EB-1 India moved 6 weeks in all of 2025. Then 11 months in a single bulletin. Then backwards 5.5 months. None of it was random. FY 27 expectation
It looks like chaos. It is actually five sentences of statute running like clockwork. I went deep on the mechanics using only official sources (travel.state.gov bulletins, Report of the Visa Office, the INA text itself) and once you see the machine, the jumps and retrogressions become predictable. Sharing the full picture because most bulletin commentary tells you what moved without ever explaining why.
Not a lawyer, not legal advice. FAD values are from official bulletins; the projections are my analysis.
(Diagram attached: the full statutory pipe network with the actual FY26 data stamped on each node.)
TL;DR
- EB-1 India only moves meaningfully when a quarterly surplus exists, and that determination has fired in January bulletins (start of Q2), never October
- 2025 proved the baseline: without surplus, EB-1 India moves roughly zero days per year at current queue density
- China's FAD is the best leading indicator for India. The two dates converge by law when surplus exists and diverge by law when it runs out
- FY2027 setup: October 2026 should partially restore the current retrogression; January 2027 is the high-probability spillover bulletin
1. The machine
Six provisions produce everything you see in the bulletin:
| Provision | What it does |
|---|---|
| INA §201(d) | Worldwide EB level = 140,000 + prior year's unused family-sponsored numbers (the "fall-across"). FY25 came out at 150,037 |
| INA §203(b)(1) | EB-1 gets 28.6% of that (roughly 40K+) plus any numbers not needed by EB-4/EB-5 (the "fall-up") |
| INA §202(a)(2) | The 7% per-country ceiling. Applied to EB-1, each oversubscribed country's base share is about 2,800/year |
| INA §202(a)(5) | The surplus switch (AC21, 2000). If EB demand in a calendar quarter won't use all available numbers, the 7% cap "shall not apply." This single sentence creates all spillover |
| INA §203(e)(1) | Visas issue in strict priority-date order, nationality-blind. Cap off means surplus flows to the oldest unreached dates worldwide |
| INA §202(e) | The snap-back. When numbers stop going unused, oversubscribed countries get pro-rated back to their share. The heaviest consumer gets retrogressed |
DOS controls the pacing (9 FAM 503.4), but the direction of every flow is mandatory. DOS cannot route surplus anywhere except the oldest dates, and cannot spare a country from proration once the limit binds. One more input matters right now: the 2025 proclamations (PP 10949, PP 10998, plus IV processing restrictions) suppress issuance in two places at once. Suppressed family-sponsored usage enlarges next year's EB pool via the fall-across, and suppressed rest-of-world EB demand creates the unused numbers that trip the surplus switch.
2. The receipts: four years of FAD data
| Bulletin | EB-1 India FAD | Movement |
|---|---|---|
| through Jan 2023 | Current | (no cutoff) |
| Feb 2023 | 01 Feb 2022 | cutoff imposed |
| Mar to Jul 2023 | 01 Feb 2022 | held |
| Aug 2023 | 01 Jan 2012 | ▼ 10 yrs (crash brake) |
| Sep 2023 | 01 Jan 2012 | held |
| Oct 2023 | 01 Jan 2017 | ▲ 5 yrs (FY24 reset) |
| Nov to Dec 2023 | 01 Jan 2017 | held (Q1 flat) |
| Jan 2024 | 01 Sep 2020 | ▲ 3 yrs 8 mo (Q2 spillover) |
| approx. Mar/Apr 2024 | 01 Mar 2021 | ▲ about 6 mo |
| approx. Jul/Aug 2024 | 01 Feb 2022 | ▲ 11 mo (pre-crash level regained) |
| Sep 2024 to Mar 2025 | 01 Feb 2022 | held |
| approx. Apr 2025 | 15 Feb 2022 | ▲ 14 days |
| May to Nov 2025 | 15 Feb 2022 | THE FREEZE. Total 2025 movement: about 6 weeks |
| Dec 2025 | 15 Mar 2022 | ▲ 1 mo |
| Jan 2026 | 01 Feb 2023 | ▲ about 11 mo (Q2 spillover; converged with China) |
| Feb 2026 | 01 Feb 2023 | held |
| Mar 2026 | 01 Mar 2023 | ▲ 1 mo |
| Apr 2026 | 01 Apr 2023 | ▲ 1 mo (peak) |
| May 2026 | 01 Apr 2023 | held |
| Jun 2026 | 15 Dec 2022 | ▼ 3.5 mo (proration) |
| Jul 2026 | 15 Oct 2022 | ▼ 2 mo (now) |
Pattern: every meaningful advance in four years came in exactly two bulletin types. October resets after a fiscal-year-end crash, and January Q2 spillovers. Everything else was flat or backward.
Proof #1: the 2025 freeze was a control experiment. FY25's worldwide limit dropped to 150,037 as the COVID fall-across washed out, rest-of-world EB-1 demand was strong (the global EB-1A surge), the ~40K pool was fully subscribed, so the §202(a)(5) quarterly test failed all year. Cap never lifted. India got base share only. The January 2025 bulletin came and went with no jump, proving January is a test, not a scheduled release. And a frozen date does not mean zero green cards: India consumed its full allocation all year. A static FAD means newly qualified demand with dates before the cutoff materialized at the same rate as supply. A treadmill. The "first applicant who could not be reached" was the same person for 11 straight bulletins. The lesson: at 2022/2023-era queue density, EB-1 India's organic clearing rate on base share alone is approximately zero days per year. All progress is surplus-dependent.
Proof #2: China caught the mechanism on camera. While India froze in 2025, China crept: small advances of two weeks to a month, all year. That asymmetry is diagnostically perfect. Under §203(e)(1), even scraps of surplus MUST flow to India first (oldest dates), so India frozen = surplus was zero, full stop. China creeping = base share against a thinner queue. Then FY26 showed the other two states. Convergence: the Jan 2026 surplus fired, India jumped ~11 months while China moved 8 days, and both landed on the identical date (01 Feb 2023). The law sends essentially the entire surplus to the trailing country until the dates match. Lockstep: Feb to May 2026, both dates moved as one blended nationality-blind queue to 01 Apr 2023. Divergence: the worldwide limit bound in June, §202(e) fired, India retrogressed 5.5 months across two bulletins while China kept advancing to 01 Jun 2023 on base share alone. Practical read: China's FAD marks where the blended queue has already been processed. When surplus returns, India's convergence target is China's current date, so every month China gains while India sits retrogressed widens the guaranteed catch-up at the next surplus event.
Why the backlog is so dense right there: FY26 was itself a measurement. An unusually large, ban-inflated allocation (India's FY26 use will print well above FY24's 8,809 when Table V publishes) sustainably cleared only Feb 2022 to Oct 2022, about 8 months of priority dates with a full year's supply. DOS reached Apr 2023 transiently, but so much Indian demand materialized at 2023 dates that two retrogressions followed. The late-2022 through mid-2023 cohort is heavily loaded (the EB-1A self-petition boom plus priority-date retention from older I-140s). And note: USCIS's pending I-485 inventory only shows domestic AOS filers. Consular demand is invisible until it materializes at NVC, which is partly why DOS keeps overshooting.
3. FY2027: what to expect and when
The current setup mirrors the FY24 opening (a fiscal-year-end retrogression to unwind), not the FY26 one (a flat year with nothing to restore). The July 2026 bulletin explicitly warns EB-1 India may retrogress further or go Unavailable before Sep 30.
The key insight is that two different spillovers arrive at two different times:
- The pool enlargement (§201(d) fall-across) is automatic on October 1. The proclamations are suppressing family-sponsored issuance in FY26, and those unused FS numbers roll into the FY27 EB level by statute. Watch for the USCIS/DOS FY27 limit estimate around mid-September (the FY24 version was the ~165,000 announcement). A number well above 140K = a big cushion.
- The India-specific spillover (§202(a)(5) cap-off) is a quarterly determination that has fired in January in both recent surplus years (Jan 2024: +3.7 yrs; Jan 2026: +11 mo), with Oct to Dec flat both times while DOS gathers actuals.
Scenario weights (my analysis, not DOS):
- Hybrid year (~55 to 60%): October 2026 partial restoration (likely somewhere in the Dec 2022 to Feb 2023 range, echoing Oct 2023's partial restore), Nov/Dec flat, then a January 2027 surge toward China's date, followed by demand materializing and another spring/summer retrogression. An FY26 rerun at a higher level
- Sustained surplus (~25 to 30%): a bigger fall-across plus continued ban suppression keeps the pool ahead of demand; the January advance holds through FY27
- 2025 redux (~10 to 15%): proclamations lifted or amended, ROW demand rebounds, small fall-across. October's restoration still happens (that is base-share arithmetic, not surplus-dependent) but the date then freezes below the 2023 cohort on 2025 physics
Watch list, in order: (1) the September FY27 annual limit estimate; (2) October's restoration size; (3) whether ROW EB-1 stays "Current" with no warning language in the Oct to Dec bulletins; (4) any amendment or litigation touching the proclamations, the single input that could flip everything back to the freeze; (5) China's FAD trajectory, i.e. the convergence target.
For consular-processing folks: a January surge can be transient (FY26 proved it). The date must be current at NVC scheduling AND at issuance. If your PD sits in the plausible convergence zone, get documentarily complete and keep everything interview-ready before the January bulletin drops. NVC schedules in documentarily-complete order and sends appointment letters roughly 2 to 3 months out, and post-level scheduling backlogs vary wildly (check the official IV Scheduling Status Tool for your post).
So: if your PD is in the 2022 to 2023 zone, which scenario are you planning around, and what does your post's scheduling backlog look like? Genuinely curious whether the consular folks are seeing the same picture as the AOS crowd.
Sources: travel.state.gov monthly visa bulletins FY23 to FY26; Report of the Visa Office 2024, Table V; INA §§201(d), 202(a)(2), 202(a)(5), 202(e), 203(b)(1), 203(e)(1); 9 FAM 503.4; DOS "Operation of the Numerical Control Process"; NVC IV Scheduling Status Tool. Two mid-FY24 rows in the table are approximate to the month; the values are right. Happy to be corrected on any specific bulletin value, just drop the official link.