The Washington PostDemocracy Dies in Darkness

Fact-checking the Trump administration’s immigration fact sheet

A person looks through the border wall toward the United States in San Diego. (Mike Blake/Reuters)

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 “myths vs. facts” document from the Executive Office for Immigration Review (EOIR), the arm of the Justice Department that runs the U.S. immigration court system.

The Facts

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

In some cases, it seems more as though 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 eleven percent (11%). Eighty percent (80%) of immigration courts, 50 out of 62, have a grant rate of twenty-one percent (21%) or lower. Only 12 out of 62 courts have grant rates more than ten percent (10%) above the median grant rate. Only 1 out of 62 courts has a grant rate above fifty percent (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 percent) and the highest (58.5 percent) were still far apart from their peers in Los Angeles. In that city, the 34 immigration judges ranged from a denial rate of 29.4 percent to 97.5 percent.

Those are two of the biggest immigration courts by case load, but these discrepancies also pop up in other parts of the country, as the EOIR chart below shows.

The median grant rate for asylum claims was 11 percent in fiscal 2018. Some localities were lower (Atlanta, El Paso, Las Vegas). The cities represented by red bars (Baltimore, Boston, Chicago, New York) were much higher than the median. Honolulu was above 60 percent — five times the median rate.

“Eighty percent of immigration courts are clustered within 10 percent 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 (44%) of all non-detained removal cases end with an in absentia order of removal due to 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.

At first glance, this seems like a case of cherry-picking data. The EOIR document focuses on in absentia orders for 44 percent 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 percent of all cases in fiscal 2013, a rate that grew to 28 percent in fiscal 2017. Looking at asylum cases alone, the rate was 6 percent in 2013 and 11 percent 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, though she did not break down what share of the total they represent. Mattingly did not address the “all cases” category in her response.

“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 Trump’s repeated claim that immigrants “never show up” for their court hearings.

There's little evidence supporting Sessions' claim that attorneys are helping illegal immigrants scam the asylum system. (Video: Meg Kelly/The Washington Post)

After we published this fact check, a policy analyst at the American Immigration Council, Aaron Reichlin-Melnick, pointed out another issue with this EOIR claim.

The “in absentia” rates are not a fully accurate measure of immigrants who fail to attend a scheduled hearing. Instead, these rates track “initial case completions” in a given year.

In other words, some immigrants are appearing in court but — since there are more than 850,000 backlogged cases — their hearings are rescheduled. In that scenario, their court appearances would not be factored into the in absentia rates.

“An example I often give to illustrate this is to imagine 10 people are scheduled for court on a given day,” Reichlin-Melnick said. "Of that 10, one person fails to appear, and his case is ‘completed’ when the judge orders him removed ‘in absentia.’ The other nine then appear in front of the judge, who ‘completes’ one other case and reschedules the remaining eight people for another hearing.

“For that day, despite just 10 percent of people missing court, 50 percent of ‘completions’ were for missing court. ... But when EOIR used this measurement in their ‘Myths vs. Facts’ document, they do not make this clear, which falsely implies that almost half of non-detained immigrants skipped out on court. When you look the government’s data more precisely, you see that the vast majority of immigrants appear for their court hearings.”

MYTH: Most aliens with representation are granted asylum in immigration proceedings.
FACT: The asylum grant rate for cases with representation is approximately twenty-one percent (21%). The asylum denial rate for cases with representation is approximately forty-eight percent (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 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,” Reichlin-Melnick wrote in a Twitter thread analyzing the “myths vs. facts” 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 percent of cases, yet for those with lawyers, the rate was a fraction of that — 18 percent.

“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: VTC [video-teleconference technology] is unreliable, and its 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 one percent (.0052%) of EOIR VTC hearings, 310 out of nearly 60,000, are continued due to a VTC malfunction.

In support of this claim about holding hearings remotely via video-teleconference, Justice Department officials tabulated data for cases postponed due to technical difficulties (“less than one-tenth of one percent”), though we note that the time period for this statistic is not included in the “myths vs. facts” document.

Is that the right metric to assess reliability and fairness?

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.”

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

“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).”

The Pinocchio Test

We were on the fence between Two and Three Pinocchios. Taken individually, each of the four claims we analyzed relies on a dubious read of the data or tells 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 “myths 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. However, EOIR should include a breakdown of detained asylum seekers in its reports if officials believe the in absentia statistic for asylum cases by itself is misleading.

As for video-teleconference hearings, we note that the GAO report said 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.” 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

(About our rating scale)

Send us facts to check by filling out this form

Sign up for The Fact Checker weekly newsletter

The Fact Checker is a verified signatory to the International Fact-Checking Network code of principles