The sordid tale of Dominique Strauss-Kahn and the Sofitel maid from Guinea has ironically put the United States asylum system on trail in the popular press. On July 11, 2011, the New York Times ran an article on a “shadowy industry dedicated to asylum fraud.” The Times piece and an associated “Room for Debate” supports the general thesis that there is “evidence of widespread fraud” in the asylum system. In the August 1st edition of the New Yorker magazine, Suketu Mehta tells the story of “Caroline,” an asylum seeker from Congo who admits to lying in order to make her already justifiable asylum claim more convincing; Mehta’s story was picked up on July 25th by National Public Radio’s All Things Considered, which connected it explicitly to the DSK story. Both New Yorker and New York Times stories cite numbers from a report by the U.S. Department of Justice’s Executive Office for Immigration Review. The use of this report to support the idea that fraud is widespread displays a general ignorance of the asylum system and commits a major oversight of what the numbers really say about recent trends in asylum claims.
The numbers and the asylum process
The Times piece cites a figure that immigration courts in the U.S. granted (approved) 51% of cases as evidence of the system’s credulity. Setting aside assumptions about what the “true fraud percentage” should be for the moment, this figure makes sense if you think the asylum system is a single process. However, ever since the 1996 Immigration Reform Act, when asylum was tightened in order to prevent fraud, getting asylum has been a two-tiered process, one called Affirmative asylum, and one called Defensive asylum. The Affirmative process is the initial review of applicants’ cases by an Asylum Officer during a single Asylum Interview. Asylum Officers are experts in conditions of the applicant’s country, and are charged with approving cases that are consistent with facts on the ground in applicants’ countries, and to deny cases that lack sufficient evidence or are suspect. Here is the DOJ’s description int he report cited by the Times and New Yorker:
Aliens who file affirmatively with DHS [Department of Homeland Security], but whose requests are not granted, may be placed in removal proceeding and referred to the appropriate immigration court for further review of the case.
In other words, the Affirmative process is designed to identify and approve the easy cases and refer the rest on to Immigration Court (see below). According to the DOJ’s report, the approval rate in the Affirmative system was 61% in 2010, lower in the five years prior.
Defensive applications for asylum are processed in Immigration Court. Immigration Court is the redheaded stepchild of the U.S. justice system, with no stated rules of evidence and dockets that would make a Criminal or Civil Court judge nauseous. Asylum cases in Immigration Court generally proceed over two or three meetings over the course of two years (lately even longer), leaving applicants in legal limbo in the meantime. The Defensive process is designed to be adversarial, with a prosecutor poking holes in the applicants’ cases and applicants – or applicants’ lawyers – defending their cases. According to the DOJ’s report, approval in the Defensive system was 35% in 2010, and it has been at this rate (between 34-37%) since 2006.
The Times article notes that rates of approval are up – but according to DOJ they are only up in the system designed to deal with the easy cases. Rates have held steady, at just over a third, in the system designed for the cases with insufficient documentation – and, speaking to the interest of the recent media coverage, presumably more likely to be fraudulent. Bottom line: The part of the system most likely to handle fraudulent cases rejects two-thirds of cases, and that proportion has not changed substantially in years.
There’s a larger problem, of course, in stating that any percentage of approvals is evidence of fraud: There’s no way of knowing what the “true fraud percentage” really is. The Times piece admits this, but suggests that the rates in the DOJ report are too high, singling out New York City approval rates of 76% in 2010. The New York rate comes from the division of total completions in New York in 2010 by receipts (Table 6 in the DOJ report), and is, first of all, wrong — completions are not grants, and if you look at the chart you’ll see that the same math in other districts where completions outnumbered receipts would result in rates that are greater than 100% — but is more fundamentally misleading in that it assumes that 24% is an underestimate of a true percentage that just can’t be seen in these numbers. The DOJ report was a simple review of approvals and denials, not an experimental study or even an observational study of asylum fraud. There is no way to justify the description that fraud is “widespread” from the report cited in the article. The New Yorker article is a little bit better in its use of numbers, but Mehta’s portrayal on NPR interview that “most of these stories” asylees tell include “a little bit of untruth” is completely unknowable from the evidence he gathered. He hung around one woman, whose story may or may not be like “most” others. Her story is compelling but that does not make it representative.
The story the DOJ numbers do support
Now, how about those numbers actually displayed in the DOJ report that the article cites? Consider the following, directly from the DOJ report:
As shown in Figure 14 below, asylum receipts declined by 42 percent and asylum completions declined by 30 percent from FY 2006 to FY 2010.
Here’s Figure 14, below:
It’s a little fuzzy here, but the story is clear: Since 2007, asylum claims (“receipts”) and completed cases (both approvals and denials) have dropped by a little less than half. Completely ignored by the Times and the New Yorker, fewer and fewer people are claiming asylum, and those that are wait longer to have their cases completed. Why might this be? Has stepped up immigration enforcement resulted in more legitimate cases not getting to the U.S.? Or are the numbers simply a result of the economic downturn? And why the marked difference in receipts and completions in 2010? Hm, there’s a story.
There’s another larger story about asylum’s place in the immigration debate in the DOJ report, though you’d have to contextualize it a bit: In 2010, among completed cases, approved and denied cases together accounted for just under 20,000 cases nationally. For those concerned that asylum provides a “free pass” for undocumented immigrants, consider this: Of an estimated 12 million undocumented immigrants in the U.S., those attempting to get in through asylum in 2010 accounted for approximately 0.17% (that’s 20,000 divided by 12 million), or less than one fifth of one percent.