I just can’t get you (the Bradley Effect) out of my head

cross-posted at my blog and PostBourgie

Andrew Hacker’s essay in the latest issue of the New York Review of Books is something of a mixed bag. The piece is an attempt to measure the possible impact of race – specifically voter registration laws, and the “Bradley Effect” – on the election.  And to some extent, Hacker is successful; he does an excellent job at giving an overview of the current state of voter rights in the country. Specifically, he looks at the Supreme Court’s ruling in Crawford v. Marion County Election Board, and the 2002 Help America Vote Act, and shows how the impact of both – whether intentionally or not – has been to suppress voter registration and turn out among African-Americans and other minorities. (This is kind of long, so the rest is after the jump.)

In Crawford v. Marion County Election Board, the Court “upheld a 2005 Indiana law requiring voters in that state to produce a government document with a photograph at the polls.”  Which means that in the state of Indiana, voters need to provide either a driver’s license, passport, or government-issued ID in order to vote. Moreover, by upholding the ruling, the Court opened the door for other states to pass similar regulations. Hacker notes that this could potentially disenfranchise thousands of poor and working-class minorities, particularly African-Americans:

Requiring a driver’s license to vote has a disparate racial impact, a finding that once commanded judicial notice. To apply for the state ID card that Indiana offers as an alternative, moreover, nondrivers must travel to a motor vehicles office, which for many would be a lengthy trip. While licenses do not record race, Justice David Souter cited relevant studies of the race of license-holders in his dissent, which was joined by Justices Stephen Breyer and Ruth Bader Ginsburg. In one survey, made by the Department of Justice in 1994, black residents of Louisiana were found to be four to five times more likely not to have the official photograph needed for an identifying document. (Not to mention access to a car; recall how many couldn’t leave as Katrina approached.) A Wisconsin survey published in 2005 was more precise. No fewer than 53 percent of black adults in Milwaukee County were not licensed to drive, compared with 15 percent of white adults in the remainder of the state. According to its author, similar disparities will be found across the nation.

The problems, however, don’t stop there.  The 2002 Help America Vote Act mandated that every state must maintainan  ”electronic “statewide voter registration list,” to be linked to every precinct.”  As Hacker points out though, African-American families are more likely to move, and like most people, probably won’t remember to notify election officials.  This, he writes, has already had a detrimental effect on voter registration:

When Ohio purged 35,427 returned names in 2004, a review found that the addresses were in “mostly urban and minority areas.” Here too, getting back on the rolls can be like mending a mistaken credit rating.

Besides the Supreme Court’s ruling and the Help America Vote Act, there’s the recurring concern of felon voter rights:

Among inmates, black men and women outnumber Hispanics by more than two to one and whites by nearly six to one. This is another reason why a much higher ratio of black citizens will be unable to vote this year, because they are among either the 882,300 who are currently incarcerated or the two million who have served sentences but continue to be disenfranchised. According to Restoring the Right to Vote, a report by the Brennan Center for Justice at New York University Law School, 13 percent of black men cannot cast votes; in three states 20 percent cannot because they are locked up or formerly were.

Hacker correctly deducts that all of these things will impact the number of African-American voters Obama receives, and could hurt him in states where African-Americans could make the margin victory (like Virginia).  To use a fairly recent – and for many, still painful example – analysis’ of the 2000 election found that in Florida, more than 14,000 voters were purged from the rolls, many of whom were African-Americans wrongly identified as felons.  The margin of Bush’s victory in 2000 was a little over five hundred.

As good as the quantative portion of Hacker’s piece is, he begins to stumble when he tries to further draw out the political implications of race in the election.  He begins by noting that – like the gay marriage initatives in 2004 – the various anti-affirmative action initiatives on the ballot this year could trigger racial resentment in white voters, leading them to vote against Obama.  He ties this to the oft-mentioned Bradley (or Wilder) Effect, which for those unaware, is:

an explanation advanced as the possible cause of a phenomenon which has led to frequent inaccurate voter opinion polls in many American political campaigns between a white candidate and a non-white candidate. Specifically, there were instances in which such elections saw the non-white candidate significantly underperform with respect to the results predicted by pre-election polls.

There are two problems with this argument: first, there’s no real consensus over the impact on the Bradley Effect, and there’s even evidence to suggest that it’s impact isn’t as great as we once thought.  At FiveThirtyEight, Nate Silver – using data from the primary election – suggests that the Bradley Effect didn’t have much, if any impact on the Democratic primary:

On average, Barack Obama overperformed the Pollster.com trendline by 3.3 points on election day.

There are some important differences by region. Using regions as defined by the US Census Bureau, Barack Obama overperformed his polls by an average of 7.2 points in the South. This effect appears to be most substantial in states with larger black populations; I have suggested before that it might stem from a sort of reverse Bradley Effect in which black voters were reluctant to disclose to a (presumed) white interviewer that they were about to vote for a black candidate.

Obama also outperformed his polls in the Midwest and the West (although there is not much data to go on in the latter case). The one region where Hillary Clinton overperformed her numbers was in the Northeast, bettering the pre-election trendline by 1.8 points. Recall that the Bradley Effect phenomenon describescovert rather than overt manifestations of racism. It may be that in the Northeast, which is arguably the most “politically correct” region of the country, expressions of racism are the least socially acceptable, and that therefore some people may misstate their intentions to pollsters. By contrast, in the South and the Midwest, if people are racist they will usually be pretty open about it, and in the West, which is nation’s most multicultural region, there may be relatively little racism, either expressed or implicit.

And second, we can’t necessarily infer decreased support for Barack Obama from opposition to affirmative action, at least not in the same way that we could – in 2004 – infer support for George W. Bush based on opposition to gay marriage.  For one, affirmative action occupies a much different – and recently less contentious – position in the discourse than gay marriage.  The general consensus among Americans concerning affirmative action is that it either needs to be A) changed or B) eliminated.  Moreover, this transcends party lines; you’re just as likely to find anti-affirmative action Democrats as you are Republicans.  Furthermore, you’re far more likely to find a single-issue voter driven by gay marriage than you are a single-issue voter driven by affirmative action.  Whereas in 2004, it was highly likely that a voter who touched the screen (“pull the lever”is such an anachronism) against gay mariage would also touch it for President Bush, it’s no where near as likely that a voter who touches the screen against affirmative action will also touch it against Barack Obama.  Indeed, if you’ll allow me to make some unfounded conjecture (which of course you will, this is a blog), I’d say that the very fact of Barack Obama’s candidacy could even serve to reinforceopinions against affirmative action.  I can imagine a hypothetical white voter saying, “If a black man like Barack Obama can make it to the presidency, then we don’t need preferences for minorities.”

I know that some Democrats are worried about the impact of affirmative-action measures on the election. My advice is not to panic.  It’s not at all clear that they will actually have any impact on Barack Obama’s electoral performance, and if they do, the extent of said impact is still very much up in the air.  My guess is that if there is any impact, it will be slight, and increases in voter registration among all groups, as well as turn out, will more than make up the difference.

As far as Hacker’s essay is concern though, I’d say that on the whole, he does a pretty good job on both shedding light on the various electoral controversies popping across the country, and addressing the recurring issues in voter enfranchisement (like felon’s rights, and such).  Where Hacker stumbles is in tackling the potential electoral impact of race; with just a little bit more data, and some better reasoning, this essay could have made the leap from merely good, to pretty great.