How a Brief Experience Can Spark Anti-Immigrant Bias

The following is adapted from Let Them Eat Tweets: How the Right Rules in an Age of Extreme Inequality by Jacob S. Hacker and Paul Pierson, with permission from Liveright Publishing Corporation. (Copyright © 2020 by Jacob S. Hacker and Paul Pierson). Hacker and Pierson recently discussed their book during an online event hosted by New America.

Picture a crowded commuter train station at rush hour in a typical suburb. Two young men walk onto the platform and start speaking to each other. They’re good-looking, cheery, and well-dressed; people who see pictures of them describe them as “friendly.” They also look Hispanic—​like “immigrants,” according to those same photo-viewers—​and they’re speaking in Spanish. Virtually everyone else at the station is white.

Within a few days, the white passengers—​commuting from the overwhelmingly Democratic suburbs of Boston—​offer substantially more conservative responses to questions about immigration than they had before the arrival of the two men. They are more likely to say immigration from Mexico should be reduced and less likely to say the children of undocumented immigrants should be allowed to stay in the United States. Two people among hundreds encountered for a short time—​that is, signals that might seem pretty subtle—​are enough to create backlash among citizens inclined to support immigration.

The signals sent by American immigration over the last generation have been anything but subtle. Between 1990 and 2016, the share of the U.S. population self-identifying as Hispanic doubled from 9 percent to 18 percent. Immigrants and their immediate descendants now represent one in four Americans. More than half of children under age five are nonwhite. Births now contribute more to the growing Hispanic population than immigration. More than 90 percent of Latinos younger than 18 are citizens. Within Latino America, a group larger than San Francisco’s entire population becomes eligible to vote each year.

Against this backdrop, the story of those Boston commuters—​an actual experiment carried out by a Harvard political scientist in 2012—​looks more ominous. When outsiders breach the boundaries of established social groups, those within them often react with resentment, even revulsion. “In-groups” don’t just feel threatened by “out-groups;” they may seek to exclude them and deny them the benefits of community membership, with the force of law if necessary. According to decades of research, the outs don’t even need to be numerous relative to the ins for resentment to set in, certainly not numerous enough to pose any real threat. Even small changes in their numbers, if visible and proximate enough, can create a visceral response.

“In-groups” don’t just feel threatened by “out-groups;” they may seek to exclude them and deny them the benefits of community membership, with the force of law if necessary.

Demographic changes in America since the 1990s have not been small. By the midpoint of this century, the United States is expected to become a majority-minority nation. In truth, this well-known forecast is misleading. For one, the voting-eligible population greatly lags the national population aggregates, both because many of America’s minority residents aren’t citizens and because recent immigrant populations vote at lower rates than either Blacks or native-born whites. For another, much of the change will be driven by mixed families—​especially Asian-white and Hispanic-white families—​and children in these families often identify as white. But the shift is still dramatic. More important, today’s native-born whites see it as dramatic.

In a series of clever experiments, the psychologists Jennifer Richeson and Maureen Craig have shown that simply sharing population projections predicting that whites will become a minority produces big reactions, including anger, fear, greater identity with whites, and greater resentment toward nonwhites. At the same time, it produces a significant shift to the right on a range of issues, from those related to race (affirmative action, immigration), to those not about race but clearly racialized (health care, taxes), to those with no obvious racial connection (oil and gas drilling). Telling white people that they’re losing their dominant status produces a large and broad-based conservative response.

The conclusion seems unmistakable: white backlash is inevitable, and it invariably helps Republicans. Yet that’s not the only possible conclusion. Although the perception that out-groups are gaining ground does trigger in-group fear and anxiety, social scientists have found that the response of elites—​those with the power to shape how these changes are understood and how politics and policy get reoriented around them—​is crucial in determining the consequences. In Richeson’s experiments, simply telling white Americans that their social status wasn’t likely to change because of increased racial diversity wiped out all the effects of the demographic forecasts. Just reframing the projections that showed whites would become a minority—​without changing those projections in any way—​seemed to reassure white Americans that their initial feeling of threat was unwarranted.

No one in the Boston commuter train station experiment made immigration an issue; the commuters were forming their own attitudes based on their own perceptions. Those attitudes, it turned out, softened as time wore on. Within 10 days, the commuters were responding to the same questions about immigration with more welcoming views. By this point, it was impossible to be statistically confident their views were any different from those of commuters who hadn’t encountered the two Hispanic men. “People have started to recognize and smile to us,” one of the young Mexican Americans told the Harvard researcher. A passenger initiated a conversation with them by saying “the longer you see the same person every day, the more confident you feel to greet and say hi to them.” Presumably, if the experiment had continued for months or years rather than weeks, attitudes would have softened further. They might even have turned positive.