Also in Türkçe · Español

Two independent analyses published this month put numbers on something self-petitioners have been feeling for a year. The category is harder, it is far more crowded, and for Indian and Chinese nationals the line behind an approval has become the whole question.

What the two studies actually are

Two documents landed this month, eight days apart, and neither is a government announcement. On August 18, 2026 the American Immigration Council published an analysis of USCIS quarterly data spanning FY2016 through FY2025. On August 26, 2026 the National Foundation for American Policy published a policy brief on employment-based waiting times. Both are third-party analyses of government data. No agency issued a new standard or a new regulation here; outside researchers counted, and the count is unflattering.

The distinction matters: no officer is bound by either document. What they give you is the environment your petition is filed into, and a self-petitioner still planning around 2022 is planning around a world that no longer exists.

The denial numbers, in order

Start with volume. Pending I-140 petitions more than quadrupled over the period the Council examined: 43,509 in FY2016 Q1, 180,439 in FY2025 Q4.

Then the denial rate. The overall I-140 denial rate rose from 6.6% to a peak of 16.5% in FY2025 Q3. That is the aggregate across employment-based categories, and aggregates hide the ones that moved. Underneath it: by FY2025 Q4, nearly two-thirds of EB-2 NIW petitions and nearly half of EB-1A petitions were denied. Denial rates in both categories roughly doubled during FY2025 alone — not across the decade, across four quarters.

The pattern does not stop at the petition stage: I-485 denials averaged 11.8% during the first three quarters of the second Trump administration, against an 8.0% decade average.

What the data does not say

This is the paragraph most coverage will skip, and it should govern how you read the rest. A rising denial rate is not the same thing as a rising denial rate for well-prepared petitions. The published data does not break out petition quality: it does not separate a documented ten-year research record from a filing assembled off a template in three weeks. It reports the composite of everything submitted, and that composition has itself changed.

So the honest statement is narrower than the headline. These numbers do not establish that strong cases are now being denied, and they do not establish that strong cases remain safe. The data does not support the inference either way. Anyone claiming it proves adjudicators turned against good petitions is going past the evidence; so is anyone claiming it merely reflects weak filings.

For the same reason, nothing here is a promise. This firm does not promise an approval rate and does not publish its own outcome figures. Past adjudication statistics describe a population, not an individual case.

A category that got much more crowded

NIW filings nearly tripled between FY2022 and FY2025. That growth was large enough to reorder the queue: EB-2 passed EB-3 and now accounts for 46.9% of the pending I-140 backlog. Nearly half of everything waiting sits in one preference category.

Crowding does two things at once. It lengthens the wait behind any approval — the backlog problem below. And it changes what an ordinary file looks like: when a category triples in volume, a record that read as distinctive against the 2022 field is now read against a much larger, better-coached one. That is an observation about the pool, not a claim about how officers decide.

For Indian and Chinese nationals, the backlog is the whole question

The NFAP brief is where strategy actually turns. Its projections — again, from a private research organization, not a government office — are these. An Indian national with a PERM or I-140 filed January 2026 or later faces a projected 179-year wait in EB-2, 38 years in EB-3, and 4–5 years in EB-1. For China the projections are 25 years (EB-2), 7 (EB-3), 5 (EB-1). For the Philippines, 6 years in EB-3.

These are projections built on current queue sizes and annual limits, not entitlements and not published schedules. Even discounted heavily, the gap between 179 years and 4–5 years is not one of degree: it is the difference between a category that can deliver a green card within a working life and one that, on these assumptions, cannot.

The queue itself keeps growing. The combined EB-1, EB-2 and EB-3 backlog went from 1,048,342 in April 2020 to 1,264,495 in December 2025 — an increase of 216,153, or 20.6%. Indians are 79% of that queue, an estimated 996,599 people. The Philippine EB-3 backlog grew 91.7%, from 20,557 to 39,408. And upstream of all of it, for anyone who still needs labor certification, DOL listed 403 days average PERM analyst review and 290 days for audit review as of August 2026 — before the I-140 is even filed.

The argument for building an EB-1 record is not that EB-1A is easier. It plainly is not — nearly half of those petitions were denied too. It is that for an Indian or Chinese national, an EB-2 approval buys a place in a line measured in decades.

What the operative Visa Bulletin says today

The September 2026 Visa Bulletin (Number 18, Volume XI) was issued August 10, 2026. It is not this week’s news; it is the bulletin in force as this is published. On the employment-based Final Action Dates, EB-2 India is Unavailable and EB-5 Unreserved India is Unavailable. EB-2 China holds at 01SEP21, while EB-2 worldwide, Mexico and the Philippines remain Current.

In EB-1, worldwide, Mexico and the Philippines are Current; China is at 01JUL23 and India at 15OCT22. In EB-3, worldwide and Mexico are at 01SEP24, China at 01JAN22, India at 01JAN14 and the Philippines at 01AUG23.

The bulletin also carries three explicit warnings: EB-1 India may become unavailable if India’s pro-rated limit is reached before the end of the fiscal year; EB-2 may retrogress or become unavailable; and EB-5 Unreserved may retrogress or become unavailable. The FY2026 limits behind all of this are 226,000 family-sponsored, 186,317 employment-based, and a 28,862 per-country ceiling. Separately, DV-2026 entitlement ends September 30, 2026.

Who is NOT affected

An approved petition is approved. These are adjudication statistics for a closed period; neither study reopens or revisits a petition USCIS has already granted. If you hold an approved I-140, the numbers describe the line behind you, not your approval.

If your category and country of chargeability are Current, the backlog arithmetic is not yours. EB-2 worldwide, Mexico and the Philippines are Current in the September bulletin, as are EB-1 worldwide, Mexico and the Philippines. A Turkish, Colombian or Spanish national in EB-2 reads the NFAP projections as background, not as their own wait. The retrogression warnings still apply to everyone.

Consular processing is a separate track from the I-485 figure. The 11.8% average against an 8.0% decade average is an adjustment-of-status statistic. If you will finish at a consulate rather than filing an I-485 here, that number is not describing your step. The I-140 figures apply either way.

And the PERM timelines do not touch a self-petitioner. The 403-day and 290-day DOL figures are labor certification numbers; a national interest waiver or an EB-1A self-petition does not go through PERM — a genuine structural advantage of the self-petition route, even in a harder year.

What this means for how you file

Choose the category before you build the record, not after. The two ask for different things: EB-1A asks you to document a record of extraordinary ability, while the national interest waiver asks the government to waive the job offer and labor certification requirement in your case. Evidence built for one and relabeled as the other is visible, and in a crowded field it reads as generic.

For an Indian or Chinese national the sequencing question is the strategic one. If the record is close to an EB-1 showing, the case for investing another year in publications, citations, judging, press coverage or a documented leading role is far stronger than three years ago — because the alternative is not a slower approval, it is a place in the queue the projections describe. That is a judgment call about a specific record, worth making early rather than after a denial.

Assume your 2022 exemplar is stale. A friend’s approved petition, a template circulating in a lab group, recommendation letters written to a pattern — each succeeded against a smaller field, before NIW filings nearly tripled and before denial rates roughly doubled in a single fiscal year. Independent corroboration, specificity about impact, and evidence that does not originate with you or your co-authors are what distinguish a file that survives scrutiny.

The pipeline, further upstream

One more dataset belongs here, because it describes who files these petitions a decade from now. The Common App End-of-Season Report, published August 20, 2026, reports international applicants fell 10% in the 2025–26 cycle — a second consecutive annual decline. By region, Asia was down 11% and Africa down 17%. By country, India fell 15%, Ghana 34%, Ethiopia 29% and Nigeria 21%. Of the ten highest-volume sending countries, only Kenya grew, at +29%, while the Americas rose 5%. The platform itself grew overall, up 2% in first-year applicants.

This has no bearing on any petition being adjudicated now. It is context for a longer horizon: the countries facing the deepest backlogs are also sending fewer undergraduate applicants. Whether those facts are connected is not something this data can answer.

Key dates

What you should do now

• Read both studies as analysis, not as rules. The Council and NFAP documents are third-party work on government data. They describe the environment; they do not change any legal standard, and no officer is bound by them.
• Pick the category first. Decide between an EB-1A record and a national interest waiver before you commission letters or assemble exhibits. Evidence built for one category and relabeled for the other is the file that reads as generic in a crowded field.
• If you are Indian or Chinese, price the queue into the decision. On NFAP’s projections a January 2026 or later filing means 179 years in EB-2 for India and 25 years for China, against 4–5 and 5 years respectively in EB-1. That gap, not relative difficulty, is the strongest argument for building an EB-1 record.
• Retire the 2022 template. NIW filings nearly tripled between FY2022 and FY2025 and EB-2 is now 46.9% of the pending I-140 backlog. A record that looked distinctive against the old field is being read against a much larger one.
• Track the bulletin’s three warnings. EB-1 India may become unavailable if the pro-rated limit is hit before the fiscal year ends; EB-2 may retrogress or become unavailable; EB-5 Unreserved may retrogress or become unavailable. Plan filings so a date movement does not strand you.
• If you still need labor certification, start earlier than feels necessary. DOL listed 403 days average PERM analyst review and 290 days for audit review as of August 2026, all of it before the I-140 is filed.
• Get a candid read on the record you actually have. The useful question is not whether the category is harder — it is — but whether your documented record supports the category you are about to file in. That is answerable before you file and expensive to discover afterward.

Sources: American Immigration Council, analysis of USCIS employment-based data · National Foundation for American Policy, research briefs · Visa Bulletin for September 2026 · Common App End-of-Season Report 2025–26.

This article is general information about the American Immigration Council analysis of USCIS quarterly data published August 18, 2026, the National Foundation for American Policy policy brief published August 26, 2026, the September 2026 Visa Bulletin (Number 18, Volume XI, issued August 10, 2026), and the Common App End-of-Season Report published August 20, 2026. The Council and NFAP documents are third-party analyses of government data, not government pronouncements. This is not legal advice and does not create an attorney-client relationship. Past adjudication statistics describe a population and do not predict any individual outcome; no approval rate is promised or implied. Content is current as of August 29, 2026. Sources: https://www.americanimmigrationcouncil.org/report/uscis-employment-based-immigration/ · https://nfap.com/research-2/ · https://travel.state.gov/content/travel/en/legal/visa-law0/visa-bulletin/2026/visa-bulletin-for-september-2026.html · https://www.commonapp.org/files/DAR/deadline-updates/2025-26/Common-App-End-of-Season-Report_25-26.pdf

Ozek Law Firm, LLC · 4500 East West Highway, Suite 150, Bethesda, MD 20814 · +1 (202) 854-8545 · info@ozeklaw.com · www.ozeklaw.com
Consultations in English, Turkish, Spanish.