This post was written with my colleague Andrea Levien.
The Justice Department announced this week that it will sue North Carolina over its new restrictive voting law. Although the law wiped out a range of pro-democracy measures, including FairVote priorities like instant runoff voting for judicial vacancies and voter pre-registration for young people, the lawsuit will focus specifically on the shortening of early voting periods and the requiring of government-issued photo identification in order to vote, policies the Justice Department claims illegally discriminate against racial minorities.
The Civitas Institute, a North Carolina group founded by the controversial Art Pope, which had great influence in designing the law and has long pushed for strict voter ID requirements, has been highly critical of the prospective lawsuit. In a blog post this week, the organization's election policy analyst Susan Myrick (not to be confused with former North Carolina Congresswoman Sue Myrick) claimed that the effects of restrictive voting laws on racial minorities are overblown. Unfortunately Myrick used a misleading statistic in her blog post and ignored many others that contradict her argument, just as Civitas president Francis De Luca did when he testified last spring before the North Carolina state legislature in favor of voter ID.
On the one hand, Myrick accurately points to the fact that the number of votes cast by African Americans in Georgia increased much more than the number of white votes cast did in the 2008 presidential election, one year after Georgia's voter ID law went into effect. This pattern was true in many southern states with large African American populations, and was likely spurred by the fact that the election featured Barack Obama, the nation's first black major party presidential nominee. Although voter ID laws can be highly problematic based on how they are implemented, they can still allow the great majority of determined people to find a way to exercise their right to vote.
However, Myrick's next statement is deceptive. She writes: "In the 2012 presidential election, African-Americans again turned out in record numbers. In all, Georgia's turnout with voter ID was 3 percentage points higher than North Carolina's without ID. Clearly, a requirement for a photo ID has not stifled voting among minorities in Georgia and will not hinder any eligible voters' opportunity to vote in North Carolina."
While Myrick is right that black voter turnout nationwide reached a record high in 2012, she is incorrect in her assertion that voter turnout was higher in Georgia than it was in North Carolina in 2012. The number she refers to as "turnout" is the turnout of registered voters, i.e. the percentage of people who are registered to vote who cast ballots. This number can vary widely based on how many people are registered to vote in the state, as North Carolina and Georgia powerfully demonstrate.
In 2012, 72.2 percent of registered voters cast ballots in Georgia and 68.4 percent did in North Carolina, a difference of 3.8 percentage points. However, no reputable authority would compare turnout between states by using registered vote turnout. The better measure of turnout is that of eligible voters -- people who are citizens, at least 18 years old, and are not disenfranchised by state laws relating to felony convictions. Using this standard methodology, a different pattern emerges. As reported by Dr. Michael McDonald, only 58.7 percent of those eligible to vote in Georgia in 2012 cast ballots compared to 65.2 percent of eligible voters in North Carolina, a difference of 6.5 percentage points in North Carolina's favor.
In fact, over time turnout in North Carolina has been pulling ahead of Georgia's in presidential elections. In 2004 -- the last presidential election without Barack Obama on the ballot and before Georgia passed its voter ID law -- North Carolina's turnout was 2.3 percentage points higher than Georgia's. In 2008, the difference between the states rose 3.4 points, and in 2012, the difference was 6.5.
Georgia's voter ID law may not be the reason for this growing disparity. After all, North Carolina was one of our nation's lucky swing states in 2012. But any fair discussion of the potential impact of voter ID on turnout should not cherrypick data. Consider that:
- Alabama and Mississippi are also non-swing southern states with large black populations, but without strict voter ID laws in 2012. Both trailed Georgia in presidential turnout in 2008, but surpassed it in 2012. (Mississippi did pass a strict voter ID law in 2012 that will go into effect for the 2014 elections, however.) South Carolina, another similar neighboring state without a strict voter ID law, outpaced Georgia in its increase in presidential turnout from 2004 to 2012.
- Kansas and Tennessee implemented strict voter ID requirements for the first time in 2012. Kansas' presidential turnout dropped 5.2 percentage points, while Tennessee's presidential turnout dropped 4.8 points. Overall, all four states with strict voter ID laws in 2012 (including Georgia and Indiana) experienced a turnout decline higher than the national turnout decline of 3.5 points.
- Analyst Nate Silver projected that strict voter ID laws would have reduced turnout by some 1 percent to 2.5 percent if they had been implemented in certain states in 2012.
Proponents like Myrick should not ignore such inconvenient facts. While Civitas' repeated comparison of registered voter turnout in Georgia and North Carolina is not as problematic as Civitas' national ally True the Vote's initial attempt earlier this year to "prove" that voter turnout rose in states with strict voter ID laws by comparing eligible voter turnout in 2008 with registered voter turnout in 2012, it is still deceptive.
As a final point, the contrast between voter turnout among eligible voters and registered voters in North Carolina and Georgia underscores one of the deepest flaws in our election administrative process: our voter registration system. North Carolina likely has many ineligible voters on its voter registration rolls that have either moved or passed away, while Georgia has clearly registered far too few of its eligible voters. Our nation's failure to embrace voter registration modernization has left us with a system that undercuts the goals of representative democracy. There are millions of people on state and county rolls who should not be registered there, and there are tens of millions of eligible voters who have not been registered at all. Even if we continue to argue about voter ID requirements and fairer ways to implement them, it is time that we come together in support of common sense voter registration reform.