For years, states have been sounding the alarm about voter fraud and pushing laws to prevent it. One such law would require voters to prove their citizenship, with a birth certificate, passport, or the like, before casting a ballot. This month a federal court slapped down proof-of-citizenship laws, but not for good. The opinion leaves wiggle room, and state lawmakers are not giving up. They are, however, wasting their effort. Anti-fraud measures can make elections safer in some circumstances, but usually they either have no effect (other than creating red tape) or make matters worse. My research proves it.
Let's get up to speed. In 2004, Arizonans approved an initiative requiring voters to prove their citizenship before they could vote. In 2013, the Supreme Court held [PDF] that federal law—specifically, the National Voter Registration Act—preempted the initiative. That Act requires states to use a form, developed by the federal Election Assistance Commission, to register voters for federal elections. The form requires would-be voters to swear, under penalty of law, that they are US citizens. States can ask the EAC to add state-specific instructions to the federal form, including additional requirements on citizenship, but they cannot demand it.
Other states—Alabama, Georgia, Kansas—adopted laws like Arizona's, and they worked many channels to get them enforced. But their efforts failed. States courts, federal courts [PDF], and the EAC [PDF] put proof of citizenship on ice.
Until January. That's when Brian Newby, the EAC's executive director, granted permission to Alabama, Georgia and Kansas to require proof of citizenship while using the federal form. Newby's decision, which came with little explanation, generated heat. Thomas Hicks, an EAC commissioner, publicly rebuked [PDF] Newby, saying he exceeded his authority and contradicted EAC policy. When civil rights groups sued, the Department of Justice refused to support Newby, siding with the plaintiffs instead and urging a federal judge to block the decision. This left Kris Kobach, the Kansas secretary of state, to defend the EAC in court.
On September 9, the DC Circuit Court of Appeals ruled against [PDF] Kobach and company, ordering the EAC to remove proof-of-citizenship requirements from the federal form. On September 26, the court issued an opinion [PDF] explaining its order. With respect to Newby's decision, the court wrote, "it is difficult to imagine a more clear violation of the APA's requirement that an agency 'must examine the relevant data and articulate a satisfactory explanation for its action[.]'" The court also rejected the argument that eliminating proof of citizenship would leave states vulnerable to voter fraud.
Fear of voter fraud anchors this entire controversy. According to Kobach, "there is a significant problem in Kansas and in the rest of the country of aliens getting on our voting rolls." Fraudulent voting violates law, cancels out lawful votes, and could, in just the right case, swing an election. Proof-of-citizenship requirements should mitigate that risk, or so goes the argument.
But the risk of fraud is infinitesimally small. As the DC Circuit Court wrote, "there is precious little record evidence" of voter fraud. After much digging, Kobach identified 18 non-citizens who tried or successfully registered to vote in Kansas between 2003 and 2015, only one of whom used the federal form. These numbers are consistent with other studies, and they are trivial. Kobach has secured only a handful of convictions, and none of his targets showed intent to game the system.
If proof-of-citizenship requirements came with no downside, these small numbers wouldn't matter. All fraudulent voting is bad, and less is better than more. But there is a downside: Many citizens can't prove their citizenship. Lost birth certificates, tattered passports, missing driver's licenses, forms, fees and trips to government offices—these obstacles would keep lawful voters from the ballot box if proof-of-citizenship laws took effect. Kobach claims to have found 18 fraudulent voters in 12 years. How many people lost their birth certificates during that stretch?
The tug-of-war between stopping fraud and suppressing turnout has dominated headlines and litigation, but it misses another, deeper problem. Anti-fraud efforts might not work. They might make the problem, such as it is, worse.
Consider this example. Candidates Amy and Bob compete in an election, and each unwittingly receives 1,000 fraudulent votes. That's a lot of fraud, but it all cancels out, and lawful votes determine the outcome. In the next election, Kobach rides to the rescue, imposing a proof-of-citizenship requirement. It prevents non-citizens from registering and voting, but it does not stop citizens from voting multiple times or other forms of voter fraud. Thanks to Kobach, the number of fraudulent votes shrinks, but not to zero. Now Amy gets 500 fraudulent votes and Bob gets 50. That's a lot less fraud, but it doesn't cancel out, and it might determine the election. Bob should win, but thanks to her big lead in fraudulent votes, Amy wins instead. The anti-fraud measure caused the problem it's supposed to prevent.
Behind this example lies a simple but important idea. The number of fraudulent votes cast for each candidate doesn't matter. What matters is the gap between them. In the example, the gap went from zero to 450, even as the number of fraudsters fell. The election became less safe.
To see this from another angle, suppose Amy and Bob used to get 500 and 50 fraudulent votes, respectively, when they competed in elections. In recent years, fraud surged, and now they get 1,000 illegal votes apiece. No one welcomes fraud, but that doesn't change the logic: The election has gotten safer.
Let's be clear about what this means. Even if proof-of-citizenship laws do not suppress a single lawful vote, they can backfire, causing exactly the problem they are designed to avoid. To prevent this, lawmakers need information not about the total amount of fraud but about the fraud gap.
One might challenge my example. There is no reason to suppose that proof-of-citizenship laws will systematically cut one candidate's fraudulent vote totals by more than the other's. Rather than going from 1,000 fraudulent votes apiece to 500 and 50, a proof-of-citizenship law might change the fraud totals to, say, 400 apiece. That could happen—and if it does the law is pointless. The election is just as safe as before, only now poll workers and voters have to deal with a cumbersome rule.
One can keep tinkering with these examples. One can complicate them by adding the realistic assumption that proof-of-citizenship laws not only affect fraudulent votes but also suppress some lawful votes. One can extend the analysis by thinking about proof-of-citizenship's cousin, voter ID laws. In a recent paper, I did this work, and I found that in some circumstances, anti-fraud measures make elections safer. But in a much larger set of circumstances, they either have no effect (but still impose regulatory costs) or they make matters worse.
Which circumstance prevails in Kansas? Does Kobach know whether a proof-of-citizenship law would help rather than hurt? Of course the answer is no. Lawmakers champion these laws without having thought through their basic logic.
The DC Circuit's decision stopped proof of citizenship, but only temporarily. The case involved a preliminary injunction, not a decision on the merits, and litigation marches forward. But perhaps courts will catch on. Perhaps they will see that, like many policy issues, making elections safer is tricky, and ham-handed efforts cannot be justified. Perhaps they will see that even if proof of citizenship does not suppress votes—a heroic assumption— proponents of these laws are still fumbling in the dark. They have neither the theory nor the information necessary to sustain their position.
Michael Gilbert is Sullivan & Cromwell Professor of Law at the University of Virginia School of Law.
Suggested citation:Michael Gilbert, Why Proof of Citizenship Won't Improve Election Integrity, JURIST - Forum, Oct 3, 2016, http://jurist.org/forum/2016/10/michael-gilbert-election-integrity.php
This article was prepared for publication by Yuxin Jiang, a Senior Editor for JURIST Commentary service. Please direct any questions or comments to her at commentary