The conventional wisdom on felony disenfranchisement laws is that they hurt the Democratic Party. The logic: A lot of the people locked up are black, and a lot of black people tend to vote for Democratic candidates. When Virginia Governor Terry McAuliffe decided to restore the voting rights for people with past felony convictions last August, Donald Trump called it “crooked politics” at a rally in Rhode Island: “They’re giving 200,000 people that have been convicted of heinous crimes, horrible crimes, the worst crimes, the right to vote because, you know what? They know they’re gonna vote Democrat,” he said. “That’s how disgusting and dishonest our political system is.”
The belief that restoring voting rights to ex-offenders would benefit the political left and their urban power base is a common one among conservatives, but a group of economists who recently examined the issue was unable to find much proof to support it. In their study, “The Voting Rights of Ex-Felons and Election Outcomes in the United States,” released March 17, economics professors Tilman Klumpp of the University of Alberta, Hugo M. Mialon of Emory University, and Michael A. Williams of the research firm Competition Economics wrote:
Focusing on elections for seats to the U.S. House of Representatives, we found a positive but statistically non-significant effect of ex-felon voting rights on the vote share of Democratic candidates.… no House majority would have changed in any year between 1998 and 2012, had all ex-felons been allowed to vote in all states.
In other words, allowing those who’ve already served their time for felony convictions to vote would not “swing” elections to the Democrats. According to the economists, this population has not voted in large enough numbers over the last 20 years to tip any kind of political advantage to either party. If there were no felony disenfranchisement laws, and former felons turned out to vote at the same rate as their demographic peers (mostly male, African Americans), this would give Democrats a slight bump—enough to pick up a few seats in Congress in certain years, but never enough to give Democrats the majority of districts in Congress, as seen in the chart below:
“Our finding that [voting] rights restoration has no tangible effects on election outcomes removes one potential political obstacle from reforming the criminal justice system towards one that places a greater emphasis on rehabilitation,” reads the study.
If only it were that simple. Virginia is one of four states that imposes a lifetime voting ban on those convicted of felonies. After serving out one’s sentence in these states, a person can only appeal to the governor, or some other state authority, to have those rights restored. Governor McAuliffe tried to remove that hurdle by signing an executive order that would automatically restore voting rights to anyone who’s served their sentence, but Virginia Republicans rebuffed the measure, taking their challenge all the way to the state’s supreme court, which disallowed it. Consequently, the governor is currently in the process of signing individual restoration orders for each of the 180,000 Virginians with past felony convictions—which will keep him occupied for a long while.
In Florida, which disenfranchises the largest portion of its voters, the Fair Elections Legal Network filed a lawsuit on March 13 against Governor Rick Scott over the law. The organization is arguing that the governor’s voting rights restoration process is arbitrary and violates the Constitution’s equal protection clause. Scott’s predecessor, Charlie Crist, eased the rights restoration process when he was in office, which allowed 150,000 people to vote again. Scott tossed that process out, however, when he took office, and put in place a more restrictive system. Only about 2,500 people have had their rights restored since that change was made in 2011. The Florida Supreme Court is currently considering allowing a ballot referendum that would let Floridians vote directly on whether or not the state’s felony disenfranchisement laws should remain in place.
In Kentucky, another of the four states with strict voting bans for former felons, the number of disenfranchised voters due to felony convictions has increased 68 percent since 2006.
It should be noted that while Democrats strongly support ending felony disenfranchisement, they don’t typically emphasize that the influx of voters might make their party stronger. Rather, the case for re-enfranchising voters with felony convictions is based upon the idea this will help those who’ve been incarcerated better reintegrate in society. “Bringing people into the political process makes them stakeholders, which in turn helps steer them away from future crimes,” writes Erika Wood of the Voting Rights and Civic Participation Project of the Impact Center for Public Interest Law in a December report on felony disenfranchisement:
The data gathered under the Scott rules supports this theory. The average recidivism rate in Florida has hovered around 30 percent for the last five years. In 2011, of the 52 people granted RCR, zero were returned to custody. In 2012, out of the 342 people granted RCR, only one re-offended. In 2013, out of 569 people granted RCR, zero re-offended. In 2014, of 562 people granted RCR, three reoffended. In 2015, of 427 people granted RCR, one re-offended. While there is not enough data to conclude that there is a direct correlation between restoring civil rights and decreased recidivism, the 0.4 percent average recidivism rate for those who have had their rights restored is eye-opening, and should be studied in more depth.
Not to mention, there’s an explicitly racist history behind why these laws still exist in the first place. That history is recounted in the “Free the Vote” report released by the NAACP Legal Defense Fund and the Sentencing Project in December:
The racial disparities caused by felony disenfranchisement laws are not a coincidence. Many felony disenfranchisement laws were revised in the years following the Civil War for the specific purpose of limiting the political power of newly-freed Black people. Indeed, many state legislatures tailored their felony disenfranchisement laws to require the loss of voting rights only for those offenses disproportionately prosecuted against Black people, or thought to be committed most frequently by Black people.
For example, guided by the belief that Black people were more likely to commit less serious property offenses than the more “robust” crimes committed by white people, the 1890 Mississippi constitutional convention required disenfranchisement for such crimes as theft or burglary, but not for robbery or murder. Through the convoluted reasoning of this law, one would be disenfranchised for stealing a chicken, but not for killing the chicken’s owner.
The “Free the Vote” report states that felony disenfranchisement is felt strongly at the state level—most especially in Florida, where one in five African Americans are unable to vote due to felony convictions. Last August, a group of political scientists from the University of Florida and Dartmouth University examined the effects of “disproportionate disenfranchisement” of African American on local races.
These effects aren’t addressed by the study the three economists released earlier this month (“The Voting Rights of Ex-Felons and Election Outcomes in the United States”). “We know from our previous research that the effects of certain laws can be magnified on the state and local level,” says co-author Hugo Mialon of Emory. “Of course, the counterargument is that ex-felons may be less likely to vote for state/local office candidates, compared to candidates for Congress who receive more media coverage. The big disadvantage of looking at state/local elections is that it’s very hard to get data on the number of votes. This information exists, of course, but not in one place.”
Until that data gets analyzed, the idea that one party gains a political advantage from allowing people who’ve paid for their crimes the right to vote—as is normal in the majority of states in the U.S.—should be considered empirically suspect.