RELEASE: The Racially Discriminatory Effects of Provisional Ballots

Washington, D.C. — One week ahead of the 2014 midterm elections, much has been made of the new, more suppressive voting laws in states across the country. These laws often disproportionately impact communities of color and could disenfranchise hundreds of thousands of voters, but new data show that the votes of those who make it to the voting booth may not be counted either. A first-of-its-kind look at 2012 data from the Center for American Progress has identified 16 states where the casting of provisional ballots significantly correlated with high shares of minority and non-English-speaking voting populations. Meanwhile, an increase in restrictive laws could mean that provisional ballots will play a key role in competitive 2014 elections.

There were 2.7 million provisional ballots filed in the United States in 2012, and nearly one-third of those were not counted. Rejections happen for many reasons, including cumbersome voter registration procedures, restrictive voting laws, poorly maintained voter lists, election office mismanagement, and voter error. But the analysis found a statistically significant correlation between minority voting-age population and the number of provisional ballots cast at the county level in 16 states: Arizona, California, Colorado, Kansas, Maryland, Montana, North Carolina, Nebraska, New Jersey, New Mexico, New York, Ohio, Oklahoma, Pennsylvania, South Dakota, and Utah. The list of states and the number of provisional ballots cast—nearly 2 percent of all in-person ballots—suggest the option is being used beyond what was originally intended.

“Provisional ballots were designed to be a fail safe to ensure voters can have their voice heard under unusual circumstances,” said Michele Jawando, Vice President for Legal Progress at the Center for American Progress. “The sheer number of provisional ballots being cast suggests that some states are using them in place of effective election administration. That nearly 25 percent of these ballots are rejected entirely suggests that more than 500,000 Americans are being silenced on Election Day, and they are often minority voters.”

Many states use provisional ballots for different reasons. New York and New Jersey saw high numbers of provisional ballots in 2012 because both states allowed voters displaced by Hurricane Sandy to cast provisional ballots at any polling location in the state. But a state such as California—where all five of the demographic groups tested were disproportionately affected by provisional balloting—shows that while laws lowering barriers to access certainly help with access, there is still work ahead to ensure eligible voters are enfranchised at polling places.

However, instead of addressing issues such as poll worker education and more-progressive voting laws, many states have gone in reverse since 2012. Several states have implemented more-restrictive voting laws, including adding voter ID requirements, cutting same-day registration and early-voting days, and even limiting provisional balloting in North Carolina. Georgia is currently struggling with questions about unaccounted for voter registration forms that could lead to the casting of provisional ballots, making the state’s failure to report their use of provisional ballots in 2012 troubling. Five other states were not required to report this data because they have same-day registration or only one jurisdiction, in Alaska’s case. No reason was given for Georgia’s failure to report. Colorado is one of a few states that have improved voting laws since 2012, including moving to mail-in balloting.

“It is important to recognize that this analysis only looks at one way a voter can be left out on Election Day,” Jawando said. “In many states, strict laws prevent people from ever even entering a polling place to cast a ballot. Our voting laws should ensure that all eligible voters have open access to voting, and we need more research to understand the impact of more-restrictive voting laws. Today, there are too many obstacles for too many voters, preventing hundreds of thousands of voices from being heard in our elections.”

Background

The analysis found a statistically significant correlation between minority voting-age population and the number of provisional ballots cast at the county level in 16 states: Arizona, California, Colorado, Kansas, Maryland, Montana, North Carolina, Nebraska, New Jersey, New Mexico, New York, Ohio, Oklahoma, Pennsylvania, South Dakota, and Utah. It tested for relationships between the allocation of provisional ballots and five county-based demographic variables:

  1. Percentage of citizen voting-age population that is African American
  2. Percentage of citizen voting-age population that is Hispanic
  3. Percentage of citizen voting-age population that is Asian
  4. Percentage of citizen voting-age population that is overall minority
  5. Voting Rights Act Section 203 counties—those with significant voting-age populations who speak a language other than English

In 2012, 2.7 million provisional ballots were submitted, up from 2.1 million in 2008. Of those, 24.1 percent were rejected entirely, and 6.7 percent were only partially counted—meaning not all of the races on the ballot were counted.

For more information or to talk to an expert, please contact Benton Strong at 202.481.8142 or bstrong@americanprogress.org.

###