In the US, state voter identification (ID) laws are hotly debated by politicians. Some people believe they are necessary to stop voter fraud, while others argue they only make it easier for voter discrimination to occur.
As with any politically-charged debate, the truth is hard to come by. But thankfully, data scientists are here to help.
Two researchers from Tufts University and Harvard University have developed a new and precise statistical method that is able to accurately determine the scope of discrimination caused by voter ID laws.
Recently, the researchers applied the method to a 2011 Texas voter ID law, which was investigated by the US Department of Justice for being potentially discriminatory. The legislation was deemed one of the strictest voter ID laws in the country, and it was immediately taken to court.
Working on behalf of the US in its litigation against the state of Texas, the two researchers developed a robust algorithm to figure out how this new law really impacted voters.
The findings are worrisome.
"Our evidence suggests a smaller number of people lack ID than recent survey evidence suggests, and it also suggests a discriminatory effect of the law, in line with concerns of those who believe these laws disproportionately affect minorities," said co-author Eitan Hersh, an associate professor of political science at Tufts University.
"Specifically, we found that white registered voters are significantly more likely to possess a voter ID than African-American or Hispanic voters."
As a result, the courts ruled that such a voter ID law was illegal under Section 5 of the Voting Rights Act.
Wow, imagine that: hard data informing policy.
The statistical method was designed to link individual records with databases, which sounds easy but is surprisingly challenging - mostly because these databases are huge.
The researchers had to compare 13 million records on the Texas Election Administration Management (TEAM) voter file with the 25-million-record file of State of Texas IDs and various national lists that are maintained by the federal government.
Plus, these absolutely massive databases have a ton of errors and missing values that make it difficult to match records to social security numbers.
But the overwhelming challenge did not stop Hersh and his colleague Stephen Ansolabehere.
Using a combination of factors, including address, date of birth, gender and name, the two researchers were able to match records about as accurately as if they had the individual's social security number.
That's pretty damn precise.
"Voter identification laws have become a source of debate and controversy, and we think this is the strongest evidence to date about the magnitude and the discriminatory effect of laws on protected minorities versus white voters," said Hersh.
Before the creation of this statistical method, voter ID laws used to be analysed using surveys. And as we all know after last year's election, surveys have serious limitations.
Hersh and Ansolabehere believe the ability to link records with a high degree of accuracy is informative to research in general.
They hope to apply their method not only to voter ID laws, but also to a whole range of issues, including public health, criminology, marketing and government censuses.
The study was published in the journal Statistics and Public Policy.