The Department of Homeland Security wrongly claims that the policy in its new immigration rule did not increase H-1B petition denials while it was in effect during Donald Trump’s first term. At the start of FY 2018, in October 2017, U.S. Citizenship and Immigration Services, a DHS component, ended deference to prior findings of fact for adjudications via a memo, which especially affects applications for “continuing employment” for existing H-1B employees. After the memo was published, the denial rate for H-1B petitions for continuing employment rose 140%, from 5% in FY 2017 (before the memo) to 12% in FY 2018 and FY 2019, according to a National Foundation for American Policy analysis . In FY 2021, after USCIS rescinded the memo in April 2021, the denial rate was only 2%.

“DHS data also show that historical denial rates did not statistically differ during periods when the deference policy was in effect compared to when it was not in effect,” DHS asserts in its Notice of Proposed Rulemaking published on October 14 (see footnote #13). However, it does not cite the actual H-1B denial-rate data in FY 2018 and FY 2019 after Trump officials rescinded deference. The data contradict its statement.

“If a government agency is not considering relevant data, or incorrect data, that goes to the State Farm analysis and is grounds for vacating a rule,” said Jonathan Wasden of Wasden Law in an interview. Motor Vehicle Manufacturers Association of the United States, Inc. v. State Farm Mutual Automobile Insurance Co. , often referred to as State Farm, is a 1983 Supreme Court case that helped define the limits of federal regulatory action. The Supreme Court ruled that the National Highway Traffic Safety Administration violated the Administrative Procedure Act because its action was “arbitrary and capricious; the agency failed to present an adequate basis and explanation.”

Eliminating deference for prior approvals means USCIS will adjudicate and challenge facts it already found true in a previous adjudication. As a result, eliminating deference will make it more challenging for employers to retain existing employees.

While eliminating deference affects several immigration categories, the Trump administration has directed much of its regulatory firepower at H-1B visas. H-1B temporary visas are often the only way for high-skilled foreign nationals to work in the United States long term. Employers must pay the higher of the actual or prevailing wage paid to U.S. professionals with similar experience and qualifications. Companies recruiting at U.S. universities find that international students make up about 75% to 80% of full-time graduate students in AI-related fields, such as computer and information sciences.

The Trump administration is proposing or implementing several policies to restrict employer access to foreign-born talent. These measures include:

  • A Labor Department proposed rule to substantially increase the prevailing wage required for H-1B visa holders and employment-based green card applicants.
  • A DHS proposed rule to enact a $103,265 fee on new H-1B professionals subject to the annual limit.
  • An upcoming DHS rule to impose new restrictions on the H-1B visa category.
  • A proposed rule to eliminate the ability of many spouses of H-1B visa holders to work in the United States.
  • A proposed rule to end the 60-day grace period for H-1B visa holders who lose jobs.
  • A final rule to restrict international students and J-1 visa holders to fixed four-year periods. A federal district judge recently postponed the effective date of the final rule.
  • A proposed rule that requires universities to pay $70,000 for an international student to work on Optional Practical Training, or OPT, and an additional $30,000 to be employed on STEM OPT.

A DHS Immigration Analysis That Does Not Mention Relevant USCIS Data

Discussing its claim that deference policy did not affect denial rates, DHS writes in the Notice of Proposed Rulemaking: “One possible reason for this is that even when deference did not apply, petitioners may have been able to overcome issues raised in an RFE (Request for Evidence) or NOID (Notice of Intent to Deny) by submitting additional evidence or information. DHS believes, however, that the impact of the deference policy on officers’ fact-finding authority is best reflected in RFE rates rather than approval or denial rates.”

In discussing the proposed rule, DHS notes that a deference policy to prior approvals was established in 2004, rescinded in a memo issued on October 23, 2017 and reinstituted in April 2021 during the Biden administration. Biden officials later published a final rule codifying its policy on deference, which the Trump administration seeks to rescind in its proposed rule.

Trump officials engaged in a series of immigration actions that judges ruled unlawful, and that also affected H-1B denial rates. On June 17, 2020, USCIS was compelled to issue a new policy memo and withdraw a February 2018 memo on “Contracts and Itineraries Requirements for H-1B Petitions Involving Third-Party Worksites” after losing a court case and then agreeing to a settlement with the business group ITServe Alliance.

However, DHS does not discuss mitigating factors, such as its illegal actions that affected denial rates, in detail. Rather, it states, “DHS data also show that historical denial rates did not statistically differ during periods when the deference policy was in effect compared to when it was not in effect.”

Request for Evidence rates skyrocketed in the H-1B category after the Trump administration changed deference policy, according to readily available USCIS data. According to USCIS data, the percentage of completions with RFEs for H-1B petitions rose from 21.4% in FY 2017 to 38% in FY 2018 and 40.2% in FY 2019. That’s an increase of 78% to 88% in the percentage of H-1 B petition completions with RFEs in the first two fiscal years after USCIS rescinded its deference policy.

However, rather than cite readily available USCIS data on the percentage of H-1 B petitions with Requests for Evidence, DHS produced an analysis that contradicts its own published data showing significant increases in H-1B RFE and denial rates after the new deference policy began in FY 2018.

DHS states, “As shown in Table 3… the RFE rate increased by 2.9 percent more than the RFE rate for the comparison group after the deference policy was rescinded in 2017.” According to the agency, “DHS applied the same DID (difference-in-differences) approach to denial rates and found no corresponding change.”

DHS does not reference in the Notice of Proposed Rulemaking real-world data published by USCIS that show the percentage of completions with RFEs for H-1B petitions increased from 21.4% in FY 2017—before the deference policy was rescinded—to 38% in FY 2018 and 40.2% in FY 2019, after the deference policy was put in place. It also does not cite USCIS data showing the denial rate for H-1B petitions for continuing employment increased by 140%, from 5% in FY 2017 (before the memo) to 12% in FY 2018 and FY 2019 after the Trump administration ended the deference policy established in 2004.

DHS does not explain why its internal analysis, which shows little to no impact on denial and Request for Evidence rates, differs so sharply from its other published data.

DHS released no information on its analysis beyond a brief explanation and Table 3 to support its claim that Request for Evidence rates increased by only 2.9% “more than the RFE rate for the comparison group.”

Mark Regets, a labor economist and NFAP senior fellow, notes that for a difference-in-differences analysis to be valid, one must believe that, absent the policy change, the two groups would have the same outcome. He said in an interview that this seems unlikely in this case because the two groups DHS compares are normally adjudicated quite differently.

According to DHS, its difference-in-differences analysis “compares petitions affected by the deference policy (affected group) and petitions not affected by the deference policy (comparison group).”

DHS states that the affected group included I-129 petitions that checked boxes b or c, which cover petitions for “continuation of previously approved employment without change with the same employer” and “change in previous approved employment.” DHS said the comparison group included I-129 petitions that checked boxes a, d or e, covering petitions for “new employment,” “new concurrent employment,” and “change of employer.”

DHS roughly divided the two groups into what it elsewhere labeled initial employment and continuing employment—and that suggests a significant flaw in the DHS analysis. The adjudication experience for these two types of petitions differed significantly during the Trump years. In FY 2018, the denial rate for H-1B petitions for initial employment was 24%, twice the 12% denial rate for H-1B petitions for continuing employment. Similarly, in FY 2019, the denial rate for H-1B petitions for initial employment was 21%, which was significantly higher than the 12% denial rate for H-1B petitions for continuing employment.

Another issue with the DHS difference-in-differences analysis: employers can use an I-129 form for 16 different categories and subcategories. This could affect or prevent a proper comparison of the policy change’s impact on H-1B petitions, the category most likely affected by the change in deference policy. DHS did not state which categories it included in its analysis.

“DHS does not show its work sufficiently for the public to evaluate it completely, and it does not explain why its analysis of denial rates and Request for Evidence rates differs so drastically from available H-1B data that it publishes and which are easily accessible,” said Mark Regets.

Attorneys Expect A Negative Impact From The Immigration Rule

Immigration attorneys expect the proposed rule will harm employers and visa holders. “DHS has tried this before, and the results are telling,” said Vic Goel of Goel & Anderson in an interview. He points to the significant increase in the denial rate for H-1B petitions for continuing employment after deference was rescinded at the start of FY 2018. “It is difficult to reconcile those numbers with the assumption that removing deference will have no effect on outcomes, especially now that USCIS has given officers authority to deny petitions without first issuing a Request for Evidence.”

Goel notes that DHS estimates the cost at about $22 per RFE, but the real cost is the delay, which DHS puts at approximately 64 additional days for petitions that receive a Request for Evidence.

“For companies, the best approach is to treat every extension as if it were an initial petition,” he said. “That means fully documenting eligibility rather than relying on the prior approval notice, making sure job duties, wages, and worksite information are consistent with prior filings, and explaining any differences up front. With the possibility of a denial without an RFE, the petition as filed has to stand on its own.”

He believes employers should also build more time into their planning, including filing extensions as early as permitted, generally six months before expiration, considering premium processing for employees in critical roles or with upcoming travel, and taking a fresh look at any prior approvals that may not hold up under closer review. He also thinks it makes sense to have supporting documents ready before a Request for Evidence arrives, since response windows are shorter and USCIS may treat a partial response as a request to decide the case on the existing record.

Goel advises employees that, before traveling internationally while an extension is pending, to talk with their employer or immigration counsel, because leaving the country can affect a pending request and they may need an approved petition to return.

“Employers need predictable immigration decisions to plan their operations and retain talent,” said Kevin Miner of Fragomen in an interview. “If USCIS has already approved a petition and the relevant facts have not changed, employers should be able to rely on that decision when seeking an extension.”

The new policy, combined with other Trump administration immigration policies, is expected to make it harder for businesses to plan. “When an approval can be revisited even though the relevant facts have not changed, companies lose the stability they need to make long-term decisions,” said Miner. “These constant changes to our immigration rules make the United States a less attractive place for large companies to invest. Some may choose to put key projects, and the jobs those projects create, in countries with more stable and predictable immigration policies.”