Here’s how we did our research
We drew our conclusions from a type of experiment that researchers call an audit study. It was designed to determine whether state legislators would respond differently to requests for information from constituents with different demographic characteristics — immigrant vs. native born, white vs. Latino, and so forth.
Other researchers have found that, on average, politicians tend to respond less to racial/ethnic minorities’ requests. There are a number of reasons this bias might apply to immigrant constituents. Ours is the first study to separate the different factors that might lead to bias toward immigrants and test their effects for immigrants from a range of racial/ethnic backgrounds.
Our sample includes all legislators who publicly provided their email addresses to the Sunlight Foundation, along with all legislators from the state of California. In total, the sample consists of 5,087 legislators from 42 states, or approximately 69 percent of all state legislators.
We created accounts from which we sent the legislators emails, written as if from a constituent. Each email contained a simple request for information, such as how a constituent could track the progress of a piece of legislation or information about visiting the state capitol.
The experimental manipulation involved changing the constituent’s name to suggest a particular race or ethnicity. The text also varied concerning whether the constituent stated that they were born in the United States or not, and described themselves as voters or not.
Overall, legislators responded about 36 percent of the time, which is similar to the response rates in other studies.
State legislators were as likely to respond to immigrants as to U.S. natives
Two of our results were somewhat surprising. First, we found that constituents who described themselves as foreign-born were as likely to get a reply as constituents who described themselves as U.S. natives. Our emails included no information about whether an immigrant email-writer was in the United States illegally or not, so we cannot say whether politicians are more or less likely to respond to undocumented residents.
Nor did it matter much whether email-writers described themselves as voters or not. Legislators might assume that anyone who contacts their offices is politically engaged enough to vote, whether or not they say so.
But state legislators were less likely to respond to racial and ethnic minorities
Compared with whites, the probability of a constituent with a black-sounding name getting a reply from a state representative was 3 percentage points lower. The probability for our Latino name was 7 percentage points less likely, and the probability for our Asian name was 9 percentage points lower.
That’s on average — but for Latinos, it varied by party. Democratic politicians were equally likely to reply to whites and Latinos; the probability of a Republican legislator replying to a Hispanic constituent was 9 percentage points lower, compared with a Democratic legislator. We can’t say whether this is a result of prejudice or whether GOP legislators assume that someone with a Latino name is probably a Democrat, and less worth their time. In either case, such a pattern could make it more difficult for Republicans to reach out to Hispanics, something that has been a concern for some in the party.
On the other hand, both Democrats and Republicans were equally unlikely to respond to a constituent with an Asian name. That may be because Asians and Latinos vote at a significantly lower rate than do whites and African Americans. And while Asians and Latinos lean Democratic overall, significant portions of both of these groups say they are ambivalent about partisan politics and are apolitical. But if both parties fail to engage with this growing constituency, members of these groups may well have reason to feel less than enthusiastic about politics and to stay away from the polls, making this self-reinforcing.
Micah Gell-Redman is an assistant professor in the departments of International Affairs and Health Policy and Management at the University of Georgia.
Neil Visalvanich is an assistant professor in the School of Government and International Affairs at Durham University in Britain.
Charles Crabtree is a PhD candidate in the Department of Political Science at the University of Michigan.
Christopher J. Fariss is an assistant professor in political science at the University of Michigan.