When the dust settles, there are likely to be a number of interesting observations to be made about the role of provisional voting in the unprecedented quasi-national primary that occurred yesterday.
For example, at Election Law @ Moritz we’ve been following reports that folks who wanted to cast a party ballot were denied the opportunity to do so—even provisionally—on the ground that they were not party members as required by state law. For reasons my colleagues explain, that denial seems contrary to both the letter and spirit of HAVA, which embraces the basic notion that no one who sincerely believes they are entitled to vote should be turned away without the opportunity to cast a provisional ballot. The ballot, of course, won’t count if the voter is incorrect in this belief. But it is so much better to resolve the immediate dispute between voter and poll worker by letting the voter record his or her electoral preferences, which will remain in limbo until eligibility is ultimately determined, than to send the voter home with no recourse in the event (however unlikely) that the poll worker is mistaken. If Super Tuesday is evidence that this philosophy has yet to become second nature among poll workers, then one lesson is that more work must be done to inculcate this mindset.
Here, however, I want to focus briefly on the report that 10-12% of all ballots cast yesterday in New Mexico were provisional. That is an astonishing figure, much higher than what we saw in 2004 or 2006. In a chapter on provisional voting that I’ve written for a forthcoming book on U.S. elections, I observe that in November 2004, Alaska had the highest rate of voting provisionally, at 7.5%. Three states were over 5%, six over 3%, and 16 over 1%. In November 2006, rates of voting provisionally declined, with only 12 states over 1%. (Alaska was again the highest, this time with only 5%, and only three other states were at 3% or higher: Arizona, California, and Ohio.) Presumably, the general decline in provisional voting from 2004 to 2006 was caused, at least in part, by improvement in the maintenance of state voter registration lists. (The chapter’s discussion of these and related figures is at pages 12-17 of the draft.)
Consequently, the 10-12% reported in yesterday’s New Mexico voting is an outlier by these other post-HAVA figures. It is unclear to me at this point what might have caused this remarkable spike in provisional voting there. (New Mexico’s rate of provisional voting for November 2006 was 0.5%.) This news story that reports the figure does not provide an explanation, other than to discuss the unexpectedly large turnout generally. Perhaps the explanation lies in part in voter confusion over party affiliation requirements (in which case New Mexico, unlike elsewhere, would be permitting the use of provisional ballots in this context).
I raise this possibility with no pretense that it will prove accurate, but only to underscore that this phenomenon is worthy of further investigation and analysis.
I look forward to hearing more reports on exactly what went on with provisional voting in New Mexico—and elsewhere—on Super Tuesday.