Analyzing the demand/supply balance for rural and high unemployment set-aside visas

EB-5 is an investor visa program. The promise of visa eligibility is what attracts an EB-5 investment, and visas are subject to numerical limits. EB-5 has nothing to sell beyond those limits except empty promises, so we have strong reason to track pipeline visa demand and inventory.

Like promoters for a general admission stadium concert, regional centers and issuers should know the number of seats in the stadium (EB-5 visas available) and number of tickets already sold for each section (I-526 filed by visa category). EB-5 inventory tracking is stressful considering multiple parties offering tickets, small supply, enthusiastic demand that’s historically exceeded stadium capacity, and a government that resists disclosing the status of ticket sales. But we do our best, because self-interest depends on it. I don’t only care to avoid the fraud of investor visa offerings without visas. I want to assess potential market size and predict how long EB-5 business can last.

The best post-RIA EB-5 inventory intel to date is from an AIIA Freedom of Information Act request, which got data for I-526 and I-526E filings through April 2023, itemized by TEA category and investor country. I previously linked to AIIA’s article on the data, and the AIIA member webinar was published today on Youtube for anyone to watch.  I recommend the webinar for a good collection of data slides, and for discussion by a panel including yours truly, Halston Chavez (Galati Law), Joseph Barnett (WR Immigration) and Charles Oppenheim (formerly of the Visa Control Office at Department of State, now at WR Immigration). We talked about translating I-526 receipt numbers into visa demand calculations, and analyzed supply factors including how carryovers and country caps work. The webinar is a great chance to hear from experts who approached the topic from different angles and without trying to sell anything or feed a message, just aiming to empower listeners. Thanks to AIIA for organizing, and to Galati Law for helping to obtain the data from USCIS.

This post is another attempt at the tough task of simplifying/explaining a complex picture. Table 1 summarizes the FOIA data, and Table 2 offers a big-picture visa supply/demand comparison. The headline: pipeline demand for the high unemployment set-aside category was already entering backlog territory as of April 2023, the rural category had some demand but still well below foreseeable visa supply, and the infrastructure category remained untouched.

Table 1. Number of I-526 and I-526E filed April 1, 2022 to April 30, 2023, by TEA Category and Country of Chargeability (summarized from data in the USCIS response to FOIA request by AIIA)

Category	China	India	Rest of World	Total	% Total
Rural	247	57	73	377	24%
High Unemployment	429	180	481	1,090	69%
Infrastructure	–	–	–	–	0%
Not TEA	23	15	75	113	7%
Rural and High Unemployment	2	–	3	5	0%
Total	701	252	632	1,585	100%
% Total	44%	16%	40%	100%

Table 2. Estimated Pipeline Demand for High Unemployment and Rural Visas Compared with Supply

	A	B≈A*2	C	D	E
	Input Fact: Total HU I-526 filings up to 4/30/2023 (mostly filed 9/2022-4/2023)	Estimated pipeline HU visa demand as of 4/30/2023 (if visa demand ≈ I-526*2.0 )	Estimated allocation of HU Visa Supply by Country	Approximate HU Visa Supply in Carryover Year 1	Approximate HU Visa Supply Without Carryover
Total	1,090	2,180	Total	2,000	1,000
China	429	858	Minimum 7%	140+	70+
India	180	360	Minimum 7%	140+	70+
Rest of world	481	962	Maximum 86%	1,720	860
	Input Fact: Total Rural I-526 filings up to 4/30/2023 (mostly filed 9/2022-4/2023)	Estimated pipeline rural visa demand as of 4/30/2023 (if visa demand ≈ I-526*2.0 )	Estimated allocation of Rural Visa Supply by Country	Approximate Rural Visa Supply in Carryover Year 1	Approximate Rural Visa Supply Without Carryover
Total	377	754	Total	4,000	2,000
China	247	494	Minimum 7%	280+	140+
India	57	114	Minimum 7%	280+	140+
Rest of world	73	146	Maximum 86%	3,440	1,720

Summary of conclusions from Table 2 estimates:

• Table 2 is designed to facilitate big-picture/ballpark estimates of how close we were as of April 30, 2023 to maxing out visa availability in EB-5 set-aside categories. Column B gives a pipeline visa demand estimate calculated from I-526 filings, while Columns D and E show approximate annual visa supply.
• When I look at Table 2, I first look at the totals in B and D. When Total D > Total B in a category, then we’re not looking at backlog risk for the first tranche of Chinese or Indian applicants in that category. If Total D < Total B, then I start looking down at country-specific supply/demand numbers and thinking about backlogs/wait times. Column D represents a year fattened with carryover visa supply; once the first tranche of applicants absorbs those visas, subsequent years (Column E) will have half the supply and even greater backlog risk.
• The Rural category was looking good as of April 2023, with over 4 visas available in a carryover year and over 2 visas available in a normal year for every one applicant estimated to be in the pipeline at that time.
• The High Unemployment category appears already approaching the danger zone as of April 2023, with total pipeline visa demand sufficient to exceed total annual supply even in a year with extra carryover visas. That level of demand would be twice available supply in a normal year, and place China and India far beyond their assured visa supply under country caps. Relatively high “rest of the world” demand represents a limit on the number of visas that could be left unused (i.e. available for China/India), and also a signal that even Rest of World could exceed visa availability and face wait times. I could see the existing high unemployment set-aside backlog risk being averted if I-526E denial rates prove very high (which could well happen if a few big projects get rejected), if a significant number of people who filed for the high unemployment set-aside are actually issued an unreserved visa (which could possibly happen as a result of multiple classifications on approvals), or if it turns out that demand for high unemployment set-aside investments tanked after April 2023. Absent widespread denials or diversions, however, the market for high unemployment set-asides (especially from Chinese and Indians) can’t afford to continue at the rate exhibited up to April 2023.
• Table 2 gives a pipeline demand estimate as of April 30, 2023 based on about eight months of I-526 filings. To estimate where we are today, about eight months later, you could more or less double the figures in Column A and B — depending on whether you guess that demand for each category has been more or less brisk through the end of this year. The following are notes on the numbers and assumptions in Table 2, column by column, and examples of how to use Table 2 for calculation.

Column A: I-526 input

• The input fact in Table 2 is total number of I-526 and I-526E filings for set-aside categories from when the set-aside categories become available in 2022 up through April 2023. USCIS reported the categories checked by petitioners on Form I-526 and I-526E. The classification assigned by USCIS on approval, which is what really matters for the visa, may vary.  (USCIS might not approve the requested TEA, or might choose to approve the I-526E in more than one visa category.)
• The data AIIA received is monthly, and indicates an fairly even volume of receipts September 2022 to April 2023. If you guess similar volumes after April, then double the numbers in Column A to ballpark estimate totals as of today. Or if you sense that the investment/filing pace picked up or slowed down after April 2023, for rural and/or high unemployment categories, then adjust the multiplier accordingly to estimate total filings as of today. (AIIA has already filed an updated FOIA for more recent data, but pending response we have to guess.)

Column B: visa pipeline demand estimate

• The pipeline visa demand estimate comes from estimating how many successful visa applications will result from I-526/I-526E filings by investors. This estimate considers the fact that not every petition/application will be approved, and that the investors counted on I-526 will later be joined by family members at the visa stage. I tend to multiply I-526 receipts by 2 for a rough estimate — for simplicity and assuming a 25% denial/attrition rate and average family size of 2.8. Those numbers are basically consistent with EB-5 history, but use a smaller or larger multiplier if you guess that future denial/attrition rates or actual average family size will be higher or lower.
• I call the Column B estimate “pipeline” visa demand because these people are nearly all awaiting petition processing at USCIS, and thus not visa applicants yet. But they are in the pipeline queue. The high unemployment applicant with May 1, 2023 priority date could look at the total in Column B as the estimated size of the queue ahead. If people associated with priority dates up to April 30 will claim 2,180 visas, then the May 1 applicant would wait for the 2,181th high unemployment visa.  If another 2,000 high unemployment applicants have come in since April 30, then today’s high unemployment investor would expect to wait for over 4,000 other applicants to claim high unemployment visas in advance of him. (The reality is more complicated because country cap limits eventually trump priority date order, but this is the general idea.)

Column C: Visa supply allocation

• So long as total supply for high unemployment visas exceeds total qualified demand, there’s no need for traffic control, and the visa bulletin and country caps don’t get involved. In a low-demand scenario, available visas simply get issued in priority date order to those ready to take them, regardless of country. In a year with fewer than 800 total applicants for over 4,000 total visas available, all applicants could expect a visa regardless of country of origin. On the other hand, if a year has 2,200 applicants vying for 2,000 visas, then the visa bulletin will activate to hold back 200 applicants. And the first applicants to be held back will be from countries exceeding the country cap.
• Of the country-cap limited countries (named in every Visa Bulletin Section A.3, usually and currently “CHINA-mainland born, INDIA, MEXICO, and PHILIPPINES”), only China and India also have high EB-5 demand, which is why I itemize only those countries in Tables 1 and 2. Vietnam, South Korea, and Taiwan are other countries with relatively high EB-5 demand, but the visa bulletin does not apply country caps to them because their total demand across EB+FB categories is not excessive. (As a reminder for how country caps work within categories and to whom they apply, review Section A in any Visa Bulletin, the EB-4 Federal Register explanation, this handy slide visual from the AIIA webinar, and Charles Oppenheim’s explanation from the AIIA webinar. Vietnam is a marginal case because it has appeared on the visa bulletin A.3 list in the past, starting in 2018, but it hasn’t been listed since 2021. China and India, on the other hand, have consistently been excess demand countries on the visa bulletin list every year since 2005.)
• When qualified visa demand exceeds supply, then people from country-cap-limited countries get limited to 7% of available visas plus whatever is left after rest-of-world demand. 7% is not a ceiling for China or India – it’s a baseline that can be increased to the extent that visa demand outside those two countries takes less than 86% of category supply that year. For example, what if the estimated high unemployment applicants in Table 2 Column B (2,180 total, 858 Chinese, 360 Indians, 962 Rest of World), were all qualified together in a year with 2,000 visa available. How many visas would be issued to Chinese that year? My best guess would be about 730, calculated as 140 (7% of supply) plus a share the difference between 1,700 ROW visas available and 962 ROW visas demanded.  Chinese and Indian applicants should pay attention to the Rest of World demand number even more than the 7% supply number, because ROW demand is what ultimately constrains the visa availability for China/India.  
• Can countries unlimited by the 7% country cap still potentially run short on visas? Yes, as evidenced by the visa bulletin, which currently has cut-off dates for everyone in EB-2, 3, and 4. EB-5 historically avoided this risk, because EB-5 demand used to be so concentrated in China and visa availability wasn’t so fragmented. But today, with about 40% of EB-5 demand coming from “rest of the world,” and supply numbers within each set-aside category relatively small, “rest of the world” also finds itself on the backlog radar. For example, what if pipeline high unemployment visa demand has reached 2,000 by now, and what if all those applicants reach the visa stage in FY2025? In a year with 2,000 ROW applicants for at most about 1,700 visas available to ROW (2,000 minus 7% each to China and India), the visa bulletin would have to use dates to hold back the 300 excess ROW applicants until the next year’s new supply can accommodate them.

Column D and E:

• TabIe 2 uses supply numbers rounded to the thousand to facilitate eyeball estimates. For detail of how annual supply gets calculated, with nuances from falling post-COVID EB limits and carryover as a function of usage, see Table 3 below. Considering processing times, I expect that we won’t see pipeline demand for EB-5 set-asides reaching the visa stage in a big way until FY2025. So I’m assuming few set-aside visas issued in FY2024 and thus maximum carryover numbers in FY2025. But I would love to see many set-aside visas issued this year, taking advantage of an unusually high limit and reducing pressure on future supply.

Table 3: EB-5 Visa Supply Detail

	EB Annual Visas	EB-5 Annual (7.1% EB)	Rural Annual (20% EB-5)	Rural Carryover Unused	Rural Total Annual Visas Available	Rural Visas Used
2022	281,507	19,987	3,997		3,997	0
2023	197,091	13,993	2,799	3,997	6,796	0
2024	161,000	11,431	2,286	2,799	5,085	<2,799
2025	140,000	9,940	1,988	2,286	4,274	?
2026	140,000	9,940	1,988	?	1,988	?
	EB Annual Visas	EB-5 Annual (7.1% EB)	High Unemployment Annual (10% EB-5)	HU Carryover Unused	High Unemployment Total Annual Visas	HU Visas Used
2022	281,507	19,987	1,999		1,999	0
2023	197,091	13,993	1,399	1,999	3,398	0
2024	161,000	11,431	1,143	1,399	2,542	<1,399?
2025	140,000	9,940	994	1,143	2,137	?
2026	140,000	9,940	994	?	994	?