This summer, pro-immigrant Representative Gerry Connelly (D-Virginia) wrote to USCIS inquiring about the affirmative asylum backlog. The USCIS response is instructive.

First, USCIS notes that, "The backlog is the result of the mathematical reality that USCIS receives more cases than it can adjudicate given current resources." That much is true. But here's the money shot, which is where I part ways with the Biden Administration's reasoning--

LIFO is a critical tool in controlling non-meritorious or fraudulent applications filed to take advantage of the backlog in order to obtain work authorization. But for LIFO, the backlog would doubtlessly be worse.

The evidence for this supposition is weak, and as I view things, based more on coincidence than causation. While the number of new cases decreased under LIFO, other factors--such as the Trump Administration's Muslim travel ban, "extreme vetting" for visa applicants, and the coronavirus pandemic--better explain the reduction in new asylum filings. Here, we'll examine how LIFO affects the backlog, and why I think the agency is wrong to conclude that the last-in, first-out system helps prevent fraud.

First, let's look at why USCIS thinks LIFO helps reduce the backlog. For those not familiar with LIFO, this is the "last-in, first-out" system whereby USCIS prioritizes new asylum applications over old applications. Contrast this with FIFO, the "first-in, first-out" system that was in effect from December 2014 to January 2018. Under FIFO, cases were interviewed in the order that they were received (first come, first served).

In the letter to Representative Connelly, USCIS states that LIFO was implemented to help reduce the asylum backlog from the 1990s. USCIS states--

During the first 20 years of LIFO, the backlog was reduced from a high of 400,000 to just under 5,000 cases by 2009. By the beginning of FY 2013, just over 4,200 cases were pending adjudication longer than six months.

This paragraph seems a bit disingenuous. In 2009, the backlog stood at "just under 5,000" cases, and in FY 2013, 4,200 cases "were pending adjudication longer than six months." This comparison is apples and oranges. In fact, the backlog by July of 2013 (the earlies date I could find data for) stood at 28,699, which is significantly more than the "4,200 cases" mentioned in the USCIS letter. Why is this important? Because the data shows that the backlog was growing in the years before December 2014, while LIFO was still in effect (in December 2014, just before USCIS implemented FIFO, the backlog had grown to 73,103 cases).

On the other hand, it is true, as the USCIS letter states, that the backlog grew more quickly under FIFO and then slowed down again after LIFO was re-implemented in January 2018. For the 17 months prior to FIFO, the backlog increased by an average of 2,614 cases per months, though by late 2014, the backlog was already growing at an increasing rate (about 3,500 cases per month). Throughout the FIFO period, the backlog increased by an average of about 6,264 cases per month. The question is, Is this a trend that was happening anyway or did FIFO cause the increase?

We can start to address this question by examining what happened after LIFO was re-implemented in January 2018. Unfortunately, data for this period is difficult to come by. According to the USCIS letter--

After the re-imposition of LIFO, the number of affirmative asylum applications filed per year decreased from 141,695 in FY 2017 to 106,147 in FY 2018 (-25%); to 95,959 in FY 2019 (-10%); and to 94,077 in FY 2020 (-2%). Since re-instituting LIFO, the annual growth of the backlog has dropped dramatically to only a 10% growth in FY 2018, a 7% growth in FY 2019, and a 13% growth in FY 2020.

Again, this claim is a bit dubious. While the backlog is increasing at a lower proportional rate, in absolute terms, the backlog is still increasing quickly, even under the "new" (post-January 2018) LIFO. For example, on July 31, 2020, the affirmative asylum backlog stood at 370,948 cases. Two months later, on September 30, 2020, the backlog had increased to 386,014 cases. That's an increase of 7,533 cases per month, which is 20% greater than the average monthly increase under FIFO (while the pandemic made it harder to interview cases, it also reduced the number of cases being filed since fewer asylum seekers could reach the U.S., and so it is not clear how badly the pandemic affected the backlog).

In addition, there exists an alternative explanation for the decreasing numbers of new asylum applicants post-January 2018: The Trump Administration had made it more difficult for foreign nationals to come to the U.S. If fewer people are coming here, fewer people will seek asylum. And so while there is some coincidence between LIFO and reduced asylum filings, there are simply too many factors to credit LIFO with achieving that result.

Aside from the data, the idea that LIFO reduces people's incentive to file for asylum does not really make sense. For one thing, even though the government implemented LIFO in 2018, most new cases did not receive interviews. Even those cases that received LIFO interviews and were then referred to court (presumably because they are not meritorious) ended up pending in court for many months or years. So if--as the USCIS letter claims--many people file fake asylum cases in order to get a work permit. LIFO did almost nothing to reduce that incentive.

In sum, I think the jury is still out with regards to the effect of LIFO on the backlog. There are simply too many variables to parse out whether LIFO made a difference. Maybe if USCIS releases more data, we will begin to get a clearer picture of the policy's impact on new applications, but so far, we do not have enough information to reach a firm conclusion.

One last point. There is an assumption built in to USCIS's logic with which I disagree. The agency seems to believe that a significant percentage of asylum applications are fake. In the title to this article, I refer to such logic as "Trumpian," but in fact, this logic long pre-dates Mr. Trump. While many applicants do not qualify for asylum for technical reasons (no nexus, for example), it is rare for the U.S. government to find that an asylum case is actually fake (I have done hundreds of cases and only seen such a finding two or three times). This does not mean that fraud is not a problem, but there is simply no evidence that a large portion of asylum cases are fraudulent, and so USCIS should stop using this as a justification for its policies.

Personally, I think LIFO is incredibly unfair. My clients--and most asylum seekers--do not want delay. They want a decision. Placing new cases ahead of old cases, and then leaving old cases in eternal limbo, has hurt many people and damaged the integrity of our asylum system. Given all the negative externalities, USCIS should more comprehensively review the effects of LIFO to determine whether it is accomplishing what they think it is accomplishing. If not, we should return to FIFO, which was more fair and gave people more predictability and more hope.

Originally posted on the Asylumist: