The Center for Democracy and Election Management at AU (which has been headed by Bob Pastor, who was the driving force behind the Carter-Baker commission which recommended the use of voter identification at the polls) has issued a press release, “New Survey Suggests that Voter IDs Are Not the Problem,” about this report it has now issued based on telephone surveys of registered voters in Indiana, Maryland, and Mississippi.
The findings are at odds with the Barreto et al paper, which found a much higher number of voters without identification. Surely the new paper, like the Jeff Milyo paper will be seized upon by supporters of the voter id law as proof that the law is constitutional, just as opponents of the law have relied heavily upon the Barreto study. As I’ve noted, however, none of these papers have yet appeared in a peer-reviewed journal, and there is a great deal of controversy over the methodology used in these studies. One obvious potential problem with the CDEM study is that a telephone survey of voters will tend to oversample those who can afford telephones and who are home to answer calls rather than working at the time that the calls are made.
UPDATE: Michael McDonald writes:

Let me re-state my call for putting a properly worded question about photo identification on the 2010 Census. That should put all these conflicting studies to rest, and the number should be informative for other national security-related policymaking.
Yes, the percentages reported in the survey are small, but they are being applied to a large population. I find that the 2.5 million eligible voters that would be excluded from participating in elections (if the 1.2% number is extrapolated nationwide to my 2006 eligible population estimate) is not a number that merits the description of “only” and would “not be a problem.” The confidence interval on the survey is +/-4.5%, so we cannot reject with a reasonable standard of scientific doubt that that true population (of phone respondents) is really 5.7% or 11.8 million people (again, applied to the entire country).