Skip to main content

The Journal Gazette

Friday, May 10, 2019 12:00 pm

Fact-checking the Trump administration's immigration fact sheet

SALVADOR RIZZO | Washington Post

When we see politicians or government officials writing fact checks of their own, we break out our magnifying glass.

Today, we're inspecting a new "myth vs. fact" document from the Executive Office for Immigration Review, the arm of the Justice Department that runs the U.S. immigration court system.

The five-page document, released this month, attempts to debunk 18 claims about immigration to the United States.

In some cases, it seems more like EOIR officials are misusing the fact-checking format to make a point about issues that no one is mischaracterizing.

For example, one of the supposed myths is that immigration judges and the attorney general "are prosecutors." We're not aware of any group, expert or journalist spreading this notion (and the attorney general is a prosecutor despite his concurrent role as head of the immigration courts).

In other cases, EOIR officials dispute common claims about immigration with cherry-picked data or questionable information. This is what we focused on, picking out four of the 18 claims.

MYTH: There is wide discrepancy in asylum grant rates across all immigration courts.

FACT: The median asylum grant rate for all immigration courts is 11%. Eighty percent of immigration courts, 50 out of 62, have a grant rate of 21% or lower. Only 12 out of 62 courts have grant rates more than 10% above the median grant rate. Only one out of 62 courts has a grant rate above 50%.

It's no myth that the rate of granting asylum varies widely by judge and by location.

In a detailed data analysis of immigration judges who decided more than 100 asylum cases from fiscal 2012 through 2017, the nonpartisan TRAC research center found that, out of the 35 immigration judges in New York, the one with the lowest denial rate (3%) and the highest (58.5%) were still far apart from their peers in Los Angeles. In that city, the 34 immigration judges ranged from a denial rate of 29% to 97.5%.

Those are two of the biggest immigration courts by caseload, but these discrepancies also pop up in other parts of the country.

The median grant rate for asylum claims was 11% in fiscal 2018. Some localities were lower (Atlanta, El Paso, Las Vegas). Cities including Baltimore, Boston, Chicago and New York were much higher than the median. Honolulu was above 60% -- five times the median rate.

"Eighty percent of immigration courts are clustered within 10% of the median grant rate," said EOIR spokeswoman Kathryn Mattingly. "Although 12 courts are outliers, EOIR does not consider 12 out of 62 to reflect a wide discrepancy."

MYTH: Few aliens fail to attend their immigration court proceedings.

FACT: Forty-four percent of all non-detained removal cases end with an in absentia order of removal because of an alien's failure to attend a scheduled immigration court hearing.

Immigration judges typically issue deportation orders "in absentia" when immigrants fail to show up for their scheduled hearings.

But this, at first glance, seems like a case of cherry-picking data. The EOIR document focuses on in absentia orders for 44% of "non-detained removal cases." That's one of several categories tracked by immigration officials. The numbers look different when measuring asylum cases alone or when combining all the different categories.

Judges ordered in absentia removals in 14% of all cases in fiscal 2013, a rate that grew to 28% in fiscal 2017. Looking at asylum cases alone, the rate was 6% in 2013 and 11% in 2017. That means the vast majority of asylum applicants did show up.

Asked why EOIR officials focused on non-detained cases, Mattingly said the asylum category includes some immigrants who are in custody and unable to skip their hearings. However, Mattingly did not directly address the "all cases" category.

"Only non-detained aliens are potentially subject to an in absentia order of removal," she wrote in an email responding to detailed questions. "By definition, a detained alien generally would not be subject to an in absentia removal order because the Department of Homeland Security is responsible for ensuring all aliens in its custody attend hearings. Consequently, including detained aliens in a measure of the in absentia rate would present a misleading picture of how many aliens attend hearings." (Yet this is what EOIR does in its own statistics yearbook.)

This item in the EOIR's document nonetheless contradicts President Donald Trump's repeated claim that immigrants "never show up" for their immigration hearings.

MYTH: Most aliens with representation are granted asylum in immigration proceedings.

FACT: The asylum grant rate for cases with representation is approximately 21%. The asylum denial rate for cases with representation is approximately 48%. These rates are essentially the same as the national averages.

This portion of the EOIR document lacks necessary context, since it focuses only on asylum grants and denials for immigrants with lawyers. Missing is a side-by-side comparison with immigrants who don't have lawyers.

In lieu of that comparison, the EOIR document says the grant rates for lawyered-up immigrants "are essentially the same as the national averages." But the word "essentially" is being stretched beyond meaning.

"Having a lawyer makes a huge difference across the immigration court system," Aaron Reichlin-Melnick, a policy analyst at the American Immigration Council, wrote in a Twitter thread analyzing the EOIR document.

The TRAC research center has a table comparing asylum applicants who did and didn't have lawyers from fiscal 2012 through 2017. Overwhelmingly, those with legal representation had a higher success rate.

Take the top five countries for asylum applications -- China, El Salvador, Guatemala, Honduras and Mexico. In each case, denial rates were appreciably higher for asylum applicants without representation. Chinese nationals were the biggest group of applicants. Those without lawyers were denied asylum in 79% of cases, yet for those with lawyers, the rate was a fraction of that -- 18%.

"Multiple sources have claimed that representation makes an alien more likely to win an asylum case; however, the grant rate for represented asylum cases mirrors the national average overall, and most asylum applications are not granted regardless of representation," Mattingly said.

MYTH: Video-teleconference hearings are unreliable, and their use violates due process.

FACT: VTC has been used by EOIR since the 1990s, and its use was expressly authorized by statute in 1996. It is used widely throughout many federal agencies, and federal courts have consistently rejected general challenges to its use as a violation of due process. There is no indication of a statistically significant difference in outcomes between VTC cases and in-person cases. Less than one-tenth of 1 percent (.0052%) of EOIR VTC hearings, 310 out of nearly 60,000, are continued because of a VTC malfunction.

It's unclear how EOIR officials evaluated the reliability and due process issues involved in these hearings. A 2017 report by the Government Accountability Office found that "EOIR has not adopted the best practice of ensuring that its VTC program is outcome-neutral because it has not evaluated what, if any, effects VTC has on case outcomes."

The GAO also noted that a federal government manual for internal controls (known as "the Green Book") says "agencies must have relevant, reliable, and timely information relating to internal as well as external events to manage the agency's operations.

EOIR, though, does not collect reliable data on (1) the number of hearings it conducts by VTC, (2) respondent appeals related to the use of VTC in their cases, or (3) motions filed by respondents requesting in-person instead of VTC hearings."

It seems as though Justice Department officials tabulated the data for cases postponed because of technical difficulties ("less than one-tenth of one percent"), though we note that the time period for this statistic is not included in the "myth vs. facts" document.

The GAO report indicates that salient data points were missing in 2017, such as the number of video-teleconference hearings conducted, the number of appeals related to the use of this technology, and motions requesting in-person hearings. Yet EOIR says "there is no indication of a statistically significant difference in outcomes between VTC cases and in-person cases."

"Case outcomes do not measure due process, as a respondent may receive due process regardless of whether an application for relief is granted or denied," Mattingly said. "No court has held that the use of VTC by itself is a due process violation."

She added: "Regarding case outcomes, no study has demonstrated a statistically significant difference in outcomes at agencies which use VTC for hearings."

Mattingly pointed to two research studies on video-teleconference hearings, one focused on the Social Security Administration and the other a 2015 law review article about EOIR's system. Researcher Ingrid V. Eagly found "there was no statistically significant evidence that judges adjudicated deportation cases more harshly over a video screen."

"Televideo cases were more likely to result in deportation, yet there was no statistically significant evidence that judges adjudicated deportation cases more harshly over a video screen," Eagly also wrote. "Instead, when compared with similar detained in-person cases, detained televideo cases exhibited depressed engagement with the adversarial process. Televideo litigants were less likely to retain counsel, pursue an application for permission to remain lawfully in the United States (known as relief), or seek the right to return voluntarily (known as voluntary departure)."

Mattingly did not address whether anything had changed in the two years since the GAO report.

We were on the fence between Two and Three Pinocchios. Taken individually, each of the four claims we analyzed is missing context or telling only part of the story, which tips the scale toward Two Pinocchios. The effect is more pernicious when considering all four claims together, which is an argument for Three Pinocchios.

The Trump administration's "myth vs. facts" document says it's wrong to claim a wide disparity in outcomes depending on which immigration court hears a case. But the data shows there is such a discrepancy. Compare Houston with San Francisco, Atlanta with Chicago, Las Vegas with Boston or El Paso with New York.

We don't see how the grant rate for asylum-seekers with lawyers is "essentially the same" as the national average, when the data overwhelmingly shows a higher success rate for immigrants with lawyers compared with those who lack legal representation.

The other two claims are more defensible. It seems reasonable to focus on non-detained cases rather than asylum cases when discussing in absentia orders, since the asylum metric includes people in custody who have little choice but to attend their hearings.

As for video-teleconference hearings, we note that the GAO report said "EOIR has not adopted the best practice of ensuring that its VTC program is outcome-neutral because it has not evaluated what, if any, effects VTC has on case outcomes." But on the other hand, a comprehensive and independent study found "there was no statistically significant evidence that judges adjudicated deportation cases more harshly over a video screen."

The overall effect is more "half the story" rather than "mostly false," so Two Pinocchios are in order.

Two Pinocchios