Electionline.org's Provisional Voting Report
Electionline.org has released this report on the implementation of provisional voting in the 2004 eleciton. The A.P. has this story. This is required reading for anyone interested in provisional voting or the implementation of HAVA generally. While the authors make clear that this isn't the "final say" on provisional ballots in 2004, the report offers some revealing and important information. Here are a few highlights.

- Nationwide, electionline.org finds that more than 1.6 voters received provisional ballots in the 2004 general election, and that almost 1.1 million were counted. Many of these voters would have been turned away, were it not for provisional ballots.

- The report finds wide variations in the way that provisional ballots were handled from state to state. In 18 states, provisional ballots were counted if cast in the proper county (or other registrar's jurisdiction); in 28, however, provisional ballots were counted only if cast in the correct precinct. The percentage of provisionals counted varied from a low of 6% (Delaware) to a high of 97% (Alaska). Unsurprisingly, states that counted provisional ballots cast out of precinct counted a higher percentage than those that did not.

- There were also wide variations in how many voters received provisional ballots, according to the HAVA report. For example, in Alaska, provisionals accounted for 7% of all ballots cast. By contrast, in Michigan, provisional ballots were less than 0.1% of the total vote. It's not completely clear, at least to me, which way this cuts. On the one hand, a high percentage of provisional ballots (per total voters) suggests that there may be significant problems with the state's registration list -- in other words, that a lot of eligible and registered voters arrive at the polls to find their names on the list. On the other hand, a very low percentage of provisional ballots issued suggests that there's less-than-complete compliance with HAVA's requirement that voters not be turned away, but instead receive a provisional ballot, if they arrive at the polls to find their names not on the list. Thus, my instinct would be that either an unusually high or an unusually low ratio of provisional ballots to total ballots is cause for concern.

- For those who think that the statwide registration databases, required by 2006, will reduce the importance of provisional voting ... it may not be so. States without registration databases counted a slightly higher percentage of provisionals (68%) than states with them (65%). - posted by Dan Tokaji @ 12:39 PM