Guess what? Ohio ranks especially vulnerable.
As part of my work on post-voting disputes over which candidate has won an especially close election—think Bush v. Gore or Coleman v. Franken, to cite the two most prominent recent examples—I have been contemplating the concept of “uncounted ballots”. It’s a bit different than the idea of a “residual vote,” which involves a counted ballot that records no vote for a particular race. An “uncounted” ballot, by contrast, is a cast ballot that it is not part of the initial unofficial returns on Election Night.
Conceptually, the three biggest categories of uncounted ballots are: (1) late-arriving absentee ballots that are potentially still eligible(mostly military and overseas ballots); (2) rejected absentee ballots that may have been rejected in error; and (3) provisional ballots that still need to be evaluated. Together, these ballots form a rough measure of a state’s vulnerability to a post-voting dispute in the event of a close result on election night: the larger the number of these uncounted ballots, the greater the possibility that the candidate behind on election night might be able to convert the apparent defeat into a victory. (One could combine the two categories of “uncounted” and “residual” votes, if one thought that the technology in a particular state made it likely that “residual” votes would be a fruitful avenue of litigation.)
The 2008 Election Administration and Voting Survey released by the U.S. Election Assistance Commission contains some data by which one can begin to assess a state’s vulnerability in this regard. From what I can tell so far, there is no reported data specifically on the number of absentee ballots counted after the polls close on Election Day. Still, there is data for each state on the number of rejected absentee ballots and the number of provisional ballots. Thus, one can combined these two numbers for a state to get a sense of how large a pool of uncounted ballots might be fought over if a statewide election were close enough to be worth fighting over.
By this metric, perhaps not surprisingly, Ohio is an especially vulnerable state—the most vulnerable of the likely “battleground” states in a future presidential election. In 2008, Ohio had 25,950 rejected absentee ballots, and 204,651 provisional ballots. By contrast, Missouri had only 5,403 rejected absentee ballots and 6,934 provisional ballots. The 2008 presidential election in Missouri was extraordinarily close (McCain ended up winning by 3,903 votes), but if the entire presidency had turned on the outcome in Missouri, its exposure to litigation would have been lower than Ohio’s, just because it had much fewer uncounted ballots to fight over.
California, even recognizing its much larger population, also has a high number of uncounted ballots: 130,730 rejected absentee ballots, and 798,332 provisional ballots. Although California may not be a presidential battleground, if it has a close Senate or other statewide election, numbers of this magnitude may prove troublesome.
The New York Times recently has stated that Ohio will be one of the most interesting states to watch in the 2010 midterm elections, with potentially competitive gubernatorial, congressional (including U.S. Senate), and judicial and other major races. There is legislation pending in the state legislature aimed at reducing the number of provisional ballots and thus Ohio’s exposure to post-voting litigation in a close statewide race. But if the legislation does not pass (each party controls one of the two chambers, and so far they haven’t been able to compromise on election reform), then perhaps it would actually be beneficial for the state to suffer the shock of a post-voting dispute over absentee and/or provisional ballots in 2010. Agony next year might pave the way for improved conditions by 2012. Although I wouldn’t wish another Coleman v. Franken on any state, it would be better than another Bush v. Gore.