Recently, in an effort to woo the Obama coalition, Hillary Clinton announced her vehement opposition to all things related to Voter ID. Speaking at a black college in Houston she called for universal and automatic voter registration and a 20-day (or more) period of early voting in every state, before every election.
However, Republicans were quick to pounce on the former NY Senator’s comments. John Kasich (R-OH) pointed out that Ohio has 28 days of early voting and NY state (her state) has none. Former TX Governor Rick Perry forcefully defended Voter ID laws his state put in place under his tenure.
There is little wonder why Clinton is hitting the GOP on voting rights. Numerous battleground states (WI, FL, GA, TX, PA) have put in place Voter ID laws in the last decade. Democrats and liberals believe this predominately impacts Democratic favoring minority communities and the young. But do Voter ID laws limit turnout? The evidence is mixed.
Following the GOP wave of 2010 Alabama, Kansas, Mississippi, New Hampshire, Pennsylvania, Rhode Island, South Carolina, Tennessee, Texas, Virginia, and Wisconsin all passed new voter ID laws between 2011-2013. Not all are solidly GOP states as Rhode Island is about as Democratic as it gets. In 2014 North Carolina joined the list with new requirements that limited early voting and shortened the time frame for absentee ballots to be sent and received (more on North Carolina in a minute).
In 2013 (right after Obama was reelected) the GAO released a report stating that “Voter ID laws reduce turnout.” The report’s findings “compared election turnout in Kansas and Tennessee — which tightened voter ID requirements between the 2008 and 2012 elections — to voting in four states that didn’t change their identification requirements.” The report concluded “reductions in voter turnout were about 2 percent greater in Kansas and from 2 percent to 3 percent steeper in Tennessee than they were in the other states examined. For the record the states that did not toughen their ID laws when the report came out were Alabama (it did soon after the report came out), Arkansas, Delaware and Maine.
Aha. There is the smoking gun that proves Voter ID laws reduce turnout. Or do they? For there are some very curious outliers. Take the cases of North, Carolina and Georgia. Each has implemented strict Voter ID laws in the last decade.
Georgia has had a photo ID law in place since 1977 but in 2005 the law was strengthened to require Photo IDs to be used at the polls by 2012. Like many states the law funded giving citizens free IDs to use at the polls. The Atlanta Journal Constitution ran a turnout post-mortem of the 2012 election and found that 77% of all registered black women and 66% of registered black men turned out to vote in the presidential election compared to 80% and 70% in 2008. Comparing the 2006 midterms to 2010 the report found a far greater share of black voters turned out in 2010 than in 2006. In 2014, blacks registered their highest turnout in exit polls.
Virginia is another curious case. In 2011 the legislature passed a strict Photo ID law but then Governor McDonnell vetoed the measure. However, in 2013, the legislature reintroduced the same measure and the outgoing Governor signed it. In the gubernatorial election that year turnout increased by 3% relative to 2009’s race. Turnout in the 2014 midterms was only a couple of points lower than 2010 (though in 2010 the state had more competitive Congressional races).
Lastly, we come to North Carolina. In 2013 the state legislature passed a strict Voter ID law and the new Governor signed it. Thus we only have one datapoint (2014) to compare data with. Nate Cohn at the NYT’s has a good analysis here. To sum up the highlights the white share of the vote dropped a 2.5% between 2010 and 2014 while the minority share of the electorate increased 2%. Among blacks turnout increased by 1.4%. Turnout among the young also did not suffer. In 2010 18-25 year olds were about 4% of the electorate. In 2014 they were 5%.
These states turnout results run counter to the narrative that Voter ID laws hurt turnout. If they did we should expect to see these increasingly diverse states have drops in turnout among minorities and the young. Yet, it is not there.
This simple analysis does not refute every argument opponents of Voter ID make. For example, there is a strong case to be made Voter ID laws hurt turnout in predominately rural and heavily immigrant communities. Few large-scale studies have been conducted to see if this is true or not though.
Considering the numerous factors that go into campaigns (candidate quality, money, partisan orientation) it is incredibly hard to argue one way or the other whether Voter ID laws adversely impact turnout. For example, turnout in Virginia’s 2010 midterm could have been higher because the state featured three competitive Congressional district races. In 2014, only one district was competitive perhaps leading to a drop in turnout.
This won’t stop politicians like Clinton from playing on people’s fears. Fortunately, Republicans are firing back and arguing there is minimal evidence Voter ID laws do impact turnout. Clinton’s play is a ploy to appeal to minorities who feel they are disenfranchised. Unfortunately for Clinton (if she cares), there is simply no definitive proof that Voter ID laws reduce turnout among the (supposedly) impacted groups.