On September 11, 2026, DHS's notice of proposed rulemaking states that the department proposes to remove the rule at 8 CFR 214.1(l)(2) and restore its previous, long-standing policy of not providing a grace period to nonimmigrant workers once their job ends. The current Code of Federal Regulations text states that the regulation being removed lets a worker in the E-1, E-2, E-3, H-1B, H-1B1, L-1, O-1 or TN classifications, and their dependents, keep status for up to 60 consecutive days after a job ends, or until their authorized stay expires, whichever is shorter, once per validity period, entirely at DHS's discretion. Comments on the proposal are open through November 10, 2026. Read on its own terms, this is a narrow technical correction: a discretionary grace period DHS always had the power to shorten, being shortened to nothing.
It is worth slowing down on how small the department says that correction is. DHS's own cost-benefit analysis for the rule states that a subset of 3,795 nonimmigrant visa beneficiaries may incur lost income if they need additional time to conduct a job search after departing the United States, and the analysis puts a number on no other cost or benefit. Read quickly, that is the entire quantified toll of ending the grace period: fewer than four thousand people, and only their job-search income after they have already left the country.
Just 3,795 people is the only cost DHS put a number on.
Strip out who actually stands in that line and the smallness stops holding up. USCIS's FY2024 H-1B characteristics report to Congress states that 71 percent of all H-1B petitions approved in 2024 were for workers born in India, well ahead of roughly 12 percent for China, the next largest source. DHS's 3,795 estimate covers only the beneficiaries, across every visa class the rule touches, whom the department expects to still be job-hunting after they have left the country. It puts no number on the far larger group the grace period actually exists for: workers who lose a job while still inside the United States and need time to line up a new employer before their status runs out, a group where seven in ten H-1B holders are Indian. DHS's own data on how often that happens make the gap concrete. From FY2021 through FY2025, DHS's own cost-benefit analysis states that an average of 65,752 workers a year across these same visa classes lost a job or changed employer, with a high of 80,034 in FY2023, and 99 percent of that population were H-1B holders, a group more than seventeen times the size of the 3,795 the rule counts as its cost.
What the proposal would change
DHS's plan replaces a 60-day cushion with none at all.
| Current rule (8 CFR 214.1(l)(2)) | DHS's proposed rule | |
|---|---|---|
| Grace period after a job ends | Up to 60 consecutive days, or until authorized stay expires, whichever is shorter, once per validity period, at DHS's discretion | None; restores DHS's pre-2016 policy of not providing one |
| Classes covered | E-1, E-2, E-3, H-1B, H-1B1, L-1, O-1, TN and dependents | Same classes |
| Rule created or proposed | November 2016 | September 11, 2026; comments due November 10, 2026 |
Source: eCFR, 8 CFR 214.1(l)(2); DHS's proposed rule, Federal Register; DHS's 2016 final rule.
The protection built for exactly this moment
DHS's 2016 final rule states that the 60-day grace period was created specifically to increase job portability and otherwise provide stability and flexibility for high-skilled nonimmigrant workers, H-1B holders among them, once their employment ended. The mechanism it protects is not a rare event. USCIS's FY2023 H-1B characteristics report to Congress states that continuing-employment H-1B petitions approved in 2023 split into extensions (23 percent), amendments (27 percent) and sequential employment, meaning a change of employer (20 percent). One in five approved renewals in 2023 was already a worker moving to a new sponsor, the exact move a laid-off worker must complete inside whatever grace period remains.
That population rarely clears the bar the grace period was built to test. Of the 328,758 workers across FY2021-FY2025 who lost a job or changed employers, DHS's own breakdown states that just 5.77 percent had a new employer's petition filed on their behalf within the 60-day window, another 0.59 percent adjusted status, 1.70 percent filed to extend or change status, and under 0.1 percent obtained a compelling-circumstances work permit, leaving roughly nine in ten with none of those protections completed in time. DHS's own proposed rule states that anyone who takes none of those steps "would need to depart the United States unless otherwise authorized to lawfully remain in the United States," a consequence that already applies today, before the grace period that gives them the chance is gone.
No easy fallback
Losing that window would matter less if a laid-off worker had another way to buy time. Since September 21, 2025, a presidential proclamation published in the Federal Register requires that new H-1B petitions filed for workers outside the United States generally be accompanied by a $100,000 payment, a fresh cost stacked on top of any employer switch that already has to happen fast. A pending green card is not the safety net it might sound like either. A Congressional Research Service report finds that the US employment-based green-card backlog, dominated by Indian nationals, held almost 1 million approved workers and family members still waiting for a visa number as of 2020, the most recent count in that report, with wait times for Indian applicants running into decades. More recent data do not improve that picture. A more recent Congressional Research Service report states that 758,250 employment-based petitions were still approved and waiting on a visa number as of June 2024, and the same report shows that, under the November 2024 Visa Bulletin, Indian nationals who filed EB-2 petitions in July 2012 were only then becoming eligible to proceed, a wait of roughly twelve years for that cohort alone. A worker in that queue cannot speed up their own green card to bridge a grace period; the wait is set by the queue, not by urgency.
The honest objection
The strongest case for DHS's proposal is that it changes nothing about the department's legal authority, only its use of it. The current regulation itself, the eCFR text states, says DHS may eliminate or shorten the 60-day period as a matter of discretion, so nobody has a vested right to it, and DHS's own cost-benefit analysis found only 3,795 people whose lost income it could put a number on. On that reading, critics are objecting to the department using a power it always had, over a benefit that was never guaranteed, at a monetized cost the agency's own analysis calls small.
That case is real, but it answers a narrower question than the one that matters. DHS's estimate only counts people it expects to keep job-hunting after they have already left the United States; it does not attempt to count workers who are still in the country, still employed by their current sponsor for now, who would have zero days to find and onboard a new one once that job ends. Sequential employment already accounted for one in five approved H-1B renewals in 2023, the year USCIS last reported the breakdown, and seven in ten H-1B holders approved in 2024 were born in India. A rule can sit entirely within an agency's discretion and still fall on a population its own cost accounting never tried to measure.
The Signal
DHS is not proposing to deport 3,795 people. It is proposing to remove the only built-in pause a nonimmigrant worker gets between losing a job and losing status, for a workforce where a fifth of approved renewals are already employer switches and seven in ten workers are Indian. The department's cost-benefit analysis is honest about what it measured: lost income for people who have already gone. It was never built to measure the thing the grace period was created for, which is time. Watch the change-of-employer share in USCIS's next characteristics report. If it holds near one in five, workers found a way to move fast enough anyway. If it drops, the grace period was doing more work than the department's own headcount ever showed.
Reporting basis: the proposed rule's text and mechanics are from DHS's notice of proposed rulemaking in the Federal Register; the regulation it would replace is the current eCFR text governing the nonimmigrant grace period. The cost estimate DHS attaches to the rule, and the fact that no other cost or benefit is monetized, come from DHS's own cost-benefit analysis published alongside the proposal. The India share of approved H-1B petitions and the sequential-employment share of continuing petitions are both from USCIS's characteristics reports to Congress, for FY2024 and FY2023 respectively, and are not combined into a single figure. The new petition-fee requirement is per the White House's proclamation as published in the Federal Register. The employment-based green-card backlog is a Congressional Research Service estimate dated 2020, updated with more current figures from a second Congressional Research Service report, current as of June 2024 for the backlog count and November 2024 for the India EB-2 cutoff date. The annual count of workers who lost a job or changed employers, the breakdown of what they did about it, and DHS's statement of the consequence for those who did nothing are all from DHS's own cost-benefit analysis and proposed-rule text, covering fiscal years 2021 through 2025. The grace period's 2016 origin is per DHS's own final rule from that year. No figure in this piece is The Signal's calculation; all are reported as stated by their originating source.
