The state will be one of the central battlegrounds this November, and it’s widely believed that ex-felons will vote heavily for Democrats. (More than half are African-Americans, a big voting bloc for the party.) And the big number of newly enfranchised voters — 200,000 — is actually larger than Mr. Obama’s 149,298-vote margin of victory there in 2012.
But the electoral effect of felon re-enfranchisement is likely to be modest. The best-case scenario for Democrats might be that they improve their popular vote margin by a half-point. That’s a big deal, but only in a close election.
The reason is deceptively straightforward. Ex-felons are less likely to vote than nonfelons, even when ex-felons are eligible to vote.
Part of it is demographics. Ex-felons are disproportionately young and less educated, the two most powerful demographic predictors of low voter turnout in the United States.
If disenfranchised ex-felons voted just like demographically similar nonfelons, their turnout rate might only be around 35 percent in a presidential election — far beneath the 60 percent or so of eligible voters who usually do vote, based on statistical models.
But there’s strong evidence that even this 35 percent figure is too high. Over the last decade, a string of studies has matched discharged ex-felons to voter registration files in the states that they’re permitted to vote. Although the figures vary, they usually find that around 20 percent of ex-felons turn out, even in presidential elections. Most of these studies were conducted with data from the 2008 and 2012 presidential elections, when black and youth turnout were particularly high.
Perhaps the most telling evidence comes from North Carolina, a state that’s demographically similar to Virginia. Multiple studies there have matched discharged ex-felons to the state’s voter registration file, which includes both the race and the party registration of voters.
One North Carolina study — from Marc Meredith of the University of Pennsylvania and Michael Morse of Stanford’s law school — looked at ex-felons who were discharged between the 2004 and 2008 presidential elections, a period when the Obama campaign’s vaunted ground operation mobilized hundreds of thousands of previously unregistered black voters.
It’s a time, therefore, when one would expect that ex-felons would have been particularly likely to register to vote, either as a result of mobilization efforts or because of the prospect of supporting the first black president.
The ex-felons were indeed overwhelmingly likely to register as Democrats. In North Carolina, they registered as Democrats by 55 percent to 10 percent. The tallies were similar in two other states where the study was conducted: New Mexico, at 52 percent to 19 percent for Democrats, and New York, at 62 to 9.
Yet in the end, just 33 percent of these newly eligible voters registered for the 2008 election in North Carolina, and just 21 percent voted. The registration and turnout rate in New Mexico and New York were far lower, with 13 percent turnout in New Mexico and 8 percent turnout in New York (for 2012; that study counted discharges from 2008 to 2012).
So what does that mean for Virginia? Imagine, for a moment, that the same figures for North Carolina played out in Virginia: 21 percent of the 200,000 re-enfranchised ex-felons turn out, and they vote Democratic by an 85-to-15 margin. The result would expand the Democratic vote margin by 29,400 votes. In the 2012 election, that would have increased Mr. Obama’s margin of victory to 4.6 percentage points from 3.9.
Again, this represents something of a best-case scenario for Democrats. The turnout rate among newly registered ex-felons was high: 69 percent among active registered voters.
The Obama campaign’s effort in North Carolina was aided by the state’s voter file, which includes the race of registered voters. That allowed the Obama campaign to aggressively target newly registered black voters ahead of the election. The campaign of the Democratic nominee won’t have that advantage in Virginia, where race is not listed in the voter registration file. This is not a small matter.
One thing that could help Democrats is if they’re able to obtain records on ex-felons in Virginia. They could send them voter registration forms or volunteers.
The ultimate electoral consequence is as much a matter of perspective as it is math. On the one hand, a half-point is fairly unlikely to decide an election. No state was so close in the 2012 presidential election. On the other hand, Virginia is the sort of state that could decide a close national election, and half a point could be a big deal.
There aren’t many things that can move the results of a presidential election by a half-point. The Clinton or Sanders campaign would undoubtedly spend millions of dollars to earn that kind of extra edge. Mr. McAuliffe has given it to them.