Many County Jails Don’t Provide Voting Access to People Detained: ‘It’s an Indictment of Our Criminal Justice System’

Research by the Public Safety Lab at New York University shows incarceration during election season leads to a large decrease in ballots cast.

By Matt Keyser
National Partnership for Pretrial Justice
Oct. 14, 2022

By the best estimations, around 11 million Americans are booked into county jails each year.

Many are held pretrial as they await case resolution – still innocent under the law. Others are serving sentences for misdemeanor convictions. In no state does incarceration pretrial or a misdemeanor sentence result in the loss of the right to vote.

But new research by the Public Safety Lab at New York University shows that few county jails provide voting access to those detained, resulting in a drastic decrease in votes cast by people incarcerated during election season.

“We need to call on elected representatives … to be aware of this problem,” says Anna Harvey, director of the Public Safety Lab. “This is about our democracy and about individuals’ constitutional rights.”

We spoke with Harvey about her team’s findings and why ignoring incarcerated individuals’ right to vote is an “indictment of our criminal justice system.”

This interview has been edited for length and clarity.

NPPJ: Anna, there are many great points in your research I want to discuss. But first, let’s talk about your main finding: People who are detained in county jails on Election Day or the days and weeks leading up to an election are less likely to vote. Why is that?

AH: We book a lot of people into county jails in this country. It’s hard to get a precise number because we don’t actually keep daily jail bookings on a national level. The Bureau of Justice Statistics does a Census every five years of everybody who’s sitting in a county jail on a particular day and they report that figure, but that’s only one day every five years. The best estimate we have is that about 11 million individuals are booked into county jails each year.

A lot of these individuals incarcerated in county jails are held pretrial — meaning they have yet to be convicted of a crime — or they’re serving misdemeanor sentences but still retain the right to vote.

We built a database of jail records from counties across the country and compared them to voting records, all of which are publicly available. We merged the voter data with the jail data and looked at the turnout of people who are booked into these county jails during voting days.

We found a really steep decrease in turnout. Looking within jails and averaging these decreases across all of the jails in our sample, we found there was a 46% decrease in turnout among registered voters incarcerated during voting days in the 2020 election.

NPPJ: We know that communities of color, specifically Black communities, which have been overpoliced for decades, make up a larger portion of jail populations. What does the data show relating to any racial discrimination?

AH: When we look at the effects separately by the race of the incarcerated individual, it’s even more troubling. We found that there was a 78% decrease in turnout among Black registered voters who are incarcerated during voting days. The effect is about a 40% decrease for white voters.

NPPJ: I want to stress a point you made: County jails are largely detaining people pretrial or those who have been convicted of a misdemeanor offense. We’re not talking about people in state prisons who are serving felony convictions and, in most states, aren’t allowed to vote.

AH: That’s absolutely right. Virtually everybody held in county jail does not lose their right to vote simply because of that detention. There have been concerns raised by organizations that work on behalf of incarcerated individuals that those detained in jail during voting days — even though they have not lost their legal right to vote because of that detention — are not being given adequate opportunities to exercise their constitutional right.

If you’re incarcerated, somebody has to proactively ensure that you have the opportunity to vote, either via a mail-in ballot or some other means. And you, as an incarcerated person, can only exercise your right to vote if somebody proactively enables that to happen.

Counties have a legal responsibility to ensure that people who maintain the right to vote while in jail are able to exercise that right. Our findings suggest that many counties are not, in fact, enabling the exercise of the right to vote for incarcerated individuals.

NPPJ: How would a county provide people detained the ability to vote? Is it mail-in ballots or setting up voting booths within the jail?

AH: There’s an organization called The Sentencing Project that has produced a couple of policy reports on this. They’ve found that in some counties, those who are in charge of the administration of county jails take steps to allow voting. They make sure that incarcerated individuals have paper ballots and then, importantly, that those ballots get mailed back to the county. It’s one thing to provide the ballots, but you have to secure the ballots, distribute them, and make sure there’s a way for people to mail them back.

Some counties also install voting machines in the jail.

But, again, our findings suggest that those methods or comparable methods are not consistently being used by counties to help people exercise their legal right to vote.

NPPJ: This research must have been a massive undertaking.

AH: In about 1,300 counties across the United States, I can find a list of people who are being held in the county jail today with their names, birthdates, their race or ethnicity, their charges, and other information that is publicly available. The problem is jail rosters change daily with new individuals being booked into jails and others released.

We realized that if we could write scripts to scrape these county servers every day, we could build a database of jail bookings in these 1,300 counties. We’d be able to see when somebody entered jail and when they left, and we would have information about charges, bail amounts, and other information.

It was a big undertaking because we had to write an individual, customized script for each jail, and then we had to standardize all these scripts so we could put everything in a database. It was an enormous amount of work. We have a website now,, where you can access and visualize the data. You can select counties or states or time periods, and you can look at the numbers of individuals booked into jail, the length of stay, and many other features.

Having all this data allows us to ask questions about the impacts of jail incarceration. There are, give or take, 11 million bookings into county jails each year. That means that in the two months prior to an election, there are about two million people who are booked in and out of county jails.

NPPJ: What takeaways can we learn from your research to help shape policy around elections?

AH: We need to call on our elected representatives in state, county, and municipal governments to raise awareness about this problem. It’s a problem we should all care about, despite what our partisan preferences are. This is about our democracy and about individuals’ constitutional rights.

We also need to ask our local media outlets to publicize this issue and to keep attention on elected leaders, sheriffs, and jail staff to ensure that men and women in jails are given the opportunity to exercise their right to vote.