In a multiethnic church in Columbus, Ohio, white members addressed their minister by his first name. Black members viewed that as disrespectful, believing he should be addressed as “Pastor.” Conflict also broke out over disciplining children during worship services. Black parents tended to discipline their children when they were being disruptive, while white parents tended to let their kids move around.

Another multiethnic church, one located in Los Angeles whose core members were mainly Filipino Americans, faced similar discord. The whites and the blacks in the congregation were frustrated that they could not forge the deep friendships shared by Filipino American members. Conversely, some Filipino American members didn’t want to change their worship style to the hymns or gospel music that the black and white congregants preferred.

One church in Chicago selected both a white and a black pastor. They clashed over preaching styles until the church shut down.

In other words, integrating a church is rarely easy, and often leads to a litany of unintended slights and unrecognized biases. And this can happen at the earliest and most basic level: welcoming visitors. Do Christian churches in the United States actually welcome people from different racial and ethnic groups?

To answer this question, another sociological researcher and I conducted a nationwide field experiment to see how churches respond to emails from potential newcomers. More than 3,000 congregations received an email ostensibly from someone moving to their community and looking for a new church. We measured whether the churches replied to this email and, if so, what they said. But there was a catch: We varied the names attached to the emails so that they conveyed different racial and ethnic identities. Would the names alone change how churches replied?

Yes, they did—but not for the churches that we expected.

The Racism Under Our Skins

First, some background. The past half-century has seen profound changes—mostly improvements—in racial and ethnic equality in the United States. Into the mid-1960s, racial discrimination was written into our laws. These “separate but equal” laws created inferior social conditions for racial minorities, especially African Americans. The laws affected every area of life. Georgia had a law that no “colored” barber could cut the hair of white women or girls. North Carolina legislated that white and black schools could not share textbooks. Oklahoma segregated whites and blacks when fishing.

The Civil Rights Act of 1964 and related legislation abolished these laws. But prejudicial attitudes and discriminatory behavior remained. Survey data from that time reveals a white America explicitly and openly hostile to minorities. For example, most whites in the 1970s supported laws for racial discrimination in housing sales. A substantial minority supported laws against interracial marriage. Many thought that blacks had a lesser inborn ability to learn. Not surprisingly, minorities experienced inferior social conditions. For example, in 1970, only about 55 percent of Hispanic children and 65 percent of black children graduated from high school, compared to 85 percent of white children. Disparities remained around income, wealth, health, and other important outcomes.

In recent decades, explicit racial prejudice has decreased. Now, only a trivial number of white Americans would outlaw interracial marriage. Even fewer believe that blacks have less inborn ability. Currently, about 70 percent of Hispanic youth and 85 percent of blacks graduate from high school. Even our language reflects this change. Decades ago, derogatory use of the n-word by whites was not uncommon. Now it is a deviant, punishable act.

Article continues below

Although explicit racism has subsided, another form has surfaced: implicit racial bias. An implicit bias refers to stereotypes and attitudes that affect how we unconsciously think about and act toward something or someone. It can lead us to treat people differently because of their race and ethnicity without us realizing that we are doing it. A host of studies have demonstrated the prevalence of implicit racial bias in white America. This explains why few people think of themselves as racist while discriminatory behavior persists almost everywhere. We’re doing it; we just don’t realize it.

Few people think of themselves as racist while discriminatory behavior persists almost everywhere. We’re doing it; we just don’t realize it.

To test for implicit racial bias, researchers can’t ask people what they think. Most people will tell us what they want us to hear. Instead, we watch how people behave. A common research strategy is to put people—either real or fictional—of different racial or ethnic groups into everyday situations and see if they are treated differently. They usually are. Consider the following studies:

  • Whites and blacks were sent to car lots to bargain for a used car. Dealers initially quoted a price $700 higher for blacks than for whites, and gave black buyers fewer concessions (a 1995 study).
  • Doctors were shown patient histories and asked to make a recommendation about heart disease. They recommended a helpful surgical procedure less often for blacks than for whites—even though the medical files were otherwise identical (1999).
  • Job applications were sent to thousands of employers. Some applications had stereotypically African American–sounding names, while others had stereotypically white-sounding names. The applications with African American names received 33 percent fewer replies from employers than did those with white names (2003).
  • When a black hand was shown holding an iPod for sale on eBay, the auction received 17 percent fewer bids than when a white hand held it (2010).

This type of implicit racial bias has probably not gotten worse in recent decades. We have just become more aware of it as more overt, explicit racism has diminished. It is like a surgeon removing one cancerous growth only to find another one behind it. The insidious nature of implicit racial bias poses a major challenge for the church in America—because, of course, Christians aren’t immune to it. Most harbor implicit racial bias that makes it difficult to integrate races and ethnicities in corporate life.

Why We Have a Race Problem

US Christianity remains substantially segregated by race at both the denominational and congregational levels. Only 15 percent of religious congregations of any kind are racially mixed. (Sociologists generally define a congregation as mixed if fewer than 80 percent of members are from a single racial group.) For Protestant congregations, this number drops to 5 percent. Evangelical congregations display especially high levels of segregation. Imagine walking into the Sunday morning service of a typical evangelical church. You sit down. At that moment, if you were to randomly select two people in the service, there’s a 98 percent chance that they would both be of the same racial group. That’s segregation.

Segregation in churches reflects broader social trends, what journalist Bill Bishop has called “the Big Sort.” Americans have a lot of choice with whom we interact. We choose which neighborhoods we live in, which colleges we go to, which clubs we join, and which churches we attend. When we make these choices, we often choose to put ourselves among people who are like us—in political belief, cultural background, or religious preference. This tendency has created a nation of microcommunities. Across all these communities we are quite diverse, but within the communities we are homogenous. Thus, in daily life, we mostly interact with people who are similar to us.

Article continues below

This certainly describes me. When my wife and I went house-hunting, one street in town just felt right. When we drove down it, we could easily imagine happily living there. We eventually bought a house on the street, and today we live with a bunch of other professors and university staff who have children about the same age as ours.

A similar, subtle process happens within our churches. Without realizing it, we are attracted to churches full of people similar to us, and this similarity typically includes race and ethnicity. Protestant churches might be highly segregated because there are so many of them. With lots of choices, Protestants can find the church that’s “just right” for them.

But evangelicals aren’t simply more racially segregated. They also report relatively high levels of racially prejudiced attitudes. For example, in 2008, 34 percent of white US evangelicals reported being uncomfortable with a close relative or family member marrying an African American. This was significantly higher than mainline Protestants (28%), Catholics (25%), and the religiously unaffiliated (12%). Likewise, evangelicals feel less warmly toward nonwhites, and a small but meaningful minority of evangelicals is less likely to support a political candidate if they are black or Hispanic. Evangelicals are also more likely to support laws against interracial marriage and for racial discrimination in home sales. Sociologist Robert Putnam put it this way in his 2010 book, American Grace: “Evangelicals are no less—and perhaps even more—racist than members of other religious traditions.”

Sociologists have attempted to explain the racial prejudice of evangelicals in various ways. Evangelicals have a strong in-group identity that gives them a sense of distinctiveness—an “us versus them” view of the world. (In general, religious groups that believe in absolute truth have stronger in-group identities.) This in-group identity serves Western evangelicals well in a secular society because it helps them to resist opposition from outside the church. Unfortunately, it can also foster a general preference for people who are similar along religious and other social dimensions, including race and ethnicity. Thus, a strong in-group identity in conservative religious traditions, including evangelicalism, has been associated with degradation of people from other racial groups.

In-group identity serves evangelicals helps to resist secular opposition. Unfortunately, it can foster a preference for people who are similar.

Another factor in evangelicals’ racial attitudes is their tendency to discount the power of larger, structural forces. The US historical liberal–conservative political split—with the former focusing on entrenched systems and the latter emphasizing personal morality—comes into play here. As evangelicals are more likely to align with conservative political views, they are less likely than mainline Protestants to use language celebrating racial diversity and more likely to oppose political and social changes, such as affirmative action, that create socioeconomic opportunities for African Americans.

Looking for the Right Mirror

My interest in race and religion came after publishing my 2010 book, Christians Are Hate-Filled Hypocrites … and Other Lies You’ve Been Told. In it, I examined common critiques of American evangelicals. Using sociological research, I showed that many of those critiques are overstated or flat-out wrong.

Article continues below

All except for race. Reviewing the studies I just described (and, unfortunately, there are more), I concluded that indeed, as a group, evangelicals have a race problem.

At that point, I decided to do my own study. I thought that if I could demonstrate to evangelicals just how bad the race problem was in their churches, it might effect positive change.

I didn’t want to do yet another survey on attitudes. As I stated above, attitudinal surveys can’t capture implicit racial bias. Instead, I wanted to study how evangelicals behave.

During this time, a colleague, Mike Wallace, and I had been working on a separate line of research. We tested for religious discrimination in the job market by sending fake résumés to thousands of advertised jobs. The résumés mentioned involvement in one of several religious groups. We found clear evidence that employers discriminated by religion—especially against Muslim applicants. This type of field experiment appealed to me. It powerfully tested for discrimination, and it told a compelling story. (I wrote about it in a June 2014 CT article.) But how could the experiment apply to a church setting?

I mulled over this for a month. Then one day, I remembered my pastor telling me that potential visitors would sometimes call or email him before visiting to introduce themselves and get information. That was it. We could test how churches responded to potential newcomers. I ran this idea by Mike, and we were off—blissfully unaware of how much work we were in for.

The Email Test

Our first big research decision was how to contact the churches. We thought about hiring a team to visit churches in person, but this seemed too expensive and posed geographical limitations. Instead we decided to send emails. After all, most churches publish their email addresses online, and it’s how most of us often communicate.

Next we had to write a letter. It would go to every church, so it had to be general enough that it made sense to people in different types of churches but specific enough to gather useful information. We talked to Christian friends to see which questions they would ask if they were contacting a new church, and we ended up pretending to be a family moving to the church’s community. Our letter asked about the size of the church and service times, and requested additional information that might be helpful.

But if we had just this letter and nothing else, we wouldn’t have a study of race. I suppose that we could have sent the letter to churches and learned who was best at replying to potential visitors. But we needed somehow to convey the letter writer’s racial or ethnic identity. We couldn’t figure out how to do this within the text of the letter. What would we write: “My white family and I are moving to your community”?

Instead, we varied the names that signed the letters. We had decided on using four different racial and ethnic groups: African Americans, Hispanics, Asian Americans, and whites. So we needed names that people stereotypically associate with these racial groups. We studied US Census data and talked to people from each ethnic group, and we ended up with eight names:

  • Greg Murphy and Scott Taylor (white)
  • Jamaal Washington and Tyrone Jefferson (black)
  • Carlos Garcia and Jose Hernandez (Hispanic)
  • Wen Lang Li and Jong Soo Kim (Asian American)

With the letter and names, we had our experimental intervention all ready. Now we needed churches. Each church would get only one email, so we needed a lot of churches. There are 300,000 or so congregations in the United States—far too many for us to email. Instead, we compiled the 436 congressional districts in the United States—since each district has a more or less equal number of residents—and randomly picked 65, then selected churches from their communities. We ended up with Mobile, Alabama; El Paso; South Bend, Indiana; Minneapolis; Boulder, Colorado; San Francisco; Chugiak, Alaska; and a whole lot of places in between.

Article continues below

We wanted to compare the three major Christian religious traditions in the United States: evangelical Christian, mainline Protestant, and Catholic. So we selected five different denominations in the mainline Protestant tradition: the United Methodist Church, the Episcopal Church, the Evangelical Lutheran Church in America, the Presbyterian Church (USA), and the American Baptist Churches (USA). We then selected six different denominations and networks in the evangelical tradition: the Southern Baptist Convention, the Assemblies of God, the Lutheran Church–Missouri Synod, the Churches of Christ, the Willow Creek Association, and self-described Pentecostal churches. (We also selected Catholic parishes, though after we finished our analyses, we realized that we had not selected enough of them.)

We selected churches from each of these 12 groups in the communities that we had sampled. Most denominations and organizations have online directories for finding churches by location. When they didn’t, we used Google. After this time-consuming process, we had the names and email addresses of 3,120 randomly selected churches—about 1 percent of all churches in the country.

This meant that we had to send 3,120 emails. We recruited six undergraduate students whom we knew to be careful, thoughtful workers. We offered them the opportunity to learn about social research and to aid a groundbreaking study. When that didn’t work, we told them that we’d buy them pizza.

We reserved the department computer lab for two afternoons, and the eight of us (six students, Mike, and I) lined up in front of computers. On one side of each screen was a list of several hundred churches, and on the other side was a fictitious email account. In a flurry of typing, we started sending emails. After about 15 minutes, we heard an unexpected chime. None of us had our personal email accounts open, so the sound could only mean one thing: A church had replied to one of the emails. We all stopped.

The student whose account received the reply proudly read it aloud. It was a kindly message from a Baptist pastor welcoming Tyrone Jefferson to his community. The pastor answered Tyrone’s questions and said that he looked forward to meeting Tyrone in person. The study finally seemed real. After so many months of meetings and endless detail work (you try randomly selecting 3,000 churches), we had real data. We ended up hearing back from more than half the churches.

Why the Study Works

Here’s the logic of our study. Each church would receive one email. Some churches would reply and some would not. We couldn’t know if the actions of any single church were discriminatory. For example, suppose Jose Hernandez sent a letter to St. Theodore’s in Des Moines and a church staff member didn’t reply. What would this mean? It could mean that St. Theodore’s had a really busy week launching their mixed martial arts–themed children’s ministry, and they forgot to write back. Or maybe they never reply to anyone. Or maybe they were put off by Jose’s apparent Hispanic ethnicity. We don’t know.

So no single email response tells us if a particular church harbors implicit racial bias. Where we can detect churches’ bias is in the aggregate. Since each church got the same letter at about the same time, if churches replied differently to different names, it had to be due to the names themselves. And if Jong Soo Kim’s letters received significantly fewer replies than Greg Murphy’s, the disparity reveals a difference in the status conferred upon them.

Article continues below

Experiments usually work like this. For example, if a drug company wants to test the effectiveness of a new medicine, it recruits subjects and gives the new medicine to some of them and something else (a placebo, an old medicine) to the others. If the subjects who took the new medicine have different health outcomes, it is attributed to the new medicine.

At this point, you might be questioning the ethics of our study. After all, we lied to the churches. The letters were not from real people. Universities have rigorous and mandatory procedures for evaluating the ethics of proposed studies, and the key determination is whether a study’s potential benefits significantly outweigh its costs. For our study, the main cost is that we asked church representatives to take several minutes to reply to an email. This is a relatively low cost. In return, we tested for racial discrimination in churches nationwide. This is potentially quite important.

Checking the Inbox

We didn’t know how long we’d have to wait until churches stopped replying to our emails. We ended up waiting for two months. A total of 1,830 of the 3,120 churches (59%) replied to us at least once. Most churches replied only once, some wrote multiple times, and a few even put us on their church newsletter list. Catholic parishes replied more often (66% did) than did evangelical or mainline Protestant churches (58% each). The best responders overall were churches affiliated with the Willow Creek Association (72%) and Episcopal churches (70%). The worst responders were Southern Baptist, Presbyterian, and Pentecostal churches (49% each). The church replies were usually written by a pastor, administrator, or lay leader.

As far as we know, none of the church staff figured out that the emails were from a fictitious person, though we had one close call. One of the Pentecostal letters went to a church in remote Alaska, and another one went to a church in a small town in Texas. By sheer coincidence, the pastor of the church in Alaska was relocating to the church in Texas, and the timing was such that he received a letter from us at both churches. Luckily, the two letters happened to have the same name attached to them (a 1 in 8 chance). His reply sounded a bit puzzled, but he generously offered to help once we knew where we were going.

Taking all 3,120 churches together, we found moderate differences in reply rates by race and ethnicity. For every 100 churches that replied to letters with a white-sounding name, 93 replied to those with black- or Hispanic-sounding names, and 85 to Asian-sounding names. We expected something like this. The big surprise came when we analyzed the three religious traditions separately.

Evangelical and Catholic churches varied little across the letters. For every 100 evangelical churches that responded to white-sounding names, 97 replied to black names, 100 to Hispanic, and 94 to Asian. These differences were not statistically significant. Likewise with the Catholic parishes. (Though, since we included relatively fewer parishes, our tests of them were statistically weaker than for evangelical and mainline churches.)

The mainline Protestant churches? That is where the discrimination happened. For every 100 mainline churches that replied to white-sounding names, 89 replied to black names, 86 to Hispanic, and only 72 to Asian. Think about it. A letter simply having the “wrong” name significantly reduced these churches’ likelihood of welcoming a potential visitor.

Article continues below

Do Minorities Get Worse Replies?

One of the most interesting parts of the study was reading through the churches’ replies. In general, most of the replies were warm in tone and informative in content.

“We hope you will take a minute to visit our website and decide to come see us upon your move,” one church wrote to Carlos Garcia. “We would love to have you come and check us out to see if we are what you are looking for,” another wrote to Scott Taylor.

Some replies went much further and offered spiritual or logistical support. “I’ll be in prayer for your search for a new church home … that the Lord will lead you to the right church family where you can connect and serve!” said a letter to Tyrone Jefferson. Wen Lang Li was offered a place to stay while he looked for housing. Several churches offered to meet the new family when they arrived in town and to help them unload the moving truck. That impressed me; frankly, I wouldn’t do that even for family and friends.

Several churches offered to meet the new family when they arrived in town and to help them unload the moving truck. That impressed me.

Every once in a while, though, a reply would be curt. We ended up labeling these “terse” replies, and we defined them as being short in length, cool in tone, and relatively uninformative. For example, one church replied to Jamaal Washington without any greeting or salutation. They simply wrote: “The best way to get acquainted with us is to visit our website at ------.com. Our church is about 500–600 members.”

Few churches sent terse replies, but when they did, guess which letters received the tersest replies? Yes, it was the nonwhite letters. About 5 percent of all replies sent by mainline Protestant churches were terse. When replying to whites, this figure was only 2 percent. To blacks it was 3 percent, to Asians 5 percent, and to Latinos 10 percent. (Evangelical churches displayed a similar but less pronounced pattern.)

When replying to nonwhites, mainline churches were also less likely to describe how their church worshiped; they gave overall lower quality information; they sold the church less; and they were overall less warm. Mainline churches sent the most informative and welcoming replies to whites, the least to blacks, and Hispanics and Asians were in between.

‘One Heart at a Time’

The results really surprised me. Going into the study, it had never occurred to me that mainline Protestant churches would display implicit racial bias and evangelical churches wouldn’t. These results were so unexpected that I redid the analyses several times just to be sure.

The obvious question is why. We went back to the research literature and started looking for potential explanations.

One explanation centers on evangelism. Arguably more so than mainline Protestants, evangelicals strongly emphasize spreading their faith across cultural boundaries—indeed, throughout the world. This perhaps makes them more willing to connect with people across racial and ethnic boundaries, at least to promote their religious activities.

Another explanation centers on perceptions of racial justice and fairness. What matters to evangelicals is how you treat the person in front of you. The phrase “You can’t legislate love” conveys the typical evangelical approach to transforming society “one heart at a time.” From this perspective, racial discrimination is rooted in poor relationships and personal sin, not in systems of injustice perpetuated by the government, media, or education. It happens when individuals treat other individuals wrong. The solution, then, is also at the individual level. It involves having a faith commitment and loving others—even those who are different. While this individualistic approach might lead evangelicals to distrust broader, societal-level programs for addressing racial inequality, it can make them more welcoming to the visitor in their midst.

Article continues below

Conversely, mainline Protestants’ pursuit of racial justice at the societal level appears not to trickle down into interpersonal behavior. They need more than just resolutions and presentations at denominational conventions. Perhaps mainline churches were less welcoming to racial minorities due to an implicit belief that minorities would fit poorly into their congregations. Mainline Protestant worship services typically use liturgies, music, and readings historically rooted in European cultures. This style of worship might be of less interest to people from other cultural backgrounds. In fact, studies have found that expressive, contemporary worship styles are most effective for transcending racial boundaries.

If the mainline Protestants who responded to our emails thought that people with nonwhite names might not fit well into their congregations, they might have changed how they replied without even realizing it. This is like a math teacher implicitly assuming that girls can’t do math as well as boys, and thus giving them less attention and fewer opportunities. Even if it’s not intentional, it’s still discrimination.

We also didn’t expect that Asian-sounding names would receive the fewest replies. Perhaps the mainline Protestant churches assumed that Wen Lang Li and Jong Soo Kim did not speak English as their first language. Given the centrality of language in religious practice, this would be another obstacle to them integrating into a congregation.

Our findings offer some clarity. We know that evangelical churches are mostly segregated, but they also do a good job of welcoming potential visitors from various racial and ethnic groups. This means that something other than welcoming behavior is generating the observed segregation. Maybe it’s a problem of attraction, with churches only drawing visitors who look similar to the people already there. Or maybe it’s a problem of assimilation, with newcomers who look different not being brought into the daily life and leadership of a church. Certainly evangelical churches seeking racial and ethnic integration have a long, difficult path ahead of them. But, as we have found, they have a solid foundation of welcoming behavior to build upon.

Article continues below

Bradley Wright is a sociologist at the University of Connecticut, where he studies spirituality and well-being. He blogs at brewright.com. The research described here is published in the June issue of the Journal for the Scientific Study of Religion.

Issue: