Americans’ proclivity toward binary thinking was on full display in Brett Kavanaugh’s response to allegations of sexual assault. It consisted primarily of a fuming recitation of his close women friends, the advancement opportunities he’s offered to women (albeit to women with a “certain look”), and his reputation as a committed husband, father and coach of girls’ athletics. The logic, though unspoken, was clear enough: his treating some (or most) women with decency precluded the possibility that he could treat any woman poorly. If it reminded you of the guy who insists he’s not a racist because “I have black friends,” you’re not alone.

In a recent New York Times editorial, Emily Ekins of the Cato Institute uses the same binary framing to argue that the forces of anti-racism have allies in unexpected quarters. “Religious conservatives,” she argues, “are far more supportive of diversity and immigration than secular conservatives.” They “have more favorable feelings toward African-Americans” and other groups who have faced discrimination. Religion might make conservatives stubborn opponents to same-sex marriage, she admits, but it’s been a moderating force on issues of race.

A thirty-page report, on which the editorial is based, lays out the evidence. Limiting the scope to Trump-voters, Ekins homes in on “religious conservatives”—those who attend church once or more weekly. The largest proportion of this group self-identify as evangelical: a full 60% (and 76% of the more than once-weekly crowd). The remainder are Catholic, other Protestants, a handful of other religious groups, or “nothing in particular.”

Ekins determined racial views by analyzing a series of “feeling thermometer” questions. Respondents were asked to quantify their “warmth,” or lack thereof, toward a range of racial, ethnic, and ideological groups. “Ratings between 50 degrees and 100 degrees mean that you feel favorable and warm toward the group,” the survey explains, while numbers below 50 represented unfavorable feelings. A majority of religious conservatives feel positively “toward Black People, Hispanics, and Asians.” And the highest numbers, in some cases topping 80%, correlate with frequent church-going.

According to Ross Douthat, all this suggests that “churchgoing Republicans look more like the party many elite conservatives wanted to believe existed before Trump came along—more racially-tolerant, more accepting of multiculturalism and globalization.” The religious faithful have Black friends, it seems, or at least they want to.

For white evangelicals, Ekins ostensibly offers empirical evidence that religion has positive effects on race relations. But as a historian of religion, I have my doubts. Well-meaning evangelicals like Michael Gerson have also argued for the movement’s racial progressivism—suggesting that abolitionism was a central preoccupation of nineteenth-century evangelicals. But historians know that for every anti-racist evangelical, there were scores more supporting white supremacy. In the same way, a handful of evangelicals supported the Civil Rights movement, but a much larger majority opposed it; as many, I’d wager, as the ninety-odd percent of religious conservatives who oppose Black Lives Matter today.

We would do well, then, to carefully evaluate these claims of racial enlightenment before accepting them at face value. What is this survey really measuring and what do the results actually tell us?

Tackling questions about race requires precise definitions. Fortunately, Ibram X. Kendi’s National-Book-Award-winning Stamped from the Beginning: The Definitive History of Racist Ideas in America provides exactly what we need. Racism, he explains, is simply the belief that one racial group is superior or inferior to another. Racist ideas stem from a simple question: why do the prospects of some racial groups look so different from others?

Three answers were developed to explain racial disparities. Segregationists asserted that some racial groups (especially those of African descent) were intrinsically, and irredeemably, inferior to white Europeans. A second group, anti-racists, claimed the opposite: all racial groups had equal capacities, but discriminatory laws and policies caused disparities. Both of these groups are well known. They dominate our binary narratives of racial oppression. Almost everyone sees themselves as anti-racists.

But Kendi crucially highlights a third position, assimilationist, that’s typically overlooked. Assimilationists presumed the superiority of whiteness, but blamed environmental causes for racial inferiority—from hot climate, to inferior culture, to the effects of slavery. Essentially, they tried to have it both ways. They rejected racial animosity and insisted that individuals from inferior racial groups could achieve parity with whites. But whiteness remained the standard.

Of course, the problem is that you can’t have it both ways. Either all groups are equal or they aren’t. Assimilationists still answered the question of unequal prospects by blaming the racialized victims of oppression rather than the racist system. Whether or not they liked non-white individuals was beside the point.

Acknowledging assimilationist racism explodes the simplistic binaries that often structure popular narratives. Nearly all northern white Protestants were staunch assimilationists—even those agitating for abolition and supporting the Civil War. And not all southern slaveholders harbored racial animosity. Some justified their exploitation of enslaved people with assimilationist reasoning—telling themselves that they were helping the people under their control. Some slaveholders had warm feelings toward those they enslaved. And although the Civil Rights movement drove segregationist racism underground, assimilationist racism continues openly and unabated.

Kendi’s categories also reveal what the survey results measure. Church attendance moderates racial animosity among Trump voters; that’s good news. But we can only determine whether a person holds racist opinions with a different set of questions: those that inquire about the causes of inequity between racial groups.

What’s frustrating about the report is that the raw survey data on which it’s based had questions that did precisely that—only they ‘re never mentioned. The other data tell a very different story.

“It’s really a matter of some people not trying hard enough;” one query asserts, “if blacks would only try harder they could be just as well off as whites.” To this bald claim that racial disparity exists because African Americans don’t try hard enough, 82% of frequent church-going Trump-voters agreed.

Similarly, 91% agree that “Irish, Italian, Jewish, and many other minorities overcame prejudice and worked their way up. Blacks should do the same without any special favors.”

Have “Blacks…gotten less than they deserve” in the “past few years?” Nine out of ten say no.

Asked whether “generations of slavery and discrimination have created conditions that make it difficult for blacks to work their way out of the lower class,” 87% of religious conservatives disagree. They apparently believe African-American poverty is completely unrelated to centuries of slavery and discrimination. Yet three-quarters of these voters claim that “discrimination against whites has become as big a problem as discrimination against blacks and other minorities.”

All this contradicts the assertion that religious conservatives are progressive on racial issues; overwhelming majorities agree with overtly racist myths.

But what about the main point: that religious conservatives are less racist than their non-church-going counterparts? Unlike feeling thermometers, these questions generate no discernable correlation between church attendance and less racist views. They demonstrate only that church-going Trump voters are more likely to be assimilationists, while segregationists cluster among non-church-goers. Church-goers generally reject overt forms of race hatred, but nearly all Trump-voters blame racial disparities on racial inferiority, not structural inequalities.

Sixty-seven percent of religious conservatives say racial equality is important; I take them at their word. But thanks to these sorts of misleading studies, well-meaning people don’t realize how much a part of the problem they are. This report is the product of a disingenuous fishing expedition: a hypothesis confirmed only by ignoring the totality of the data.

Evangelicals who want to become anti-racists should ignore right-wing think-tanks and the hagiographical narratives that make evangelicals into the champions of racial equality. Instead, start listening to those among you who experience the racism you insist does not exist. These activists, believers all, have said it again and again since Ferguson. White evangelicals, and religious conservatives more broadly, have a race problem—one hidden by good guy/bad guy binaries.

The appeal of binary thinking is not limited to conservatives; it’s found among all white Christians, including my corner of the religious world. Presumably, white religious Clinton voters see themselves as progressive on race, yet an embarrassing 29% of frequent church-attenders agreed that “if blacks would only try harder they could be just as well off as whites.” That’s a full 10 points higher than Clinton-voters who never attend church. Sit with that for a moment.

A recent review of Robin DiAngelo’s White Fragility noted that a binary “world of evil racists and compassionate non-racists, is itself a racist construct.” The reality, she writes, is that “white progressives cause the most daily damage to people of color.” If white Protestant progressives want to take their place in the anti-racist charge, they must first entertain the prospect that we are the worst transgressors of them all.

Faith is not a product to be marketed with claims of perfection, it’s a flawed human experience in constant need of reform. Racist ideas are a cancer that has grown within American Protestantism, conservative or not, for hundreds of years. Addressing it begins with admitting there’s a problem.