Poli-Sci Perspective is a weekly Wonkblog feature in which Georgetown University's Dan Hopkins and George Washington University's Danny Hayes and John Sides offer an empirical perspective on the issues dominating Washington. In this edition, Hopkins looks at how fears about foreign languages shape how we think about immigration. For past posts in the series, head here.

Federal immigration policy is back in the headlines, as a bipartisan group of senators released an outline of its proposed reforms and as a House committee held the first hearings of this Congress on the issue.  The deep partisan fissures on immigration are well known.  But what’s at the root of many native-born Americans’ concerns about immigration?  Ongoing political science research points in part to the issue of language.

When native-born Americans talk about immigration, language differences quickly come to the fore.  For instance, in a 2006 article, Pamela Paxton and Anthony Mughan report on focus group conversations about immigration.  To explain the importance of speaking English, one Los Angeles participant relayed a story from a trip to Burger King: he ordered a hamburger and received eight.  While Burger King might find such miscommunications profitable, the narrator certainly did not.  And he is not alone. A 2006 Pew Research Center survey found that among Americans who encounter non-English speaking immigrants, 43 percent found such encounters bothersome.  It’s little wonder that Paxton and Mughan conclude that “the importance that English-speaking Americans attach to language for successful assimilation cannot be overstated”--or that upwards of 90 percent of Americans believe that one must speak English to be American.  Questions about language are intertwined with many native-born Americans’ conceptions of national identity, as Deborah Schildkraut details in Press “One” for English.

One way to isolate the effect of Spanish is through randomized experiments.  In one set of experiments, researchers Van C. Tran, Abigail Fisher Williamson and I show that people who are exposed to just a single line of Spanish in a survey or exit poll can become more anti-immigration, especially if they encounter the language frequently in their daily lives.  Encounters like those in the Burger King seem to have a cumulative effect.  In another set of experiments, Todd Hartman, Benjamin Newman, and Charles Taber show that when a chatroom partner happens to mix English and Spanish, participants’ sense of cultural threat grows.  And separately, professor Hartman and his co-authors also find that people vary in the degree to which they push the costs of these inter-cultural exchanges onto the Spanish speakers themselves.  That is, some of their respondents ask the Spanish speaker to translate, while others take it upon themselves to do the translation.

What about Democrats and Republicans--do they react differently to these brief uses of Spanish?  To address that question, I conducted online experiments with the support of the Russell Sage Foundation in which people from across the United  States watched modified ABC News clips about immigration policy.  In those clips, the same unauthorized immigrant spoke either English or Spanish--and I provided subtitles in English in both cases, to make sure the respondents understood.  I also included a third, “clear English” condition in which the respondents heard a different person speaking with a very faint accent.  I then asked respondents about their support for creating a pathway to naturalization, one of the central questions in today’s debate over immigration.

It turns out that Republicans and Democrats respond differently to Spanish.  Two sentences of spoken Spanish have a significant, polarizing effect, making Republicans less supportive of creating a pathway to naturalization.  Indeed, the number of Republican identifiers who strongly oppose a pathway to naturalization jumps from 34.5 percent when hearing accented English to 40.5 percent when hearing Spanish--as the figure below illustrates.  Spanish has an especially negative impact among Republicans.

And the pattern isn’t just an artifact of the survey setting--it also holds when analyzing the impact of Spanish-language ballots in California’s 1998 primary election.  Republican precincts that voted with Spanish-language ballots were more supportive of curtailing bilingual education than similar precincts without such ballots.

It is critical to note, as Dylan Matthews did in this space last week, that research by political scientists Jack Citrin, Amy Lerman, Michael Murakami and Kathryn Pearson shows that today’s Latino immigrants acquire English and lose Spanish rapidly in the second generation, just like past generations of immigrants.  But concerns about English have nonetheless colored Americans’ reactions to immigrants from the days of Benjamin Franklin, and they remain an important element in the immigration debate today.  These concerns are widely shared among native-born Americans and prove especially influential among Republicans.  So don’t be surprised if proposals related to language, such as expanded English-language instruction, prove central in building legislation that can win support from Republicans and Democrats alike.