Why Turnout in Virginia Wasn't Quite as Bad as You Think

Part of the quickly congealing convention wisdom of Tuesday's election is that turnout was down dramatically in Virginia. Democrats lost because they didn't show up. While that's certainly true to some extent, I think the case has been overstated a bit.

Here's what Larry Sabato had to say:

Turnout played a huge role in the outcomes in both NJ and VA, with Republicans showing up in droves and Democrats going fishing, at least to some degree. In Virginia, one result of absentee Democrats was the lowest voter turnout for a gubernatorial election in the state's modern two-party history (1969 to 2009). The 2009 turnout of 39.8 percent of the registered voters was the lowest in forty years. Even with all the population growth since 2005, the absolute voter turnout in 2009 (1.97 million) fell below that of four years ago (2.0 million). And the electorate was barely more than half that of 2008 (3.7 million). Astounding.

Sabato's stats do show a sharp drop in the turnout rate for registered voters compared to previous Virginia gubernatorial races: This year, 5 percentage points fewer registered voters showed up than in 2005, a drop of 11.6%. That's quite a lot. The drop was even steeper compared to 2001.

However, it's worth noting that there were 500,000 more registered voters in Virginia this year than in 2005. Population growth only partially explains the change. The 2008 presidential election, in which both parties (especially Democrats) aggressively tried to find new voters in the state, contributed. For that reason, I think measuring turnout based on registered voters is somewhat misleading.

How do I know that the increase in registered voters wasn't entirely because of population growth? Because Virginia's voting-eligible population (VEP) -- the numbers of adults who aren't non-citizens or disenfranchised felons -- has only increased by approximately 225,000 since 2005. VEP is the truest measure of everyone who might have voted in an election, which is why I think it's the best standard by which to judge turnout.

Michael McDonald is a turnout guru at George Mason University who estimates each state's VEP in every even-numbered year. Based on his numbers, I came up with simple estimates for Virginia's VEP for its recent gubernatorial elections, which, of course, were in odd-numbered years. I assumed that 2001had a VEP that was the mid-point between 2000 and 2002. For 2005, it was the mid-point between 2004 and 2006. For 2009, I just assumed that the VEP had increased at the same annual rate as between 2006 and 2008.

Using those numbers, here's a chart that displays turnout based both on VEP and on registered voters.

As you can see, turnout was down even based on VEP. But, it only dropped by a couple of percentage points. That's not nearly enough to explain Bob McDonnell's 17-point victory, unless Democrats showed up in unusually small numbers and Republicans showed up in unusually large numbers (which is probably what happened).

All of that being said, there is a case to be made that the turnout stats based on registered voters are relevant too. By registering last year, hundreds of thousands of Virginians showed for the first time that they were interested in politics. As a result, you might have expected turnout to increase this year based on VEP. Perhaps it isn't surprising that people who weren't even registered to vote prior to 2008 didn't show up in large numbers for a state election, but Democrats almost certainly would have been better off if they had.