Precinct closures harm voter turnout in Georgia, AJC analysis finds

Amid widespread voter distrust of government oversight of elections and questions about ballot access, The Atlanta Journal-Constitution conducted a unique statistical analysis to learn how precinct closures and distance to the polls impact voting.

The AJC mapped Georgia's 7 million registered voters and compared how distance to their local precincts increased or decreased from 2012 to 2018. During that time, county election officials shut down 8% of Georgia's polling places and relocated nearly 40% of the state's precincts.

Most of the precinct closures and relocations occurred after the U.S. Supreme Court in 2013 ended federal oversight of local election decisions under the Voting Rights Act.

The AJC’s analysis, vetted by two nonpartisan statistics experts, showed a clear link between turnout and reduced voting access. The farther voters live from their precincts, the less likely they are to cast a ballot.

Precinct closures and longer distances likely prevented an estimated 54,000 to 85,000 voters from casting ballots on Election Day last year, according to the AJC’s findings.

And the impact was greater on black voters than white ones, the AJC found. Black voters were 20% more likely to miss elections because of long distances.

Without those precinct relocations, overall Election Day turnout in last year’s midterm election likely would have been between 1.2% and 1.8% higher, the AJC estimated.

“Seems to me, they’re making it harder for us to vote,” said Coleman, who voted in the November election for governor but didn’t cast a ballot in the primary. “I hate that they closed that place down because it was more convenient. Maybe I wouldn’t miss elections if it was still open here.”

The AJC’s analysis accounted for both large, rural precincts and small, urban precincts by measuring how far voters had to travel as a percentage of their precinct’s geographic area. Both groups were impacted, the AJC found.

The average Georgia voter’s distance to a polling place more than doubled from 2012 to 2018, according to the AJC’s analysis.

Still, in many ways, voting has never been easier in Georgia.

Georgia leads the nation in automatic voter registration, with more than 350,000 new voters signed up when they obtained their driver's licenses since last year's election. In addition, the state provides three weeks of in-person early voting and voting by mail for anyone who requests a ballot.

There are a record 7.4 million registered voters in Georgia, though about 300,000 of them are scheduled to be canceled this month because they've moved or haven't cast a ballot since 2012.

Secretary of State Brad Raffensperger said the convenience of voting resulted in record turnout during last year's midterm election, with 57% of registered voters participating.

He predicted heavy turnout, more than 5 million voters, in next year's presidential election as well.

“The General Assembly rightly gives county officials the decision about how many polling places and where they are located. They know best about traffic patterns, the needs of their citizens and their county budget,” Raffensperger said. “My goal is to give them perspective on turnout so they can make wise decisions.”

Despite the popularity and ease of early voting, however, in-person voting on Election Day is still preferred by nearly half of the electorate, so where and how many polling locations election officials choose to make available matters to millions of Georgia voters.

Decisions to close or relocate precincts, often to save tax money, can be controversial.

Randolph County in Georgia drew national attention when local election officials proposed closing seven of the county's nine precincts in 2018, a proposal that critics said would make it harder on voters in the largely black county. Officials backed off the proposal but closed three precincts in mostly white areas this year.

Though tens of thousands more people likely would have voted last year if their precincts were closer, according to the AJC's analysis, they almost certainly wouldn't have changed the outcome of last year's election for governor. Republican Brian Kemp defeated Democrat Stacey Abrams by 54,723 votes. Even with ideal voting locations, Abrams would have had to have won between 82% and 100% of those additional votes to close the gap, or 61% to 67% of those votes to force a runoff.

High court fallout

It wasn’t always so easy for county governments to make electoral changes that reduced voting access.

In 2013, the Supreme Court split 5-4 along ideological lines, ruling that Georgia and eight other states with a history of discrimination no longer had to obtain federal approval before making electoral changes, including eliminating or moving voting locations. The ruling also applied to select cities or counties in six other states.

The court's decision in that case, Shelby County v. Holder, resulted in more states implementing voter ID laws, and voter registration cancellations rose 33% over the next two years. For example, Texas implemented a voter ID law the same day as the court's ruling, and Alabama implemented a photo ID law within days. Georgia has required photo ID when voting since 2006.

The VRA and its permission requirement, known as “preclearance,” were introduced during a televised event in 1965 by President Lyndon B. Johnson in response to “Bloody Sunday,” when Alabama state troopers attacked unarmed voting rights marchers during their famous march from Selma to Montgomery. Congress passed the landmark legislation less than five months later.

Preclearance required notification to communities about planned electoral changes, fact-finding to show that minorities communities wouldn't be harmed, and feedback from local minority leaders, said Jon Greenbaum, chief counsel for the Lawyers' Committee for Civil Rights Under Law, a civil rights organization.

Under preclearance, changes that hurt minority voting rights by even the barest statistical margin weren’t allowed.

Once freed from federal oversight, precinct closures accelerated in areas previously covered by the Voting Rights Act. At least 1,688 polling places were shut down since 2012, according to the Leadership Conference Education Fund. The AJC reported last year that 214 of those precinct closures were in Georgia, third most of states previously covered by the act's preclearance provision.

Before the ruling, voters of all races were barely affected by their distance to the polls, accounting for a 0.2% and 0.4% reduction in turnout, according to the AJC’s analysis of election data from 2012. The number of Georgia voters who missed elections because of distance more than quadrupled in 2018 compared to 2012, the AJC found.

Turnout by black voters would have been between 1.3% and 2.1% higher on Election Day in 2018 if they all lived near their polling places.

Overall, black voters are also significantly more likely to live farther from their precincts than white voters, the AJC found. About 30% of black voters must now travel across half of their precinct to reach their poll compared to less than 20% of white voters.

The AJC’s analysis shows the impact of the Supreme Court’s ruling, said Donald Verrilli Jr., the U.S. solicitor general at the time of the court’s decision in 2013. The court’s majority said the Voting Rights Act covered states based on their history rather than on recent evidence of discrimination.

“This is exactly the kind of updated data the justices in the majority said was lacking,” Verrilli told the AJC. “Exactly the kind of data that suggests that the judgment of the majority of the court — the South has changed — may be in need of amendment. Maybe the South hasn’t changed as much as one would have hoped.”

A bill that passed the U.S. House of Representatives this month would restore federal supervision requirements of the Voting Rights Act. States with repeated voting rights violations in the previous 25 years would have to obtain approval for changes in their election laws.

The legislation passed along a mostly party-line vote in the House, where Democrats hold a majority. The bill now advances to the Republican-controlled Senate.

“Voters shouldn’t have to make a decision between casting their ballot and picking up a child from school or taking time off from work,” said Leigh Chapman, the director of the voting rights program for The Leadership Conference Education Fund, a civil rights organization. “We need to make sure we’re preventing potentially discriminatory policies and laws from going into effect before they’re harming voters.”

‘Up to the voter’

The erosion of voting locations takes place at the county level, especially in rural areas with tight budgets that can save money by closing precincts.

Fewer precincts are needed because so many people vote in advance, either in person or by mail, said Melessa Shivers, the election supervisor in Clay County.

“In three weeks of early voting, I find it hard to believe that you wouldn’t be coming to Fort Gaines in that time,” said Shivers, referring to Clay’s early voting location. “And a homebound person is already voting an absentee ballot. It’s every bit up to the voter to exercise their right to vote. We can’t go to their house and bring a ballot to them or carry them to the polls.”

Shivers, who oversaw the consolidation of five precincts into one in 2015, said there was little opposition when she proposed them to the county election board. Some of those cement block precincts needed repairs, lacked accommodations and poll workers didn’t want to work there, she said.

The Rev. Shirley Cody, a Methodist pastor in Clay County, said voters in her church wanted to keep their local precincts.

Cody organized two drivers to bring voters from her church to the polls last year, helping eight people on Election Day 2018.

“I knew if we didn’t give them rides, that would be votes we didn’t have from this area,” Cody said.

Other voters said distance to the polls doesn’t deter them from exercising an American birthright.

“It doesn’t matter to me — you have to show your support,” said Junior Pridgen as he fed cows at his home in rural Ben Hill County. His precinct is located 17 miles from where he lives. “We’ve got to have the right representation in the House, Senate and president’s office.”

Urban voters impacted, too

The effects of precinct relocations are also felt in metro Atlanta, where voters are located closer to voting locations but they often have to contend with severe traffic to get there.

Penny Reid, an 80-year-old Gwinnett County resident, said she always votes by mail because she doesn’t want to fight traffic and long lines before casting her ballot. Her precinct is about 2 ½ miles from her home, across busy Jimmy Carter Boulevard and farther than her old voting location at a Methodist church.

“It creates obstacles for people,” said Reid, whose bad knees limit her ability to stand in line to vote. “People in the neighborhoods around me don’t have easy access to voting.”

Distance from the polls creates unequal classes of voters. Voters who live in counties with more accessible precincts have higher turnout and therefore more voting power than those who live far away.

"You really have a system that has disincentivized people from voting," said Rich Levy, founder of Galvanize Georgia, a group that encourages voter participation and education. "It's getting harder, not easier."

In Clayton County south of Atlanta, Harold Lewis said he was able to vote in last year’s election after some hassles.

Lewis walked two blocks to try to vote at an elementary school near his apartment, but he then learned his assigned precinct was 5 ½ miles away at a senior center. Lewis, who doesn’t have a car, had to find a ride to the polls and take a bus back home.

“There are a lot of hurdles, but you can’t give up on it,” said Lewis, 51, who works in aviation. “It should be easier.”

Why precinct closures matter

Research by The Atlanta Journal-Constitution last year revealed that 214 Georgia precincts shuttered between 2012 and 2018, representing a loss of about 8% of the state’s polling locations. These precincts closed after the U.S. Supreme Court lifted a requirement for federal approval of changes in voting access in 2013.

Those precinct closures made a difference in voting behavior, according to a new statistical analysis by the AJC.

Registered voters were less likely to cast a ballot on Election Day last year when they lived farther from their voting precincts. Voters of all income levels and races were deterred from voting when they had less access to their voting location.

As a result, turnout in last November’s election for governor was at least 54,000 voters lower than it would have been if all voters lived relatively close to their polls, the AJC found.

Methodology for the analysis

AJC data specialist Nick Thieme obtained the data and performed the statistical analysis that produced the findings for this story.

Thieme used data from publicly available voter registration lists provided by the Georgia Secretary of State, economic data released by the U.S. Census Bureau, and the Georgia General Assembly’s digital maps of voting precincts to estimate the number of people who were less likely to vote because of their distance to their polling place.

He identified the location of nearly every voter’s address, determined each voter’s census tract, and assigned each voter their census tract’s median income. He also determined each voter’s voting precinct and calculated how far every voter was from their polling place.

Because of urban density and traffic, however, a voter can travel 20 miles in rural Georgia much faster than a voter can travel 20 miles through Atlanta during rush hour. To account for this, Thieme corrected distances for urban density by adjusting the calculated distance between a voter’s home and their polling place by the size of their precinct.

The impact of distance on voter behavior also largely depends on age, race, income, county, prior voting history and education.

To account for other variables, Thieme first grouped voters by their age, race, income, county and prior voting history. He then calculated turnout in those groups for those near and far from their polling places, taking the difference between those rates to be the effect of distance on voting.

M.V. “Trey” Hood III, professor of political science at the University of Georgia, Regina Nuzzo, senior adviser for statistics communication at the American Statistical Association, and Hunter Glanz, a professor of statistics at California Polytechnic State University, supported this article with vetting, suggestions and discussions of the methodology.