Project description

Amidst a broad crackdown on illegal immigration, this Reuters investigation found an asylum seeker’s chances of being allowed to stay in the United States depend largely on which judge hears the case, and where.

Analyzing more than 370,000 cases heard over 10 years in all 58 U.S. immigration courts, reporters found immigrants with nearly identical cases meet wildly different outcomes depending on the judges to whom they are randomly assigned. One judge in New York City, for example, grants 93 percent of asylum requests while another judge in Houston green lights only 4 percent.

Reporters found variance in outcome correlated to a judge’s background: Men are more likely than women to order deportation, as are judges who worked as prosecutors in immigration court. Judges with long tenure on the bench are more likely to grant asylum.

What makes this project innovative?

Reuters undertook an advanced statistical analysis of the data to account for other variables that can influence court outcomes. For example, U.S. government policy is more lenient toward people from some countries and less so for others. A judge also is bound by precedents established in the federal appeals court that oversees his district, and those precedents can vary. To account for these differences, Reuters built a multivariate regression model. Even after controlling for such factors, the analysis found that the judge who hears a case and where it is heard remain reliable predictors of outcome.

What was the impact of your project? How did you measure it?

The Reuters analysis determined that after controlling for some factors, who hears a case and where it is heard remain reliable predictors of how a case will be decided. The Reuters analysis also found that an immigration judge’s particular characteristics and situation can affect outcomes. Men are more likely than women to order deportation, as are judges who have worked as ICE prosecutors. The longer a judge has been serving, the more likely that judge is to grant asylum.The findings underscore what academics and government watchdogs have long complained about U.S. immigration courts: Differences among judges and courts can render the system unfair and even inhumane.

Source and methodology

The Reuters analysis used data from the Executive Office for Immigration Review (EOIR), the U.S. Justice Department unit that oversees immigration courts. The count of deportations included cases in which judges allowed immigrants to leave the country voluntarily.

The analysis excluded immigrants who were in detention when their cases were heard because such cases are handled differently. It also excluded cases in which the immigrant did not appear in court, which nearly always end in a deportation order, and cases terminated without a decision or closed at the request of a prosecutor.

About half the cases in the analysis were filed by asylum seekers like the two Honduran women. The rest were requests for cancellation of deportation orders or other adjustments to immigration status.

Technologies Used

Microsoft SQL Server, R.

Project members

Mica Rosenberg, Reade Levinson, Ryan McNeill, Ashlyn Still


Project owner administration

Contributor username


Click Follow to keep up with the evolution of this project:
you will receive a notification anytime the project leader updates the project page.