We Have the Facts and We're Voting Yes

Last December, Heather Gerken made the case for the creation of a system that ranks states and localities based on how well they run elections. Her new book, The Democracy Index, published by Princeton University Press, expands upon the idea, arguing that clear data about the electoral system could stimulate productive competition between localities and thereby improve the voting process.

In the following excerpt adapted from The Democracy Index, Gerken considers the practical effects that voting data -- and lack thereof -- has on ordinary citizens and the implementation of policy.

Spencer Overton doesn't fit the stereotype of an election reformer. Polished and professional, it's easier to imagine the Justice appointee in Armani than Birkenstocks. Overton draws his idealism from a civil-rights background, and he is capable of talking about the right to vote in stirring terms. But with his Harvard Law degree and measured baritone, it's as easy to imagine him relating to corporate executives as to public interest lawyers.

People have written a good deal about the new generation of reformers. Entrepreneurial and pragmatic, they eschew old political divides and attack problems with the hard head of a corporate executive. They look to a variety of institutions (the market, administrative agencies), not just the courts, for solutions. They are as likely to appeal to business-minded ideas -- accountability, competition -- as progressive values like participation and empowerment. Overton, who once taught law at the George Washington University, perfectly embodies this new style.

Overton's problem is that he is fighting for change in a world without data. Indeed, he found himself in the middle of one of the biggest election reform battles we've seen in recent years -- one that made it all the way to the Supreme Court -- and lost in large part because he didn't have the data he needed to make his case.

The fight was over voter identification -- the requirement that voters show a government-issued photo ID when they cast a ballot at the polls. Voter ID has been a significant source of contention in election circles. Conservative commentators insist that an ID requirement deters fraud. Liberal commentators counter that the requirement is a disguised effort to suppress (largely Democratic) votes. The rhetoric on both sides of the issue has been quite heated, with one side talking about stolen elections and the other side equating ID requirements with vote suppression.

Overton became embroiled in the issue when it was taken up by the Commission on Federal Election Reform, chaired by former Democratic president Jimmy Carter and former Republican secretary of state James Baker. Though most of the members of the bipartisan commission had strong political ties, it included a handful of academics, including Overton. The Carter-Baker Commission eventually staked out a position on voter ID that looked an awful lot like a political deal. It roughly tracked the compromise that would emerge if a prominent Democrat and a prominent Republican sat down to work out something both sides could live with. The commission blessed the ID requirement (something Republicans usually want) while demanding that the state take affirmative steps to distribute IDs (something that Democrats would want if forced to accept an ID requirement).

Deal or no deal, the main problem with the commission's position was that it was utterly unsupported by empirical evidence. A pure political compromise can be produced without coming to grips with the empirics; a sound decision cannot. Although the commission did an excellent job of amassing data on how our election system is run in many areas, this was not one where it managed to find much. As the commission itself stated, there is "no extensive evidence of fraud in the United States." To the extent there is any evidence of fraud, it is almost entirely due to absentee voting scams or ballot-box stuffing, not the type of fraudulent in-person voting that photo ID is supposed to deter. The only other justification that the commission offered for its decision was that a photo ID requirement would enhance public trust in the system. That claim, too, was unsupported by empirical evidence (and may have been misplaced).

Overton did his best to persuade the other members of the commission not to endorse an ID requirement. Most advocates contesting voter ID have simply invoked civil-rights rhetoric. Overton called upon that tradition, but he mainly focused on the kind of cold-blooded cost-benefit arguments that conservatives stereotypically use. Working with the Brennan Center, he tried to amass data on the effects, good and bad, of photo ID. When he failed to change the majority's mind, he published a forcefully worded dissent. I saw Overton a day after the fight went public. I've never seen anyone more exhausted.

The reason Overton faced such an uphill slog is that the data were haphazard and inconsistent. As he discovered, "No systematic, empirical study of the magnitude of voter fraud has been conducted at either the national level or in any state to date." Nor were there any good studies on an ID requirement's effect on voter behavior. Overton pulled together some basic numbers (how many voters lack ID, how many fraudulent votes might have been prevented by an ID requirement). Based on these numbers, he argued that it would be a mistake to endorse voter ID at this stage because the commission could not show that it "would exclude even one fraudulent vote for every 1000 eligible voters excluded." But Overton candidly admitted that his data, standing alone, could not tell you what would happen if an ID requirement were enacted.

Overton and the Carter-Baker Commission as a whole had the same problem: they were fighting about reform in a world without data. The Carter-Baker Commission justified its conclusions with the only evidence available: anecdote. Overton believed that anecdotal evidence led the commission to overestimate both the problem of fraud and the likelihood that an ID requirement would solve it. Overton did not spare his allies criticism, either. He rebuked opponents of voter ID because they "regularly recite talking points about threats to voter participation by the poor and minorities, but often fail to quantify this assertion." Overton's frustration about the debate remains palpable: "I'm an academic," he says. "I believe in facts."

The same year that the Carter-Baker Commission released its report, the Republican-dominated Indiana legislature passed a photo ID requirement in a straight party-line vote. The state conceded it was not aware of a single episode of in-person voter fraud in its entire history, and the legislature failed to do anything about the security of absentee ballots (the one area where Indiana had recently had a fraud problem). "Let's not beat around the bush," wrote one of the lower-court judges reviewing the case. "The Indiana voter photo ID law is a not-too-thinly-veiled attempt to discourage election-day turnout by certain folks believed to skew Democratic."

When the lawsuit challenging Indiana's law worked its way to the Supreme Court, Justice Stevens, writing for the plurality, upheld the requirement. He concluded that photo ID was a reasonable strategy for combating fraud and building voter confidence. What evidence did Justice Stevens cite in support? A funny anecdote dating back to Tammany Hall, the fact that one person had voted fraudulently in a Washington gubernatorial election . . . and the Carter-Baker Report.

The problem is obvious. The Supreme Court didn't have much evidence to cite for its view that in-person vote fraud was a problem. So it cited the Carter-Baker Report, which in turn didn't have much evidence to cite. The Supreme Court had no evidence to cite for its intuition that an ID requirement boosts voter confidence. So it cited the Carter-Baker Commission, which in turn had no evidence to cite. It's turtles all the way down.

The debate over voter ID is part of a larger story about reform in a world without data. The story has an obvious moral -- whether your intuitions are closer to Justice Stevens' or Spencer Overton's, surely you'd prefer the decision rested on data. But it also gives you a flavor for what reform debates look like in a world without data.

Note, for instance, what kind of reform proposals get traction in a world without data. Most reforms never see the light of day, as I discussed in the last chapter. The rare proposals that do get traction are those with intuitive appeal, like an ID requirement. Middle-class voters are accustomed to showing ID to get on a plane or pay with a credit card, so it's easy to frame the issue in a way that they can understand. (Think about the only other issue to get traction in recent years -- paper trails for voting machines. It's another issue people can wrap their hands around.) There's no reason, of course, to think that intuitively appealing reform is the right reform. But the best strategy for defeat mistaken intuitions -- testing them empirically -- is impossible in a world without data.

Worse, in the absence of data, reform debates are completely at the mercy of politics. The reason photo ID got passed in Indiana is because it aligned with partisan incentives and the other side couldn't build a case against it. (Lest you think I'm picking on the Republicans, I should emphasize that Democrats are similarly inclined to oppose photo ID because of their own political interests. Remember, the Indiana law was passed without a single Democratic defector.) Similarly, when the Carter-Baker Commission announced its position on voter ID, it had no empirical basis to think it was announcing good policy. All that the Carter-Baker Commission could offer was a position that both political parties could live with. Here again, there is no reason to think that "change the parties can live with" bears any resemblance to the change we need.

Just think about how hard it is to referee this fight. There are lots of accusations and few facts. The Republicans and Democrats shout about partisanship. Reformers hint darkly about voter suppression. Whether you are a voter or a Supreme Court justice, it's hard to figure out who is right unless you subscribe to Lev Tolstoy's wry claim that "among coachmen, as among us all, whoever starts shouting at others with the greatest self-assurance, and shouts first, is right."

Finally, and most importantly, note that the ultimate referees of this fight -- members of the Supreme Court -- were hungry for guidance. The Court encountered the dilemma we all face in the elections context: distinguishing between legitimate efforts to regulate the election system and illicit attempts to hijack it for political ends. The justices were plainly on the hunt for a yardstick to evaluate the Indiana law. Justice Stevens wasn't the only one to rely on the Carter-Baker Report. The dissenting justices did so as well. Unfortunately, it wasn't a very good yardstick for the justices to use. The Carter-Baker Commission had nothing to go on except atmospherics and anecdote. All it could offer is a compromise that smelled like a political deal. The voter ID fight makes clear just how powerful a yardstick can be in election debates. Even an imperfect baseline -- a bipartisan compromise unsupported by empirical evidence -- was enough to sway the Supreme Court. Imagine what a better metric could achieve.

The story of the photo ID looks a lot like the story of election reform generally. Reform battles take place in a world without data. We know more about the companies in which we invest, the performance of our local baseball team, even our dishwashers, than we do about how our election system is working. The institutions that administer our election system -- the linchpin of any democracy -- don't give us the information we need to evaluate how they are performing. The limited data that exist are often undependable, unverifiable, and too inconsistent to allow for comparisons across jurisdictions. It is remarkable that we spend so much time arguing about which direction election reform should take when we don't even have the data we need to map where we are now.

This essay is adapted from the new book The Democracy Index by Heather Gerken, published by Princeton University Press. Copyright © 2009 by Princeton University Press.