The Nonprofit Voter Engagement Network has available for download a report on "Voter Turnout in the 2006 Election." Political junkies will not find much novel in it, but nonetheless may be interested. It is constrained by the internal contradiction in such a project: nonprofits are legally barred from partisan participation in politics. Consequently it treats all enhanced voter participation as an unalloyed good, without regard for who benefits. Given the outcome in 2006, it is not hard for a progressive to feel great about additional participation, and I do. But this report necessarily comes at politics bass-ackward -- in the real world, politics is about who wins and who loses, or, at its most elevated, which policies win and which lose, so I take a "non-partisan" effort with a shaker of salt.
That said, at least this report is looking at what I think of as the right questions: where did the 2006 elections depart from historic patterns in midterm elections, why, and especially, what happened with the youth and people of color vote.
These authors conclude the most important variable to increasing participation is the presence of competitive races. However it is hard to make the case that Michigan (19 percent turnout gain) or Nebraska (17 percent) had particularly competitive years -- no Senate or House seats changed party in either state. All incumbents won. Their winning margins were solid. On the other hand, Indiana (9 percent), where the Democrats didn't even field a candidate against Republican Senator Richard Lugar, did have 3 incumbent Republicans defeated by Democratic House candidates. The report's case is stronger that simply having two statewide races, a Governor and a Senator, on the ballot, even if the races weren't particularly competitive (Michigan fits this picture) did raise turnout. It would be nice -- reassuring to proponents of democracy -- to be able to say that competitive races are the main determinant of increased turnout, but I don't think this report's data really proves this case.
On the other hand, the data do seem to be pretty clear: states that make it easier to vote because they accommodate election day registration (Idaho, Maine, Minnesota, Montana, New Hampshire, Wisconsin, Wyoming) have higher percentages of eligible voters exercise the franchise. From a partisan point of view, that is a mixed bag.
The report also highlights that turnout of voters under 30 was a full 2 percent higher than in 2002. This is a gain that probably is reliably progressive, because, within this age group, Latinos, Blacks and Asian-Americans made up a rising fraction of the whole. These communities will be part of the core of new Democratic majority.
Much the most useful part of the report for me was the section in which the authors examine various prescriptions for higher turnout, always putting them in the context of "where in place." The result is an interesting catalogue of locales that do such things as restore voting rights to felons, ensure disability and language access to ballots, and encourage early voting. While the report notes that discrepancies between local voting practices are one of the barriers prospective voters face in U.S. elections, as long as they exist, the more we know about them, the better. Uniformity is not the U.S. practice; negotiating the thickets of local practice is part of the grand game of electoral strategy.