The United States has been a nation of immigrants since its founding, but the political battle over the specifics of modern immigration policy still rages. This range of attitudes can lead to challenges for those who are newly arrived in the United States; recent academic research indicates that immigrants are often segregated in new communities or even the victims of hate crimes.

While these attitudes are not merely felt in America — negative opinions of new immigrants have been documented in countries around the world — it is difficult to know what precise mechanism is causing these negative attitudes towards other ethnic, racial or religious groups. Are people’s opinions shaped by national media coverage? Are they the result of economic concerns? Is it merely that increased contact with a different group of people prompts negative stereotyping to occur? While observing interpersonal interactions in laboratory-style experiments can be useful, it does not allow researchers to note the effects of ongoing contact between groups in a more real-world setting.

Ryan D. Enos of Harvard University has conducted an experiment designed to simulate organic demographic change in a community. In his 2014 study, “Causal Effect of Intergroup Contact on Exclusionary Attitudes,” published in the Proceedings of the National Academy of Sciences (PNAS), Enos implements a randomized controlled trial (RCT) — an impact evaluation method adopted from medical testing and increasingly used by social scientists. Simply, RCTs involve choosing a sample of people and randomly assigning a “treatment” to some and not providing it to the others. In theory, this allows the researcher to attribute any difference in outcome between the two groups to the specific treatment administered.

Designing an RCT to examine whether attitudes shift as a result of changing demographics is difficult because of the inherent challenges involved in actively manipulating the ethnic make-up of a given area. To overcome this problem, Enos selected nine commuter rail stations in the Boston metropolitan area all serving homogeneous Anglo communities. In the first phase of the study, passengers waiting on train platforms were asked questions about their demographic and political characteristics, as well as their attitudes toward immigration policy and English as the official language of the United States. Enos then randomly chose certain trains to receive the “treatment” — a pair of native Spanish-speaking people waiting on the train platform with the commuters each day for two weeks. Assuming that groups of people who ride the same trains at the same time are statistically similar overall, Enos infers that the only difference in experience throughout the experiment was the presence of these Spanish-speaking passengers. After two weeks, follow-up questions on immigration were asked of the treatment and control groups to determine the effect of the exposure.

Key findings from the study include:

• The “treatment” passengers — those who likely came into contact with the pair of Spanish-speakers — displayed increasingly exclusionary attitudes with regard to immigration policies than the control group. They were more likely to advocate for a reduction in immigration from Mexico and less likely to support illegal immigrants being allowed to stay in the United States.
• Behavior from an outgroup does not have to be overtly threatening to provoke these exclusionary attitudes; the pair of Spanish speakers on the platform were not engaging in any abnormal, antagonistic behavior.
• The exclusionary attitudes were stronger among those passengers who definitely wait on the train platform — as opposed to in a car or parking lot — and were therefore more likely to be exposed to the Spanish-speakers.
• However, “repeated long-term exposure to an outgroup can mitigate initial negative reactions.” Treatment passengers were randomly assigned to be surveyed either after three days of exposure to the Spanish-speakers or after 10 days (two working weeks). Although the 10-day group still expressed more exclusionary attitudes towards immigrants from Mexico than the control group, the sentiments were less strong than the exclusionary attitudes expressed by the three-day group.

Overall, Enos concludes that these results indicate that the mere fact of ongoing demographic change in Western nations will likely be accompanied by exclusionary impulses and conflict at first. However, with more prolonged contact, these types of attitudes could be diminished.

Related Reading: For more background on the use of randomized controlled trials in the social sciences, see resources from MIT’s Abdul Latif Jameel Poverty Action Lab (J-PAL); for recent data on U.S. attitudes toward immigration, see a 2013 survey from the  Pew Research Center; for current statistics and details on U.S. immigration, see resources from the Migration Policy Institute.

Keywords: racism, Latino/Hispanic