The Bradley Effect and other Racial Myths

As early as mid-May of 2007, The New York Times fretted in print about the supposed reluctance of Americans to cast their presidential ballots for a person of color. The Times dutifully recited the hoary historic example of Tom Bradley, an African-American candidate who lost his California gubernatorial bid in 1982 after being ahead in the polls in late October. Sayeth the Times: “In high-profile contests where one of the major party candidates is black, pre-election telephone polls have often been wrong, overstating the strength of the black candidate. In polling circles this is known as the ‘Bradley effect’ or the ‘Wilder effect’ or the ‘Dinkins effect.’ Will it also be known as the Obama effect?”

And so it has gone for the better part of a year and a half. With three weeks to go in the 2008 election, CNN wondered breathlessly, “Will Obama suffer from the ‘Bradley Effect?’” Hours later, ABC News mimicked: “Will the ‘Bradley Effect’ be Obama’s downfall?” ABC quoted a clearly worried University of Chicago political scientist named Michael Dawson as saying, “We’ve seen these types of leads disappear at election time. When you are in the voting booth, nobody is there and you can express what you really believe."

The casual manner in which liberals, academics, journalists, and some Democrats accuse their fellow citizens of mass racism has always stuck in Loose Cannon’s craw. Just this week, U.S. News & World Report posed the question of whether John McCain’s emphasis on campaigning in Pennsylvania in the campaign's last days is evidence that the Republicans are “counting on the Bradley effect.”

Yes, it’s getting a little paranoid out there, so it’s time to set a few things straight.

First of all, there never was any “Bradley effect”—at least, not in California. This is a myth, easily debunked, which I’ll undertake to do now. (Not surprisingly, the most sensible commentary of that 1982 California election comes from people, like me, who covered it personally, or who participated in it.)

In 1982, California Gov. Jerry Brown was stepping down after two terms to run for a vacant U.S. Senate seat. The Democrats nominated Tom Bradley, a respected fixture in the Los Angeles political scene. Bradley had been a track star at UCLA, and after college he joined the police department, working his way up to lieutenant in the L.A.P.D. while getting his law degree at night. Bradley became the first black city councilman in Los Angeles, and the city’s first black mayor—a job he’d held for nine years when he decided to run statewide against Republican George Deukmejian. "The Duke" was a New York native, known for his tough law-and-order credentials both in the state legislature, where he authored California’s “use a gun, go to prison” law, and as state Attorney General. That year, crime (but not race) was a defining issue in California. Although he wasn’t on the ballot, Jerry Brown’s legacy of liberal judicial appointments was a backdrop to the 1982 gubernatorial campaign. (Asked why he was running when he already had a job that seemed to suit him well, Deukmejian replied: “Attorneys general don’t appoint judges—governors do.”)

The Deukmejian-Bradley contest figured to be close, and it was. In the end, it was decided by some 93,000 votes out of about 7.7 million votes cast. Several events unfolded in that campaign that helped determine the outcome, including the following:

-- Bradley was indeed ahead in the polls in late October, but Election Day never comes in late October. In the waning weeks, Deukmejian was closing fast. The best tracking polls were done by Lance Tarrance’s firm, which was retained by Deukmejian’s campaign. Tarrance’s polls had it 49-37 (with Bradley ahead) on October 7; 46-41 on October 21; and 45-42 on October 28. Mervyn Field, an independent pollster whose numbers sometimes tilted Democratic, found a seven-point Bradley lead the weekend before the election, but Tarrance had it 45-44 on November 1. (Lance Tarrance himself wrote a smart piece for Real Clear Politics about the 1982 campaign, which you can read by clicking here.) Field’s exit polls also had Bradley winning, but on Election Day, Field was at odds with CBS and NBC, which polled different precincts. The two networks had Deukmejian the winner in a real close race. Why the disagreement on where to poll? This leads to the second variable that year, which was…

-- Proposition 15, an initiative on the state ballot that would have erected strict barriers to handgun purchases and require statement handgun registration. Bradley favored it; Deukmejian opposed it. Here’s a recounting of that issue. Proposition 15 not only failed, but it attracted to the polls a more conservative electorate than was typical for California. (The National Rifle Association and the state’s gun shops registered some 300,000 voters themselves, more than three times Deukmejian’s winning margin.) Thus, Merv Field’s exit polling samples may have been inaccurate.

-- Finally, in 1982, absentee ballot regulations were relaxed—and the California Republican Party seized on the change, mailing out hundreds of thousands of ballots to GOP voters. This election was the precursor of today’s early voting, which the Democrats are using even as I write this, to garner huge numbers of pre-Election Day votes for Obama. In California that year, the exit polls weren’t counting these voters because they weren’t at the polling places at all. Dan Walters, longtime political columnist for the Sacramento Bee, wrote about this in August. His entire piece is here.

In the end, did white voters lie to pollsters? Did Bradley even lose because of his race? There’s no real evidence for either claim. Republicans were winning statewide in California in the 1980s (Ronald Reagan carried the state, hugely, twice), and on the same day Bradley was losing, Jerry Brown lost his U.S. Senate bid to Pete Wilson by nearly six times Bradley’s margin of defeat in a contest in which both candidates were white. As for the notion that people deceive pollsters because they’re embarrassed by their own political preference, well, that seems little more than the product of a liberal theorist who has trouble conceiving of any reason for voting against Tom Bradley other than race.

This theorist, in fact, has a name. He’s Charles P. Henry, and he’s a professor of African-American studies at the University of California, Berkeley. Twenty-five years ago, Henry popularized the so-called “effect” of the Bradley loss. Lance Tarrance calls these musings “a theory in search of data.” I’d go even further: I don’t even think the underlying logic makes sense. I’d submit that human nature being what it is, if someone is embarrassed by the candidate they support—and there’s no reason anyone should have been embarrassed by George Deukmejian—they’d be more likely to refuse to answer than to lie. How do I know this? I don’t, but here’s some evidence: Two years ago, five statewide African-American candidates were on the ballot in this country. None under-performed based on the results of pre-election polling. In other words: No “Bradley effect” was present. To read Pew Research Center’s excellent examination, click here.

You’d think those results would have put this whole canard to rest. Think again. Charles Henry himself told a columnist for the Los Angeles Times earlier this year that Obama would need a double-digit lead for him “to feel confident” on Election Day. Of course, if race is the issue, the only relevant factor is whether Barack Obama’s status as the first African-American presidential nominee nets him votes: In other words, yes, many Americans (and not just whites) harbor unenlightened racial attitudes in the privacy of their own minds—and sometimes not just there—but are more Americans voting for Obama because of his race than against him for that reason?

On its face, this seems an easy question. Obama’s campaign narrative is based on his persona and his journey, and not his experience. There is, after all, a liberal Democratic senator from Obama’s state of Illinois with much more experience than Obama, and virtually the identical voting record. His name is Dick Durbin. I’ll bet you’ve never heard him mentioned as presidential timber. Of course, some academic will come up with another theory…oh wait…it’s happened again, only last month. Stanford University oversaw a poll on racial attitudes and came up with the conclusion that white prejudice might cost Obama up to six percentage points on Election Day. “There’s a penalty for prejudice and it’s not trivial,” a Stanford political scientist named Paul Sniderman told the Associated Press.

Apparently, there’s no penalty for fatuous political analysis, but that’s the subject of another blog-um. Loose Cannon is more optimistic than some people when it comes to the hearts of my countrymen. And I have still more facts on my side. Here’s one: In 1958, Gallup asked Americans whether or not they would be willing to vote for an African American for president. Fifty-three percent of the public said they would not vote for a “generally well-qualified person who happened to be a Negro.” In February of 2007, Gallup asked the same question, changing the wording slightly to reflect modern sensibilities: “Are you completely comfortable voting for a qualified presidential candidate who was black or would you have some reservations?” Ninety percent said they would be completely comfortable. That’s good news, folks, whether you’re supporting McCain-Palin or Obama-Biden.

And, oh yes, the answer to U.S. News’ conspiracy-minded question about why McCain is focusing on Pennsylvania: It’s not as sinister a rationale as racism: It’s because he’s running out of battleground states where he is competitive.