Now the REAL Debate Can Begin: How DOJ’s SC Voter ID Objection FINALLY Brings Data to the Discussion


[Logo image courtesy of the U.S. Department of Justice]

Way back in late August, I blogged about the Voting Rights Act review by the U.S. Department of Justice (DOJ) of South Carolina’s new voter ID law. In that post, I said:

Regardless of the outcome, DOJ’s review of SC’s photo ID laws will give us something we haven’t much of in the debate to date: hard data. While this data will only describe one state’s experience, the lessons learned may finally enable the field to approach voter ID – indeed, all election policy – from an empirical rather than a rhetorical perspective.

Fortunately, I was right; DOJ’s December 23 objection letter avoids the familiar (I would say tired) rhetoric about the impact of voter ID and jumps right into the data. In objecting to South Carolina’s ID requirement, DOJ makes the following observations:

  • + The State provides no evidence of the fraud that ID is designed to prevent;
  • + Nearly 240,000 (almost 9%) of South Carolina’s registered voters lack ID;
  • + Minority (“non-white”) voters are 20% more likely to lack ID;
  • + county-to-county figures for lack of ID range from 6.3% to 14.2% – and encompass several of the state’s largest minority populations; and
  • + all of these figures suggest that 81,938 minority registered voters lack required ID.

While the objection has set off the predictable partisan firestorm about the wisdom and courage (or lack thereof) of the DOJ, it has also elicited some empirical resistance that could be important if and when the objection is reconsidered, either by DOJ or a federal court.

First, as I noted just before the holiday break, the data on which DOJ based its analysis may not be reliable. Specifically, the state DMV announced on December 22 – the day before the objection came down – that the 240,000 figure provided to DOJ by the State Election Commission was incorrect because of matching problems and should be closer to 30,000. This number would shrink the percentage of registered voters without ID to a little more than 1%. That number may still be too high for DOJ – especially when county or racial disparities are taken into account – but it’s a significant drop from 9%.

Second, in a portion of the letter which (unsurprisingly) didn’t get media coverage, DOJ notes that several mitigating factors advanced by the state – free IDs, public education and outreach to citizens lacking ID – weren’t considered because the procedures to accomplish these programs weren’t final at the time of the submission. DOJ explicitly notes that such measure could “potentially mitigate” the discriminatory effects of the ID requirement – as Louisiana did with its own ID requirement in 1997 after an initial 1994 objection.

Strip away all of the rhetoric, then, and this is where we stand on an empirical debate about SC’s voter ID law:

  1. DOJ says that the data the state provided suggest that voter ID would disenfranchise 81,938 minority voters;
  2. DMV is saying that its data proves that number is almost certainly too high; and
  3. The state has yet to finalize or submit mitigating measures that could drive the number down even further.

These are important facts; yet even as I decry the rhetoric, I have to admit that these decisions will not be made in a partisan vacuum. Still, I think the debate is better off for the numbers, as I noted in an introduction to Pew’s Data for Democracy report in 2008:

We also need to understand that data merely illuminates problems. It does not solve them. Disagreements over values are resolved in the political arena — but data helps put flesh on the bones of what is often a skeletal debate over values.

A new law that will enfranchise hundreds of new voters but also opens the door to dozens of fraudulent votes will be acceptable to some and an anathema to others. Conversely, measures preventing fraud but also preventing otherwise eligible voters from casting ballots will spark similar disagreement. Data helps put these discussions on an empirical level, but the decisions themselves will go far beyond the data. Still, these disagreements are better, not worse, for the availability of data.

We’ve reached that moment in the analysis of South Carolina’s voter ID requirement. DOJ has put the data in the spotlight – now let the REAL debate begin.

2 Comments on "Now the REAL Debate Can Begin: How DOJ’s SC Voter ID Objection FINALLY Brings Data to the Discussion"

  1. I received the following email from Justin Levitt at Loyola, who graciously agreed to let me reprint it here as a comment …


    I share, of course, your applause for a discussion focused on actual numbers and actual facts (that’s a standard I’ve been trying to promote for a long time on this issue (

    On that line, though, a small note: the South Carolina DMV didn’t actually say that the real number of ID-less citizens is closer to 30,000. The DMV director’s report did point out matching problems, and those problems bring the number south of the initial 240,000 report. But as I wrote here (, if all of the data at the DMV is accurate, that still leaves 90,064 registered voters without current, state-issued ID … and that’s not counting the number of individuals who aren’t currently on the rolls who don’t have the right document. I know the press picked up 30,000 as the target number, but a look at the DMV’s actual letter shows that’s not quite right either.

    The broader point, of course — enthusiasm for actual facts rather than rhetoric — is spot-on, and one with which I heartily agree…


  2. Doug,

    Great post! The Cost of Freedom Project, a citizen-led voter education initiative, is using open data to develop the Cost of Freedom App, a location-based web app to provide voters with information on how to apply for a voter ID.

    I agree that voting rights advocates need to move beyond the tired rhetoric of poll tax. So we will also gather data to quantify the high cost of a “free” voter ID. For more information, please visit us at


Leave a comment

Your email address will not be published.