Racial discrimination vs. fraud at the polls
Harvard Law School Professor D. James Greiner is co-author of a recent study on the experience of Boston voters in the election of 2008. As another election approaches, we ask Greiner a few questions about his study and the current efforts to pass tougher voter ID laws.
Your article is titled “Can Voter ID Laws Be Administered in a Race-Neutral Manner?” What did you find?
Our study reached some discouraging conclusions. The combination of federal and Massachusetts laws limits the circumstances under which a person approaching a voting location in Boston should be asked for an ID. Despite these limits, far too many white persons who approached voting locations in the 2008 general election in Boston were asked for an ID. But the problem was even worse for African-American and Hispanic voters. Members of racial minority groups were more likely, more than 10 percentage points more likely in most cases, to be asked for IDs. The disparity among racial/ethnic groups persisted even when we used sophisticated statistical techniques to control for other things a poll worker was likely to observe when a would-be voter approached a voting area—things like gender, age, educated speech patterns and proficiency with English.
What role can law play in the solution to the discrepancy you found?
That’s the particularly discouraging thing. The “law” (meaning the law written down in the books) seems to have been pretty clear here. [It requires that an ID be requested of voters who registered by mail without including a copy of a valid ID and of those who are classified as inactive voters.] In fact, election administration officials had gone so far as to print an “I” or an “ID” next to the name of each person who was to be asked for an ID if she approached the polls. In other words, if a person trying to vote had an “I” or an “ID” next to her name, the poll worker was supposed to ask for an ID. If neither of these symbols was present, the poll worker was not to ask for an ID. It’s hard to see how the issue could have been made more simple or straightforward. And yet, racial and ethnic disparities occurred.
Other studies have shown that a majority of American voters, including blacks and Hispanics, favor voter ID laws. How does that jibe with your results?
ID laws have a strong intuitive appeal. After all, the argument goes, one has to show not just an ID, but a photo ID (Massachusetts law has never required a photo ID to vote), to board an airplane. Why not require one for voting? It turns out that there are two reasons not to require an ID for voting.
First, there is no evidence (really, none) that the kind of fraud that voter ID laws prevent has ever occurred in the United States on a significant scale. If you think about it, voter ID laws prevent only one kind of fraud: a person showing up at a voting location to impersonate someone else. To pull this off, the impersonator must know the actual voter is still on the voting list, must also know the actual voter’s name and address (both of which must be provided in order to vote), and must know that the actual voter has not already voted that day. Realistically, each fraudster could impersonate only one or two voters on a voting day, so there would need to be a coordinated effort involving a lot of advance knowledge and a lot of fraudsters to affect an election. There are so many easier ways to commit fraud (stuffing ballot boxes, misleading potential voters about polling dates or locations, prematurely purging registration rolls, etc.) that such a coordinated effort is not worth the candle. So basically no one does it. The Bush Justice Department tried (hard) to find examples of this kind of fraud so as to justify voter ID laws, but it found essentially nothing.
Second, there is some evidence (including our study) that voter ID laws disproportionately affect racial and ethnic minorities, the elderly, and the poor. We know that persons with these demographics tend to lean Democrat. That’s why Republicans like these laws. Don’t get me wrong: If these laws disproportionately affected Republicans, Democrats would be all for them. But we as citizens shouldn’t swallow this kind of garbage.
What is the national import of this study, especially in a year when there have been efforts in states around the country to pass more stringent voter ID laws than the ones that exist in Massachusetts?
It’s true that we studied only one jurisdiction, the city of Boston, in one election, the general election in 2008. Nevertheless, this is one of the last places and times one would have expected to observe racial and ethnic discrimination in the administration of voter ID laws. In 2008, the top two elections, for president and U.S. Senate, were expected to be blowouts in Boston and in Massachusetts, and in fact they were blowouts. Only one other issue decided that day, a proposal for a statewide ban on dog racing, was remotely close. Fraud is always a concern, but there was little reason to expect Boston election officials to be hypervigilant about fraud that day. Meanwhile, Boston is racially and ethnically diverse, and the Boston Election Department in 2008 had recently made some effort to hire racially/ethnically diverse poll workers. True, Boston has some ugly periods on race relations in its past (but what U.S. jurisdiction has not?), but race relations in Boston are probably better than they are in many other places in the United States. Finally, as noted above, it’s hard to see how the law could have placed more limits on the poll workers who administer these laws. Of course, more investigation is needed before one can draw firm conclusions, but our thought is, If racial disparities occurred in Boston in 2008, it seems pretty likely that they are occurring in other jurisdictions at other times. With all that in mind, one has to ask, Are these laws worth it?
Read “Can Voter ID Laws Be Administered in a Race-Neutral Manner? Evidence from the City of Boston in 2008” (Quarterly Journal of Political Science, 2012) by Rachael V. Cobb, D. James Greiner and Kevin M. Quinn.