Massive Asylum Fraud Alleged in New York Case

Charges involve more than 180 applications

By Andrew R. Arthur on March 31, 2018

One of the more common questions that I am asked is one that I cannot answer: "How common is asylum fraud?" While I can provide my own experience as an immigration judge, I cannot dispositively provide a percentage of total asylum claims that involve fraud, or even the number of total cases in which fraud has been identified by adjudicators. The first number has never been analyzed, and the second number is not captured by any source of which I am aware.

Political scientist Raymond Wolfinger is credited with coining the phrase: "The plural of anecdote is data." While this is a debatable point, in the absence of a completed study forensically analyzing fraud in asylum applications, individual cases are generally the best guide for answering this question.

On Wednesday, March 28, 2018, the U.S. Attorney's Office for the Southern District of New York issued a press release announcing that Queens immigration attorney Andreea Dumitru had been charged with asylum fraud and making false statements. According to U.S. Attorney Geoffrey S. Berman, Dumitru "is alleged to have submitted fraudulent forms for over 180 clients, each containing statements and representations she knew to be false."

In my April 19, 2017, Backgrounder, "Fraud in the 'Credible Fear' Process", I detailed the difficulties that U.S. Citizenship and Immigration Services (USCIS) and the immigration courts face in identifying fraud in the asylum process. I specifically cited a December 2015 report from the Government Accountability Office (GAO) captioned "Asylum, Additional Actions Needed to Assess and Address Fraud Risks", which had found:

USCIS and the Department of Justice's (DOJ) Executive Office for Immigration Review (EOIR) have limited capabilities to detect asylum fraud. First, while both USCIS and EOIR have mechanisms to investigate fraud in individual applications, neither agency has assessed fraud risks across the asylum process, in accordance with leading practices for managing fraud risks. ... Without regular assessments of fraud risks, USCIS and EOIR lack reasonable assurance that they have implemented controls to mitigate those risks. Second, USCIS's capability to identify patterns of fraud across asylum applications is hindered because USCIS relies on a paper-based system for asylum applications and does not electronically capture some key information that could be used to detect fraud, such as the applicant's written statement. Asylum officers and USCIS Fraud Detection and National Security (FDNS) Directorate immigration officers told GAO that they can identify potential fraud by analyzing trends across asylum applications; however, they must rely on labor-intensive methods to do so. Identifying and implementing additional fraud detection tools could enable USCIS to detect fraud more effectively while using resources more efficiently. Third, FDNS has not established clear fraud detection responsibilities for its immigration officers in asylum offices; FDNS officers we spoke with at all eight asylum offices told GAO they have limited guidance with respect to fraud. FDNS standard operating procedures for fraud detection are intended to apply across USCIS, and therefore do not reflect the unique features of the asylum system. Developing asylum-specific guidance for fraud detection, in accordance with federal internal control standards, would better position FDNS officers to understand their roles and responsibilities in the asylum process.

It should be noted at this point that Dumitru has simply been charged, and that she is presumed innocent until proven guilty. If those charges are true, however, they would represent a victory for FDNS against the odds identified by the GAO.

In that press release, USCIS New York Asylum Office Director Patricia Menges alludes to the role that FDNS played in that investigation, stating that the agency "is proud of the hard work of our fraud detection officers that leads to important cases like this one, and ensures that those who try to commit immigration fraud will face justice," although she also credits U.S. Immigration and Customs Enforcement (ICE) Homeland Security Investigations (HSI) and other law enforcement for their collaboration in the investigation.

The indictment in that case alleges that Dumitru "knowingly made false statements and representations about, among other things, certain of the Asylum-Seeking Clients criminal histories, personal narratives of alleged persecution, and/or locations" on more than 180 asylum applications that her firm had filed.

In addition to the difficulties in detecting fraud identified by GAO, asylum fraud is particularly easy to commit because of the lower standard of evidence required to satisfy an asylum claim. Section 208(b) (1)(B)(ii) of the Immigration and Nationality Act (INA) states that "[t]he testimony of the applicant may be sufficient to sustain the applicant's burden without corroboration." This is a necessary standard because, as a popular immigration axiom states, "persecutors do not provide affidavits."

Unfortunately, this lenient standard is all too easily abused by fraudulent applicants and crooked attorneys. This is problematic because fraud in any given case may have a tendency to color the view of the adjudicator in other cases. In addition, as USCIS and immigration judges wade through bogus claims, there are unable to grant protection in meritorious cases in a timely manner.

More needs to be done to identify fraud in the asylum process. A starting point would be for USCIS to undertake a full assessment of such fraud, like the partially completed asylum-based Benefits Fraud and Compliance Assessment (BFCA) that Louis D. Crocetti, Jr., former associate director of FDNS, described during testimony before the House Committee on the Judiciary's Subcommittee on Immigration and Border Security in February 2014. As I explained in "Fraud in the 'Credible Fear' Process:

The asylum-based BFCA Program was designed "[t]o determine the scope and types of fraud, and the application and utility of existing fraud detection methods" and "[t]o identify weaknesses and vulnerabilities, and propose/undertake corrective action." It consisted of a "random sampling of [239 out of 8,555] pending and completed (approved/referred) [affirmative asylum applications filed] with USCIS between May 1 and October 31, 2005." Of those 239 cases, 29 (or 12 percent) were determined to be fraudulent; 12 of those 29 cases had already been granted. While 72 (or 30 percent) of the cases did not contain any "fraud indicators" (that is, inconsistencies, derogatory, or negative information), 138 (or 58 percent) "exhibited possible indicators of fraud," not counting 27 additional cases (for a total of 69 percent) that had been referred because of fraud indicators for overseas verification requests, which had not been completed.

As Crocetti explained, "due to increasing internal differences of opinion and concerns about methodology," FDNS was only able to complete four of the eight BFCAs that it undertook, not including the asylum-based BFCA described above. USCIS and ICE HSI should undertake a similar analysis of approved asylum applications in order to assess "the scope and types of fraud, and the application and utility of existing fraud detection methods."

Such an assessment would inform the asylum debate with data, instead of just anecdote.