, election judges would still be permitted to challenge voter eligibility at the polls. Among the bases on which a voter's eligibility may be challenged is that he or she is not a citizen of the United States. Naturalized citizens may be required to produce their naturalization papers.
Imagine, for example, a 75-year old woman from China who gained her citizenship decades ago and has been voting at the same precinct for 30 years. Under the proposed bill, that citizen would be required to show her naturalization papers if challenged at the polls – even though she probably would never think to bring them with her, having voted without incident for so long. The voter would still be allowed to cast a provisional ballot, but it appears that the ballot would only be counted if she produces her naturalization papers within three days.
There are two problems with this provision. First, it's a recipe for racial and ethnic profiling, which may have the effect of intimidating and denying access to eligible voters. Second, it's likely to result in more people casting provisional ballots, thereby increasing the margin of litigation. When voters are forced to cast provisional instead of regular ballots, the opportunity for post-election litigation grows.
Not all components of Ohio's bill would increase the margin of litigation. One of the good things in the bill is a provision requiring that counties send their voters postcards notifying them of their polling places, at the time of registration and 45 days before each general federal election. This reform can be expected to decrease the margin of litigation, by making it more likely that voters will show up at the correct polling location, and therefore will cast rather than provisional ballots.
On the other hand, a related provision of the proposed bill would increase the margin of litigation. When the county postcards are returned as undeliverable, those voters' names must be "marked" in the pollbook. While a returned postcard may indicate that a voter has moved, it may also reflect a mistake on the part of the post office or county registrar. Nevertheless, a voter whose name is marked would be required to show ID and to cast a provisional ballot. Thus, the predictable effect of the requirement that voters' names be "marked" is that more provisional ballots will be cast – and that the margin of litigation will increase.
Fortunately, it now appears that Ohio's election bill will be tabled until the fall. Between now and then, state legislators would be well-advised to assess whether the proposed changes will make post-election lawsuits more or less likely. To be sure, finality is only one of the values that should be considered in deciding whether to amend election laws. Some reforms that increase the margin of litigation may still be worth enacting, because they promote electoral integrity or voter access. Still, it's vitally important that, in considering proposed amendments to election laws, legislators throughout the country consider the impact on their proposals on the margin of litigation.