Did Donald Trump's 2016 presidential campaign benefit from voters' fears of immigrants in communities experiencing greater demographic change?
New research shows the answer is "no," a finding that contradicts the conventional wisdom and which surprised even the political scientists who conducted the study. Instead, those communities actually moved more toward the pro-immigration Democratic candidate.
In a paper published in the Proceedings of the National Academy of Sciences, political scientists Daniel J. Hopkins of the University of Pennsylvania, Seth J. Hill of the University of California, San Diego, and Gregory A. Huber of Yale University, describe their novel approach to the question.
They were looking to see if demographic changes from 2012 to 2016 shifted voters toward an anti-immigration presidential candidate.
Rather than look at large geographic areas like counties, as previous studies have, they analyzed a much smaller section of communities: voting precincts.
The challenge with using U.S. counties to study the question is that some are sparsely populated and others have millions of residents, says Hopkins. He and his colleagues wanted to drill down to a more precise and local level.
Hopkins has been studying the question of local demographic changes and influxes of immigrants for years. His prior work and that of others has indicated that those changes should be expected to produce shifts in local politics, he says.
"One of the key things I've found is that people, in explaining their unease about immigration, talk about very local encounters. They say it's challenging to see all of the grocery store signs in Spanish, or that on the phone the bank asked them if they wanted to 'Press 1' for English or 'Press 2' for Spanish," he says. "We were very interested in whether local demographic changes were part of the explanation for the election of Donald Trump, and more generally for the rise in anti-immigration populist political parties and candidates in recent years."
The three researchers teamed up to collect the deep and unique data set. They compiled election results and demographic measures for more than 26,000 precincts in Florida, Georgia, Michigan, Nevada, Ohio, Pennsylvania and Washington.
They chose states that were large, diverse, and politically contested and that jointly captured some of the key demographics that drive American politics: States with large cities like Pennsylvania; western states with large Latino or Asian American populations like Washington; key northeastern and midwestern post-industrial states that were decisive in 2016, like Ohio and Michigan; and larger southern states with very diverse populations, like Florida and Georgia.
They had to work with secretaries of state and state election offices to get individual-level voter turnout data, and then they worked with those offices or other scholars to identify precinct-level electoral data. Precincts are quite small, often with just 1,000 voters, and the researchers contended they would be a much better measure of people's local, lived experiences in their immediate communities.
"A common argument for why candidate Trump won the 2016 election is that he benefited from local demographic changes in the lives of native whites. Our evidence does not support that argument," Hill says. "While it is possible Trump benefited from anti-immigrant sentiment, in the states we examined his vote gains relative to 2012 do not seem to follow from Americans' local lived experience with immigration or demographic change."
Meanwhile, Trump's opponent Hillary Clinton saw increases in votes in precincts with growing shares of Hispanics and immigrants. While the authors cannot make a causal claim about why Clinton benefited, they use their precinct observations to show that even in precincts that strongly voted Republican in 2012, increasing diversity led to movement toward Clinton. This implies the overall pattern was not generated simply by immigrants and Hispanics moving exclusively into already heavily Democratic precincts.
"It may be that native-born citizens have some discomfort with demographic change but that discomfort either declines over time or it is not as important for their presidential-vote choice as factors such as policy views or candidate characteristics," Hill says.
Nationalization is a powerful trend, Hopkins says, and immigration-related political appeals can resonate in a wide range of different communities.
"We're not saying that demographic change doesn't reshape our politics, but what we are saying is that demographic changes at the local level do not seem to be what drove many voters to Donald Trump."
Daniel J. Hopkins is a professor in the Political Science Department at the University of Pennsylvania, with a secondary appointment in Penn's Annenberg School for Communication. He assists with the coordination of the Philadelphia Behavioral Science Initiative.
Seth J. Hill is an associate professor in the Department of Political Science at the University of California, San Diego.
Gregory A. Huber is the Forst Family Professor of Political Science in the Department of Political Science at Yale University.