For the world’s most vulnerable, protection in the United States has all but disappeared.
Wait times for asylum seekers at the U.S.-Mexico border that already seemed indefinite now seem impossible. Families struggle to find food and shelter to outlast a pandemic order with no end date.
Those who cross north are sent back to Mexico in a matter of hours — or even put onto planes back to the countries from which they fled — without any opportunity to explain why they came.
In its response to COVID-19, the Trump administration achieved what it long sought, a shutdown of the U.S. asylum system. And with new regulations introduced this summer, the administration has moved to squeeze out any real chance at refuge in case the pandemic order is lifted.
But even before the current president began his campaign against asylum in the United States, people often struggled to win protection — no matter how strong their cases appeared to be.
In its 40-year history, the system has chronically fallen short of its promise of safety.
The Trump administration has used statistics about grant rates to justify closing off access to asylum, saying that those who lose their cases are illegitimate asylum seekers.
The facts show a different story: Thousands of people turned away based not on the merits of their cases, but on the capriciousness of a system so riven with inequity that many outcomes seem little more than arbitrary.
A San Diego Union-Tribune analysis of 10 years of court outcomes uncovered many symptoms of the system’s biases — shortcomings that date to the system’s creation.
Numerous factors can sway a case’s result, calling into question the administration’s assertion that a denial means an asylum seeker was lying.
Where asylum seekers wait for their day in court can mean the difference between protection and deportation.
That “where” depends on two decisions mostly out of asylum seekers’ control — whether they are held in detention and in which part of the country their hearings are scheduled.
It can ultimately influence several other important factors: their chances of finding legal representation, the judge assigned and what legal precedents the judge must follow.
Outcomes also vary by nationality, discrepancies that cannot be fully explained by the human rights violations that vary from country to country.
Mixed into all of this are the tendencies of each judge. Even among judges at the same court, grant and deportation rates vary widely.
Stories of different outcomes for similar cases, even for family members fleeing the same danger, are common.
Not a simple yes or no
When people ask for protection at the border, they enter a maze of bureaucracy that is the U.S. asylum system.
Herding them along are thousands of federal employees and contractors — asylum officers, detention center guards, deportation officers, immigration judges, court interpreters and government attorneys.
The process is an adversarial one, with a goal of determining whether the person is deportable from the United States, not whether that person merits protection.
U.S. asylum law, based on international agreements, protects people who flee persecution based on race, nationality, religion, political opinion or membership in a social group such as the LGBTQ community. Persecution must come directly from the government or from someone whom the government cannot or will not control.
In the decade of cases analyzed by the Union-Tribune, immigration judges granted asylum about 19% of the time.
These findings are based on roughly 146,300 immigration court cases with asylum applications filed that reached initial decisions from fiscal 2009 through 2018, excluding some asylum requests that didn’t originate at the border.
But asylum is not always a simple yes or no.
About a quarter of cases were closed without judges making decisions on the merits of the asylum applications. These closures generally meant that asylum seekers were allowed to stay, at least temporarily, in the United States.
Judges ordered deportations in nearly half of the total cases.
The Executive Office for Immigration Review, the agency within the Department of Justice responsible for immigration courts, did not respond to a request for comment about the various findings of the Union-Tribune’s investigation.
Where they wait
Asylum seekers often have little control over where their cases end up — and it’s not necessarily tied to where they arrived at the border.
Though the Trump administration drastically changed where asylum seekers wait beginning in 2019, during the decade analyzed by the Union-Tribune the federal government had two main options.
If immigration officials decided to keep asylum seekers in custody, they were sent to detention centers around the country depending on bed space.
If released, they went wherever someone was willing to help them — a cousin in New York, a friend in Colorado, or an unknown sponsor linked to an advocacy group.
Where this fateful combination of circumstances takes an asylum seeker can make a big difference.
Based on the 10 years of case data analyzed by the Union-Tribune, a detained asylum seeker in Texas was 9.3 times more likely to be ordered deported than a non-detained asylum seeker in New York.
Nationwide, asylum seekers who remained in custody were ordered deported at a higher rate — in 74% of cases — compared with 44% for those who were never detained. Those who were initially detained and then released were ordered deported in 37% of cases.
Being detained doesn’t necessarily reflect anything about the legitimacy of an asylum seeker’s case or hint at a criminal past. Rather, Immigration and Customs Enforcement, the agency responsible for immigration detention, has longstanding policies to keep many asylum seekers in custody regardless of their circumstances.
In or out of custody, the region where an asylum seeker ends up dictates what legal precedents will be used to decide their cases.
The example most widely cited by attorneys is the debate over the definition of what constitutes a “social group” for asylum purposes. The 5th U.S. Circuit Court of Appeals, which governs Texas, along with Louisiana and Mississippi, has long used a narrow interpretation.
How courts define which categories count as social groups makes a difference for people whose persecution claims are based on their membership in a family, or as part of a broader group such as women fleeing domestic violence in countries that don’t protect them or as young men targeted for gang recruitment, among others.
Records show judges under the 5th Circuit ordered 3 in 4 asylum seekers deported from fiscal 2009 through 2018 — more than any other circuit in the country.
At the other end of the spectrum, judges in the 2nd Circuit, which guides case law in New York, Connecticut and Vermont, ordered fewer than 1 in 3 deported.
Judges in the 9th Circuit, which includes California, ordered just over 2 in 5 asylum seekers deported.
Location also dictates how many immigration attorneys, particularly those willing to work pro bono, are available. It is notoriously difficult for asylum seekers held in rural detention centers to find attorneys.
Asylum seekers who did not have representation were ordered deported in 60% of cases in the Union-Tribune analysis, compared with 42% for those with legal help.
Unlike cases in criminal courts, attorneys are not provided to those who cannot afford them — not even when the asylum seeker is a child.
Where they’re from
Geography plays a role in another way — the Union-Tribune’s analysis revealed disparities in outcomes based on nationality.
Part of that has to do with conditions in the country and whether they create reasons to flee that are clearly defined under asylum law.
But another part may have to do with biases and preconceptions in U.S. culture about that country.
This may help explain why asylum seekers from China are much more successful than those from Somalia.
Out of the 10 nationalities with the most asylum applications filed, those two countries of origin, both with long histories of human rights violations, are near the top of the list for grant rates. China ranks second, and Somalia ranks third.
And yet, the odds of asylum seekers from China being granted asylum were 2.2 times higher than those from Somalia, according to the Union-Tribune analysis.
This disparity might be explained by the outsized focus on China in U.S. media coverage and in federal government. China’s notorious treatment of Uighurs — an ethnic minority — and the country’s violent repression of pro-democracy protests in Hong Kong are just the latest examples of human rights violations in the public eye.
In contrast, the decades-long rampant human rights abuses in Somalia that have created hundreds of thousands of refugees do not often make front-page headlines.
Add in racism and xenophobia toward people from certain regions of the world, as well as potential anti-Muslim bias, said Karen Musalo, director of the Center for Gender & Refugee Studies at UC Hastings College of the Law, and the potential for discrepancy grows.
“The relationship between the United States and the country you’re from is a big factor,” said Jeremy Slack, a University of Texas at El Paso professor and author of the book “Deported to Death.” “Chinese people get asylum right now much, much easier than most other countries because we like to poke China as a human rights abuser.”
China’s affiliation with communism is another likely influencer, especially after Congress, in 1996, made it easier for people fleeing the country because of its one-child policy to claim asylum.
These kinds of systemic biases regarding nationality have been in place since the asylum system was created.
In the early days, immigration officials who processed asylum requests relied on U.S. State Department recommendations for each individual case, guidance that was heavily influenced by U.S. foreign policy — in particular, the country’s war on communism.
Under the Reagan administration, Central Americans fleeing powerful communist leaders were granted asylum far more often than those fleeing right-leaning strongman governments because of the United States’ involvement in proxy wars in their countries.
“There was a refusal to recognize that the governments we were supporting were engaging in human rights violations,” said Lucas Guttentag, who teaches immigration law at Stanford University and Yale University.
This meant that Nicaraguan migrants, who were fleeing the left-wing Sandinistas, were granted asylum at a rate of 26%, according to a report published in the book “Threatened Peoples, Threatened Borders: World Migration Policy” in 1995, compared with Salvadorans at 2.6% and Guatemalans at 1.8%, who were fleeing right-wing regimes.
Guttentag was one of the lead attorneys in a lawsuit calling for an end to systemic discrimination based on U.S. foreign policy.
A 1990 settlement in the case allowed Salvadorans and Guatemalans to have their claims reassessed, and Congress made other changes to try to account for the system’s shortcomings.
But as the Union-Tribune’s data analysis suggests, systemic bias based on country of origin has not disappeared.
“It’s cynical to say this, but it needs to be said, which is even though the refugee definition is supposed to be applied in a neutral way, the same way to all nationalities, that has never been the case in the U.S.,” Musalo said.
The difference a judge makes
Even for nationalities with higher grant rates, family members fleeing the same persecution can be split apart by different results.
There are glaring examples among many Chinese families that sought asylum based on the country’s former one-child policy. On multiple occasions, immigration judges granted asylum to the father who was seeking refuge from forced abortions, but not the mother.
“It is difficult to imagine how a rational system of law could tolerate such inconsistent results,” appellate judges in the 2nd Circuit wrote in changing the outcome for a mother in one such case.
The difference for many of these families came down to which judges decided the cases.
The Union-Tribune found large differences in decisions among judges at the same immigration court, even when taking into account that asylum seekers held in detention facilities tend to be ordered deported at higher rates.
Take, for instance, the three judges in San Diego who heard mostly detained cases over the course of the decade analyzed by the Union-Tribune.
Judge Robert McSeveny had the highest deportation rate and ordered about 81% of asylum seekers before him deported. He also had the lowest grant rate at 13%.
Judge Carmene “Zsa Zsa” DePaolo ranked somewhere in the middle, ordering 41% deported. She granted asylum in about 55% of cases.
Judge Anthony Atenaide ordered about 20% deported and granted about 76% asylum.
These gaps between judges — some well over 60 percentage points — exist in courts across the United States.
“There shouldn’t be that much difference,” said Paul Schmidt, a former immigration judge. “It’s hard to make sense out of the system because there are so many variables superimposed on each other.”
Ashley Tabaddor, president of the National Assn. of Immigration Judges, said that different rates among judges should be expected.
There might be details in cases of two people fleeing the same harm in the same country that lead to different outcomes depending on how good their attorneys are — if they have them — as well as how much the government attorneys push back and what the judges’ own previous courtroom experiences are, she said.
Even in cases where details are exactly the same, Tabaddor said, judges can have different opinions.
“It’s not unusual for people looking at the same set of facts and same set of rules to have differing opinions about how much weight to give evidence and what the conclusion should be. That’s in every court,” Tabaddor said, pointing to differences among U.S. Supreme Court justices.
Judges’ grant rates are also influenced by their work history, according to a study from 2007 titled “Refugee Roulette” by researchers from Temple University and Georgetown University Law Center.
The Union-Tribune analysis corroborated this finding. Judges who previously worked as ICE attorneys — generally arguing in immigration court against asylum seekers and other immigrants requesting to stay in the U.S. — were about 1.4 times more likely to order asylum seekers deported during the decade analyzed.
A little more than half of immigration judges who heard cases analyzed by the Union-Tribune previously worked for ICE.
When they make the career switch, they go from one federal agency to another.
That’s because, like ICE employees, immigration judges work for the executive branch rather than the judicial branch.
The judges’ boss is the attorney general, the nation’s highest-ranking prosecutor in the Department of Justice.
Tabaddor and other leaders of the judges’ union have long argued that immigration courts should be part of the judicial branch instead — a solution that could help reduce the mistrust that many critics have toward the system.
“Why don’t you trust the judge?” Tabaddor said, addressing those critics. “It’s because you know the court is run by a law enforcement agency. You feel like the court is stacked. You feel like there’s something inherently wrong. And on those grounds I say, ‘You’re absolutely right.’”
A family separated
For one asylum seeker from Central Asia, this fateful lottery of circumstances could mean that he is deported while his family stays in the United States.
The Russian-speaking man, Mr. U, declined to fully identify himself, as well as the country that he fled, to protect family members he left behind. But he allowed the Union-Tribune to review his case files.
His experience navigating the system presents a striking example of how an asylum seeker’s physical location can impact other factors in a case — and ultimately the outcome.
Mr. U first entered the asylum system in San Diego.
He was separated from his 13-year-old son, as well as his adult stepson and his stepson’s wife, shortly after the family arrived at the San Ysidro Port of Entry in 2017.
The physical separation meant their immigration court cases were split, as well.
Though their asylum claims were all tied to the stepson’s political activities, their separation would ultimately result in dramatically different asylum outcomes.
Mr. U would spend the remainder of his case locked up at Otay Mesa Detention Center while his son was taken to a facility for unaccompanied migrant children in Chicago.
His adult stepson and the stepson’s wife ended up at a detention facility in Adelanto, a few hours north. Mr. U did not know where they were.
The day before his trial, Mr. U got some of the best news he’d received since his arrival. He finally found a pro bono lawyer willing to take his case.
“I was very happy,” Mr. U told the Union-Tribune through a translator. “I had a new hope to see my son sooner.”
On the day of his trial, he handed the judge a written statement explaining that the attorney, who then worked at Catholic Charities, was unable to be in court that day because another client had a hearing.
The statement added that Mr. U’s adult stepson was also in immigration custody and that their cases were related. Mr. U hoped that his stepson could be a witness in his case.
Mr. U asked to postpone his trial for about a month and a half.
“I do not wish to prolong my case any longer than necessary,” Mr. U said in his statement. “My young child has been taken from me and is in the custody of the government far away. He is alone and without me, and I need to get back to him.”
Immigration Judge Scott Simpson insisted that Mr. U proceed with his trial that day without the attorney and without his stepson as a witness.
“You’ve been detained for over eight months,” Simpson told Mr. U, according to court records. “That’s ample time to find an attorney. So, there’s no good cause to continue any longer.”
On his own, Mr. U, who has a high school education, struggled to explain his story clearly to the judge.
The Russian interpreter also struggled to understand him, frequently interrupting the dialogue between Mr. U and Simpson to get clarifications.
Both Simpson and ICE attorney Guy Grande called out details from his testimony that were slightly different from what a fellow detainee had translated into English for Mr. U’s asylum application.
For instance, Mr. U testified that his wife had loaned money to someone, but his application said, “I lent” the money.
He also testified in court about an incident in which police threatened him that he did not mention in his application. His application did mention other instances of threats.
In the end, Simpson did not find him credible. He ordered Mr. U deported.
Mr. U felt fear flood his body, terrified what might wait for him in his country.
“It was a big shock and a hit for me. I immediately felt bad,” Mr. U said. “My blood pressure spiked, and my heart was hurting because I cannot return home.”
Simpson had the highest deportation rate of judges at Otay Mesa Detention Center, according to the Union-Tribune analysis.
The judge, who previously worked as an ICE attorney, ordered asylum seekers deported in more than 80% of the cases he heard. He granted relief in about 15%.
About a month after Simpson decided Mr. U’s case, the stepson and the stepson’s wife were granted asylum by a different judge. They did not have a lawyer either.
That judge, Ian Simons, had the highest grant rate and one of the lowest deportation rates among judges who heard cases at Adelanto at the time. He granted asylum in more than 30% of cases and ordered asylum seekers deported in over 60%.
Holding out for appeal
Appeals are an increasingly important part of the path toward protection, a reality often reflected in high reversal rates among some judges.
The Board of Immigration Appeals, or BIA, part of the same agency in the Department of Justice that employs immigration judges, is the first step in the process.
In the cases analyzed by the Union-Tribune, the BIA told more than 1 in 5 judges nationwide that their decisions were wrong at least 20% of the time.
Judges with fewer than 25 appealed cases were not included in the analysis.
A reversal rate of 20% is a common measure when looking for red flags with judges’ decisions, according to University of San Diego law professor Shaun Martin.
Four judges had reversal rates of 40% or higher.
“A consistently high reversal rate like that would cause you to look very closely to see if the judge was doing something systemically wrong,” Martin said.
Cases appealed beyond the BIA go to the federal circuit courts of appeal, leaving the immigration court system and entering the traditional legal process in the judicial branch. Data analyzed by the Union-Tribune did not include those court decisions.
Attorneys worry that more and more asylum seekers will have to go to the circuit courts to be granted relief. Under Trump, most recent appointees to the BIA are former immigration judges who had some of the highest asylum denial rates in the nation.
Ten members of the current 23-person board are former immigration judges appointed to their positions during the Trump administration.
All but one had deportation rates of more than 70% in at least one of the courts where they heard cases during the decade analyzed by the Union-Tribune. Eight of them had asylum grant rates below 10%, including two judges who didn’t award a single grant of asylum while at a particular court.
The board includes two other Trump appointees who worked in the Office of Immigration Litigation in the Department of Justice, which argues against asylum seekers who have appealed their cases to federal circuit courts, and one Trump appointee who worked for ICE before the administration hired her as an assistant chief immigration judge and ultimately a board member.
Six board members were appointed under the Obama administration, and four were appointed by earlier administrations.
So much is riding on appeals that attorney Dree Collopy, who wrote a 1,680-page legal guide on asylum, encourages lawyers to make sure that the record created in immigration court is incredibly thorough.
“We honestly can’t depend on immigration courts or the Board of Immigration Appeals to give any kind of due process or meaningful review of asylum cases anymore,” Collopy said.
But navigating the appeals process can be daunting and often takes years. Those who are already detained usually stay in custody for the duration. Many give up before they get that far.
Mr. U is better positioned than many.
Because of the outcry over family separation at the border and a class-action lawsuit in San Diego, he was released from detention and reunited with his family in Chicago. And he has a pair of attorneys — Bardis Vakili with the American Civil Liberties Union of San Diego and Imperial Counties and Luis Gonzalez, the attorney who’d originally agreed to take his case — working on his appeal.
They argued that Simpson denied Mr. U his right to an attorney, as well as his right to present evidence.
The BIA sided with Simpson.
Now Mr. U is waiting to find out if the 9th Circuit will change the outcome in his case.
He’s not allowed to work while he waits, so he has to depend on his stepson, who already has a green card.
He hopes to learn English and one day find a profession that will allow him to be useful to the country he wants to call home.
And he worries about the family members left behind.
Though asylum has become an increasingly polarized topic, building enough political will to make the system more equitable has historically been difficult.
“Asylum was always a political football,” said Ruth Wasem, who specialized in immigration policy at the nonpartisan Congressional Research Service before becoming a professor at the University of Texas at Austin.
The asylum system is part of a worldwide effort that grew out of failures to protect Jewish migrants during the Holocaust. It took the United States decades to fully codify its commitment to help victims of future atrocities.
Immigration court records are collected, tracked and released monthly by the Executive Office for Immigration Review within the Department of Justice. The San Diego Union-Tribune used the June 2019 release in its analysis.
Since case information is entered manually, various columns throughout the more than 50 million rows of data in the various tables contained slight inconsistencies, and the Union-Tribune cleaned these entries when necessary.
The analysis includes any cases with asylum applications that were completed from fiscal 2009 through fiscal 2018. Cases flagged as legal permanent residents, rider cases, cases originating with U.S. Citizenship and Immigration Services, and cases that did not include either a charge of being present without admission or arriving without a valid entry document were excluded from the analysis. Cases with incomplete information on these distinctions remained in the analysis.
Judges who heard less than 50 cases in a particular location and nationalities with less than 100 cases during the 10-year period were excluded to prevent skewed results.
Judge work histories were gathered by the Union-Tribune based on summaries released by EOIR when the judge was hired. When necessary, histories were confirmed or clarified using news clippings, court records, law firm biographies, law school newsletters, and in some cases, contacting individuals directly.
With guidance from statisticians, the Union-Tribune performed various statistical tests, including logistic and multivariate regressions, to determine the significance of findings.
Case data and analyses steps can be found on the Union-Tribune’s GitHub page.
With few exceptions, most of the changes to the system since it was created have made the process more difficult for asylum seekers, particularly for those who come to the border.
Among recent changes is the Trump administration’s “Remain in Mexico” program, which forces many asylum seekers to wait for their cases across the border.
Most recently, the administration published new rules that would narrow longstanding definitions in parts of asylum law and fundamentally shrink what options asylum seekers have if they manage to get their cases into immigration court. A second set of proposed rules would allow the government to bar people from asylum because of the pandemic.
The changes would further erode access to a system that is meant to function as part of an international screening process, one that determines which migrants should be recognized as refugees.
“I see our current moment as one of real crisis,” said Denise Gilman, co-director of the immigration clinic at the University of Texas at Austin School of Law. “Even as limited and paltry as the system was, we’re experiencing a moment of doubt as to whether it will even exist in any meaningful way.”
The Trump administration has argued that because the majority do not win their cases, many people applying for asylum do not have valid reasons to ask for help. Recent outcomes appear even lower than in previous years, partly because the administration changed the way outcome statistics are calculated in official reports.
Larry Gollub, a retired asylum officer, said that low grant rates are not proof that asylum seekers are filing frivolous or fraudulent claims.
“It’s just proof that they can’t meet the high standard for asylum,” Gollub said.
And they are struggling with the obstacles built into the system that make it less likely for them to win.
Morrissey and Schroeder write for the San Diego Union-Tribune. Union-Tribune intern Celina Tebor contributed to this report.
The view from Sacramento
For reporting and exclusive analysis from bureau chief John Myers, get our California Politics newsletter.
You may occasionally receive promotional content from the Los Angeles Times.