Friday, July 12, 2019

Age Assessments for Unaccompanied Children: Alternative Facts, Policies, and Logic

Click image to enlarge 

Interim Report on Age Assessment Policy Violations by the U.S. Department of Health and Human Services, Office of Refugee and Resettlement

Jackie Stevens 
Professor, Political Science Department, Founding Director, Deportation Research Clinic, Buffett Institute for Global Affairs, Northwestern University
Research assistance from Political Science Department Farrell Fellow Daisy Grace Conant and Posner Fellow Khadeejah Milhan

This is a preliminary report on findings about ORR Age Assessments.  It is being released in conjunction with producer Nadia Reiman's "This American Life" July 12 broadcast as an interim alert to the public, including those in custody, and attorneys, as we await the full production from ORR and ICE.

U.S. officials are unlawfully moving asylum-seeking children from costly shelters into relatively cheaper Immigration and Customs Enforcement jails.  The Deportation Research Clinic has reviewed hundreds of age assessment cases since January 1, 2016.  Among the 205 formal Memorandums of Age Determination coded to date we found ORR officials blatantly misstating the results of medical assessments and official policies.

Along with keeping adults out of facilities meant for minors, officials from the Office of Refugee and Resettlement are placing unaccompanied children into ICE jails.  We found ORR collaborating with Immigration and Customs Enforcement officials, violating the policies for placement, and also writing bogus reports that could hurt children's asylum claims.
Journalists previously have highlighted ORR officials turning children over to ICE if, per ORR policy, a dentist says there is at least a 75% likelihood that their wisdom teeth, roots and all, look like those of someone 18 years or older.  If you're reading this a second time and furrowing your brow, you're not alone: the expert consensus is that dental radiographs cannot be used to reliably distinguish the teeth of a 17 year-old from those of a 23 year-old, much less an 18 year-old.  And if the teeth are those of someone from the vast majority of the world's population that has no reference sample for a comparison, forget about it.

U.S. policy says the radiograph reports cannot be used without other evidence of adult status, such as a birth certificate.  In at least three separate federal lawsuits since 2016, judges have found ORR violating this policy.  

It's bad enough that the government is using tooth X-rays for immigrant age assessments.  But we found something even more alarming and not previously reported: the government ordering children moved to ICE custody even though they had submitted their own bona fide birth certificates and the dental radiographs affirmed their claims or were never referenced.

The L.A. Times article recently lamented cases in which people had their ages reassessed because they passsed the 75% threshold. Our cases show ages being reassessed upward even when the radiograph reports show probabilities below this threshold, including indicating someone is not likely to be an adult.
Think Magic Eight Ball.  The government asks the Magic Radiograph, "Is Taj 18 years or older?"  The Magic Radiograph says, "Most likely."   The shelter worker gives Taj the news and tells him ICE is on the way to put him into jail.  Taj says, "What about my Bangladeshi birth certificate?"  "Tough luck," the case worker might tell him.  

But what about when the Magic Radiograph answers the question of whether Taj is at least 18 with "Cannot Predict Now," or even "Very Doubtful"?  That's what happened in about a quarter of the cases we reviewed that used the dental forensic reports.

Taj should have lucked out.  But in the world of alternative facts, he still loses.  The government is taking these results and emphasizing that according to the Magic Radiograph, there's some chance Taj could be 18 years or older.  And then, based only on adulthood simply being possible, the government is disregarding Taj's bona fide documents, and his narrative, and that of his parents, and still telling him ICE is on the way to put him into jail and hold him there until his final hearing. 

Interim Key findings:

(1)  ORR officials are misrepresenting the facts and policies in their age assessment memorandums.  Instead of catching fraud, ORR is perpetrating it.

The ORR shelters regularly encounter people whose narratives, statements on file with the government, official documents, and biometrics, contain discrepancies. There is nothing inherently unlawful or even surprising about this. Children often travel using false documents attesting adulthood so they can leave their country of origin, especially if their journeys require air travel from, say, Bangladesh, India, or Camaroon.  The registries of many developing countries are known to contain errors and missing data.   

Minors benefit from laws allowing them more lenient supervision and release while awaiting hearings. Not surprisingly, some adult noncitizens have entered the United States with their own altered birth certificates or a bona fide birth certificate belonging to a younger friend or relative and used these to enter the juvenile shelter system.   Case managers in shelters are catching some of those, mostly from pressuring people into confessions. 

Case managers also review cases remotely, including employees of a military firm with ORR contract.  They conduct reviews of migrant children digitally and from their homes.  One employee claimed she had elucidated an ORR client was lying after she found a Facebook page that, if it really did belong to the person in their custody, proved the person had graduated from college.  Again, if, it didn't belong to another, say, Juan Mario Gomez from El Salvador.) 

ORR's Age Assessment policy states:
Each agency acknowledges the challenges in determining the age of individuals in custody. These challenges include, but are not limited to: 
•Unavailable  documentation;
• Contradictory or fraudulent identity documentation and/or statements
• Physical appearance of the individual; and
• Diminished capacity of the individual.
The TVPRA requires the age determination procedures, at a minimum, to take into account multiple forms of evidence. Accordingly, under these procedures, each case must be evaluated carefully based on the totality of all available evidence, including the statement of the individual in question.
ORR blatantly disregards these requirements, as documented below.

(2) The payments for these investigations are coming out of ORR's health care budget for these children.
The payments to the dentists performing the X-rays and analysis are coming out of ORR's health care budget for the shelters.  This would be like the Federal Bureau of Investigation paying for forensic hair analyis from our Medicare budget.  It's exactly the sort of shuffling of funds from child care to law enforcement that makes it difficult for the Democrats to support Trump's plans to increase even ORR funding without extensive oversight by the House.  (ORR has funds to pay hundreds of thousands of dollars for junk science reports, but not toothpaste for kids in CBP custody?)  UPDATE: 3/17/2020 - The documents released indicated offsite age assessments were being performed by four dentists, with David Senn and Peter Arvanitis conducting the vast majority.  Arvanitis runs a family dentistry practice in Arizona, was not listed as a member of the American Board of Forensic Ondontology, and did not respond to inquiries about his qualifications for conducting such assessments.
(3) The age assessments rely heavily on data and narratives from the consular officials of the children's country of origin.
The cases are frequently asylum claims.  Relying on the consular officer's disavowal of claims made by those seeking asylum is like discrediting a wife's claim of abuse because an accused husband claims the narrative is not plausible.  Further, the vast majority of the age assessments are for people from Guatemala or Honduras, both  notorious for national registries with inaccuracies, missing information, and even, in the case of Honduras, criminal fraud.  (Widespread problems with the registries and certificates in Latin American countries are described by Emory University Law School Professor Polly Price in Chapter 1 of Citizenship In Question: Evidentiary Birthright and Statelessness.)

(4)  When attorneys file habeas challenges to the relocation of a client to ICE custody based on the dental X-rays, judges convey the common sense indignation one would expect from any well-educated professional.
The judges' orders note that those charged with enforcing the Trafficking Victims Protection Reauthorization Act seem unfamiliar with its most basic instructions.  The orders suggest that the judges are encountering government officials who inhabit a cave of illiterate simpletons.

For instance, in an opinion of October 26, 2018, Judge Diane Humetewa writes:
There is no apparent plausible construction of the TVPRA, or the ORR Guide, under which an ORR official's nonspecific, unsubstantiated speculation of what they perceive to be adult behavior suffices as 'evidence' that may be considered and relied upon in making an age determination.  Indeed, ORR does not include appearance or behavior as criteria for evaluating whether an individual is an adult or juvenile, but instead lists those factors as a challenge to the age determination process.
The judge ridicules not only reliance on the case managers' subjective opinions but also the accuracy of  information in ICE databases.  She writes: "DHS was aware of this conflicting information in Bitmap at the time it designated Petitioner as an UAC" and so inferred "these records appear to be questionable at best."

(5)  Military and security firms are using massive cross-country databases with information that is not only bogus and harmful, in this case to the children from whom it was elicited without consent, but also prone to hacking and other illicit uses.

General Dynamics Information Technology handles juvenile case management services for ORR.  Last year GD acquired the firm that contracts with Homeland Security to operate the biometric database the judge above was rightly questioning.

Everyone knows that the Biometric Identification Transnational Migration Alert Program database cannot distinguish fraudulent passports children use to leave their countries from their own from bona fide documents they present on arriving at the U.S. border.) 

Here is a link to a GDIT juvenile case management job posting, courtesy of The Daily Beast.    (When you read about the difficulty in reuniting children with their parents, blame General Dynamics.) 

General Dynamics is the world's fifth largest military contractor and the country's third largest federal contractor.  Vertical integration is the name of its gruesome game.  Just like Purdue Pharma makes money first from addicting people to opioids and then from selling drugs for overdoses, General Dynamics profits first from selling cluster bomb missiles, and then the federal services for the children made refugees from war.   That's why countries and university faculties are calling for ending contracts with GD.

The MADs - Overview of Results to Date

The Memorandum of Age Determination is the key document ORR uses when coordinating with ICE to remove kids from a juvenile shelter, at a cost of about $750/day, and put them in ICE jail, about $125/day, while awaiting an immigration hearing.

The triggers for the reviews typically are omitted or stated vaguely.  105 MADs state the age review was initiated because it "became apparent" during an intake interview that the person was likely an adult, with no explanation of what exactly was "apparent" that prompted the case worker's belief the client was an adult. 

43 of the reviews were initiated due to discrepant documents, 21 after people admit they lied, and the balance are initiated based on "behavior" or "physical appearance."

Email shows the shelter case workers and ORR often are taking cues for age assessments from ICE officials, even though ORR is legally the guardian for unaccompanied children.  The case managers include people who staff the nonprofit shelters as well as folks who research cases from their homes and work for General Dynamics Information Technology.

The statements by these contract workers in records systems and emails, including those that are not made lawfully, may be used to move noncitizens out of shelters and also to impugn their credibility in asylum cases.  Conflicts of interest are a serious concern.  If the shelter is under scrutiny for self-dealing and overpay, as is Southwest Key, a major contributor to these MADs -- responsible for all but nine of the 205 MADs we reviewed -- could they be selling out those over whom they have legal guardianship by cheerfully providing assessments the contracting agency wants to hear?  If they cooperate, might this make it less likely ORR will suspend SWK from federal contracts, as the record shows should happen?

A typical MAD is between one and two pages and follows a template that narrates facts in the person's file and a list of the transfer documents, including a TB report.  The MAD concludes with an assessment the person is 18 years or older or just a statement that per policy ICE should pick them up.  About 80% include no medical age assessments. 

We are still tabulating, but so far found nine cases in which the dental radiographs put the chances of someone being an adult below 75%, sometimes well below.  ORR official memorandums misstated the outcomes or policy, or both.  Just to be clear, these cases have weak or no other evidence to impute adulthood.

We've extracted a few examples.  Here's one from November 3, 2018, for a teenager who possesses a birth certificate that shows his date of birth is 2001, and a reference to a Consulate report on the birth certificate with no claim the Consulate questioned its validity.  There is also a dental forensics report.  It estimates a 51.4% probability that the person claiming to be 17 is 18 years or older, the equivalent of the Magic Eightball being clueless.   (The Memo does not report the country of origin but other email and notes suggest the person is from Bangladesh.)

There is exactly one observation inconsistent with this person being a minor: "On November 5, 2018, ORR received information that Mr. [REDACTED] entered the US with a date of birth of [REDACTED] 1998, making him 20 years old."  There is no reference to the nature of the information, its source, or how information received on November 5, 2018 could be included in a Memo of Age Redeterminated dated two days earlier.

Compare the above memorandum with one typical of those referencing actual evidence that would hold up in court.  Information included in this one and not in the one above is: an admission of lying, specific evidence of an altered birth certificate, and a summary of a conversation with the Guatemalan consulate.

Another memorandum, also from early November, 2018, but authored by a different ORR official covering the Chicago region, where Yong was held, follows a similar trajectory as the first bogus one -- crediting unsourced information from Bitmap and ignoring bona fide documents and a radiograph that says there is only a 30% likelihood that the wisdom teeth belong to someone who is 18 years or older.

Here's a third example of a MAD gone wrong, by a third official, but in the same time frame and also for a teenage boy from Bangladesh.  The memorandum indicates the Bangladesh birth certificate is authenticated and that the forensic results are below the ORR threshold.  Nonetheless, the memorandum concludes, "In accordance with ORR age redetermination policy and procures, MR. [REDACTED] dental forensics results are 68.45 and ORR requests that FOJC [Field Office Juvenile Coordinator] coordinates and ensure[s] the immediate transfer of custody of this individual to an adult DRO facility." 

The media have made a big deal about the dental radiographs.  Not previously reported is that fewer than 20% of the MADs actually reference any forensic report.  For an unknown number of cases, radiographs appear to have been taken and assessed but not reported.  After we receive all the documents owed us from the FOIA request, we should be able to assess the discrepancies between references to dental X-rays in the shelter records and their absence from the MADs.
The MADs contain numerous other problems unrelated to radiographs.  In one case without a radiograph, the ORR in late 2018 noted it had evidence that a teenager from Guatemala was born in 2001, not 2000, as he claimed.  ORR's memo stated he was 20 years old.  The research assistant who coded this MAD conveyed to me her concern that people in ORR cannot add.

I told her my worry: under the impression no one would ever look at these, officials are filling in the blanks, maybe even copying and pasting, and then asserting whatever is necessary for the desired bottom line: "Please refer this case to DHS ICE Juvenile Coordinator to arrange for transfer to ICE adult custody."

In another case with a 2016 MAD, the U.S. government stated that someone was using travel documents with a 1999 date of birth, making the individual 17 years old.  However, a shelter worker claimed the individual stated he was born in 1997.  The government in this case relied on the older age supposedly stated by the teenager and ignored its documents.

 Cherry picking the oldests age based on the documents available is a pattern throughout the MADs.  In November, 2018 an ORR official in Washington, D.C. writing a report requesting an ICE pick-up from a Phoenix children's shelter writes: "Dental forensic report result indicate adult age over 18: Avg. Mean Age 17.02."  
Among the 205 memorandums going back to January 1, 2016 that we have coded to date, just 38 include forensic results.  Among these, 22 rely on these exclusively to claim the person is 18 years older.  And, about one of out four, are using the age ranges to claim people are adults, even though the probability is below the legal threshold or even states the person is unlikely to be an adult.  

One MAD indicates the radiograph report indicated just a 28% probability that the individual is an adult, meaning a 72% probability the dental radiographs were from a juvenile.  A prior communication on this same case explains the surprising outcome: "[Because] the possible age range includes having reached the age of majority, it is our determination that the original determination that they are adults should stand" (emphasis added).  This suggests the official disregarding not only ORR policy but also the math on which it is based. 

Most dental radiograph from ORR shelters are likely to yield an age range that could include 18 years or much older.  This is because the dental radiograph ranges used for age determinations include two standard deviations above and below the mean.  Indeed, because the dental changes used for tracking wisdom teeth max out in our teens or early twenties -- we're not rodents -- radiograph analyses cannot distinguish someone who is 21 from someone who is 50.  In other words, more than 90% of the population falls within the two deviations of the 17.2 mean age associated with this radiograph. 

Here's the ORR policy on dental radiographs:
If an individual’s estimated probability of being 18 or older is 75 percent or greater according to a medical age assessment, and this evidence has been considered in conjunction with the totality of the evidence, ORR may refer the individual to DHS. The 75 percent probability threshold applies to all medical methods and approaches identified by the medical community as appropriate methods for assessing age.
Clearly ORR is ignoring this threshold. 

We also noted MADs leaving out information ORR requires to be included, especially the narratives of those in custody and their parents, as well as close coordination with DHS, especially for the individuals from Bangladesh.  An ORR official notes that "two dental records that indicate the UAC are minors."   In an earlier email the ORR official writes that the "Embassy said that the mother called and said when she gave birth and she should know since she is the one who birthed him."  This information does not appear in any of the MADs for the teenagers from Bangladesh, in violation of ORR policy.  
More generally, shelter workers claimed individuals confessed to being adults in about half the cases, but in the balance, the clients insisted they were indeed children.

Again, we are still in litigation.  The next status hearing is in late July and we will be asking the judge to order the government to release the balance of responsive documents, including the memorandums currently withheld in violation of the FOIA.

One important question is about the seeming discrepancy between the references to ordering dental forensics -- well over a hundred -- and the 38 memorandums of 205 referencing them.  Is the number referenced low because they were only ordered for these 38 cases, and when we receive more MADs they numbers will change significantly?  Or because an underling decided her boss might not want to see it? 

In an email of December 29, 2016, under the subject heading "re: Issue of Age," an intrepid ORR employee writes, "I'm happy to share the dental if you want it, I didn't send it in yet since it is below 75%." 

While ORR regularly uses the highest age range for kicking children out of ORR, the government also has been known to claim it has a very different policy, one requiring it use the lowest age.  "This American Life" producer Nadia Reiman encountered just such a case.  A Laotian woman arriving at O'Hare airport with a fiance visa was put into a juvenile shelter and held there despite a bona fide passport proving she was 19.  The rationale?  "[P]er ERO's standard operating procedures of age determination... DHS will base her age on the lowest age range from the 9/18/17 report (Dr. Senn's) at 15.35 age and minor's new DOB will be adjusted to September 1, 2002 for their records." 

With the exception of Yong's case, I have not seen a single report in which either ICE or ORR relied on the lowest age range of a dental forensic report.  And of course, if this policy were in place, then the cases that include age ranges under 18 -- the vast majority -- would not be moved to ICE custody and we've seen cases above showing that is happening.  For instance, the report of November, 2018 indicating a mean age of 17.02 states the lowest age for the person is 13.78 years. Yet the request is for immediate removal to ICE custody.

Finally, the stakes for the age assessment reports are not just custody while one waits for an asylum hearing.  A May, 2019 memorandum makes the MADs relevant to the actual asylum review as well:
Documentary evidence establishing the applicant's age and identity will be especially important if the applicant appears to be over the age of 18 at the time of filing, or if there are questions about the applicant's identity. As the party invoking users' jurisdiction, the individual filing for asylum bears the burden to establish that he or she met the UAC definition, which includes the applicant's burden to establish his or her age, and that the applicant was unaccompanied, at the time of first filing the asylum application." 
The USCIS memo misstates the Trafficking Victims Protection Reauthorization Act and ORR policy.  ORR states that challenges to accurate assessment of someone's age include "unavailable documentation" and "physical appearance."  In other words, officers need to make sure that just because someone is missing a birth certificate and may appear 18 years of age, the officials do not assume the person claiming to be a juvenile is lying.

To be continued!  Thanks to Nadia Reiman for her great reporting on Yong Xiong, and Daisy Conant and Khadeejah Milhan for their research assistance. 

No comments:

#End read more