The Electionline Briefing of Dec. 2004 strikes a tone favorable to statewide registration as it reports that, “Statewide registration databases were used in 16 states
and the District of Columbia, making for a smoother election process by reducing the number of double registrants and better tracking voter movement between jurisdictions.” As we will see below, however, in Georgia “better tracking” may be the best explanation for why the state tossed out 70 percent of provisional ballots cast.
According to the electionline database the 16 states with statewide registration are: Alaska (1985); Arizona (2004); Connecticut (2005); Delaware (1990); Georgia; Hawaii; Kentucky (1973); Louisiana (1987); Massachusetts (1993); Michigan (1998); Minnesota; New Mexico; Oklahoma; South Carolina (35 years); South Dakota; and West Virginia.
Pennsylvania reports that 56 of 67 counties are covered so far.
Georgia offers an interesting example how the use of a statewide database may make it more difficult to get your vote counted. In Georgia, 70 percent of the state’s nearly 13,000 provisional ballots were tossed out. And electionline speculated that one reason for the high rate of rejection might be the use of a statewide database to more easily identify out-of-place voters (p. 5). The possibility that Georgia’s database resulted in more effective post-election “discipline” is just the kind of thing that worries us.
We would rather see a statewide database used to explain higher rates of participation and success in having votes counted. It is what we call the frontend-backend test of registration technology. Did the statewide database enable voters to more easily vote or did it enable state officials to more easily discount votes after they were cast?