EB-5 Updates

Complete I-526 and I-829 data for FY2019 Q1, by country

Buried deep in the Electronic Reading Room, where USCIS probably hoped no one would ever find them, are two Freedom of Information Act files that individually record every I-526 and I-829 receipt and adjudication from October to December 2018 (FY2019 Q1).

• I-526 Approvals, Denials, and Receipts for Q1 of FY19 (PDF, 812 KB)
• I-829 Approvals, Denials, and Receipts for Q1 of FY19 (PDF, 258 KB)

Being diligent, I discovered the files, and immediately converted them to Excel and got to work with pivot tables. This data allows fact-based answers, at least for one quarter, to questions generally subject to rumor and speculation.

• How do USCIS processing times reports relate to actual processing times?
• Have petition processing times differed by country?
• Do approval rates differ by country?
• From which countries are I-526 receipts coming?

Before considering answers to these questions from FY2019 Q1 data, consider FY2019 Q1 in context.

IPO apparently made dramatic processing changes between 2018 and 2019, and FY2019 Q1 has one foot on either side of that change.  So what happened in FY2019 Q1 isn’t necessarily representative of what came before or after. But for what it’s worth, here’s analysis of exactly what happened in that one quarter.

• Processing Times Questions
  • Back in December 2018, the USCIS processing times report gave an “Estimated Time Range” of 20.5-26 months for I-526, and 30-39 months for I-829.  USCIS claims that in this range, “The first number is the time it takes to complete 50% of cases (the median). The second number is the time it takes to complete 93% of cases.” Presumably, these percentages get calculated from data for the previous month or two.
  • In reality, according to FY2019 Q1 data reports, 64% of I-526 adjudicated in October-November 2018 had been pending less than 20.5 months. 79% of I-829 adjudicated in October-November 2018 had been pending less than 27.5 months. Meanwhile, 16% of I-526 and 14% of I-829 adjudicated had been pending longer than the outer limit of the reported estimated time range.
  • In FY2019 Q1, the actual processing times for adjudicated I-526 and I-829 were quite a bit lower on average than the processing times report would suggest, and also had more deviation from average. The link between the contemporary processing times report and actual performance is not clear. Generally, the reality was somewhat better than the report.
  • The average I-526 approved in FY2019 Q1 had been pending 17.5 months, while the average approved I-829 had been pending 26 months.
• Country Questions
  • Data on I-526 adjudications for FY2019 Q1 shows differences by country, but not enough to suggest that USCIS was already using a visa availability approach last year.
  • The average processing time for Chinese I-526 approved in FY2019 Q1 was just two months longer than the average for other countries.
  • The average processing time for India I-526 approved in FY2019 Q1 was almost five months shorter than the worldwide average, likely due the influence of expedite requests. Indians accounted for 30 of the 36 I-526 processed in FY2019 Q1 within six months of filing.
  • Indians accounted for a majority (31%) of the I-526 filed in FY2019 Q1, followed by China (15%), Vietnam (11%), and South Korea (6%). Indians filed enough I-526 just in Q1 to use up over a year and half of the EB-5 visa quota for India.
  • Chinese, as might be expected considering past demand trends, accounted for the majority of I-829 filed (81%) and adjudicated (81%) in FY2019 Q1.
  • The I-526 approval rate in FY2019 Q1 was over 90% for most countries, but just 81% for China. I suspect this is due to USCIS’s surreptitious policy change regarding currency swaps, which particularly affects China.
• Other Notes:
  • The records show that USCIS codes at least two kinds of I-526 denial: Denied Fraud, and Denied Others. In FY2019 Q1, only one petition was denied due to fraud.
  • USCIS may not have its best and brightest on data entry and record-keeping. The “country” column for I-526 receipts, for example, includes 20 petitions coded as coming from Falkland Islands (presumably standing for Great Britain, where DOS categorizes Falkland Islands), 13 from “Unknown,” 8 from USSR, and 1 from United States. Also the totals for the quarter do not exactly match the official report of I-526 and I-829 data for FY2019 Q1.  (For example, 1,808 I-526 receipts in the official quarterly report; 1,743 I-526 receipts recorded in this detailed report.) However, please do not be shy USCIS: publishing slightly inaccurate records is a thousand times better than hiding data, leaving the industry to rumor and speculation.
  • It’s always been clear that EB-5 processing is not simply First-In-First-Out. The USCIS Estimated Time Range for processing would obviously not be so broad under a FIFO system, and the range in actual processing times is even broader. But what explains why some petitions have been processed years earlier or later than others? One factor that’s obvious in the data — denial decisions go “out” much later than approvals.
  • To repeat: petition processing has not been strictly FIFO.  This is clear, looking at the dates of petitions that received decisions in this one quarter. The PDF files linked above record individual decisions. If the FY2019 Q1 record shows that one I-526 with X filing date got approved or denied, does that mean that every I-526 with X filing date has been adjudicated? No.

And now some charts based on the FY2019 Q1 data.

I’m compiling materials for a new data room, and hope to launch a new processing time estimate service later this month following the EB-5 stakeholder meeting with USCIS.

And a few legislative notes. Senator Mike Lee continues to work on the S.386 Fairness for Highskilled Immigrants Act to eliminate country caps and reorganize the order of EB visas. Could the bill that’s been on the table since 2011 actually move in 2020? I doubt, but Lee is pushing hard.  The competitor RELIEF Act has just a few sponsors so far.

Meanwhile, bright-eyed representatives Cardenas (D-CA) and Stivers (R-OH) have introduced H.R. 5971 Case Backlog and Transparency Act of 2020. This bill refers back to P.L. 106-313, which was passed in the year 2000 with this beautiful sentence in section 202: “It is the sense of Congress that the processing of an immigration benefit application should be completed not later than 180 days after the initial filing of the application.” (p. 12) H.R. 5971 proposes to revive that deadline, and require DHS to report in detail on backlog reduction efforts. Lovely! If only the current Congress could agree that immigration benefit applications should be processed efficiently.

And as always, my PayPal link is open. If my work to find and analyze data is helpful and time-saving for you, consider making a contribution to support the work. And thanks to my past contributors!

FY2019 EB-5 Visa Stats by Country

The Report of the Visa Office 2019 has been published, with EB-5 visa statistics in Table VI Part IV (visas issued through consular processing) and Table V Part 3 (consular processing plus I-485 status adjustment).  The statistics reflect number of green cards issued for conditional permanent residence by country of origin.

In a sense this is old news – not only because Charles Oppenheim summarized this data at the IIUSA conference last October, but also because EB-5 visas issued in FY2019 reflect EB-5 investment decisions made at least two years ago (for most countries, considering I-526 processing times) or five years ago (for China, considering the visa bulletin). To understand current EB-5 demand, we need per-country I-526 data from the beginning of the process. But USCIS resists disclosing such I-526 data, so we make do with visa statistics that reflect usage midway through the EB-5 process.

A few questions that occur to me, as I look at visa statistics:

• How close did Department of State get to its goal of issuing the total visas available for the year under numerical limits?
• How are EB-5 applicants divided between people living abroad and people already in the US? What populations in the United States are using EB-5 to adjust status?
• Beyond the few top countries, how is the EB-5 market diversifying or concentrating?
• How many EB-5 visas are actually going to investors, and how many to spouses and children?
• Which data points deserve a film contract?

The EB-5 numerical limit is not a fixed number, but 7.1% of a total number of EB visas that varies each year, further divided by the 7% per-country cap. The EB numerical limit for FY19 was 141,918 visas, which put the EB-5 share for FY19 at 10,076 visas, and the individual country share at 705 visas. In practice, it’s not possible to hit the targets exactly. In FY2019, DOS unluckily undershot the worldwide target (issuing only 9,478 total EB-5 visas) but slightly overshot the per-country target for India and Vietnam (which each ended up with more than 705 visas).  The worldwide visa numbers don’t reflect lack of demand (there were plenty of visa applications left pending at the end of the year), but complications in the process (p. 2-3 of this article explains some reasons).

US entrepreneurs promoting EB-5 investments may wonder: should I buy plane tickets, or can I find potential EB-5 investors in my own back yard? Visa statistics for consular processing vs adjustment of status can help answer this question. The data shows, for example, that nearly half of South Americans who received EB-5 visas in FY19 were not living in the South America, but already residing in the US on different visas. Likewise 31% of the EB-5 visas to Europeans, and 34% of those to Indians, went through status adjustment in the U.S. By contrast, 90% of EB-5 visas issued to China-born people in FY19 went through consular processing. (But China being China, even the 10% from status adjustment in the U.S. is still a large number: 433 people). Africans got a record (for Africa) 334 visas in FY2019, most of them issued abroad.

The Report of the Visa Office does not itemize visas by principals and derivatives, but the DHS Yearbook of Immigration Statistics does. I’ve added a pie chart below with the most recent data (2018) as a reminder that the 10,000 or so annual EB-5 visas do not – as Congress intended – support 10,000 investments in the US economy, or 100,000 jobs. Because Department of State believes that it needs to fit whole families into the numerical limit, the EB-5 quota is only able to incentivize around 3,300 investments annually. In FY2018, just 3,363 EB-5 visas went to principals i.e. EB-5 investors. The majority of EB-5 visas (42%) went to children. (Interestingly, nearly a third of EB-5 applicants in FY18 apparently immigrated without spouses.)

Back to the Report of the Visa Office, FY2019 was similar to FY2018 in terms of country diversification, with similar regional distribution and number of countries contributing to the visa total.  Growing diversification was more evident between FY17 and FY18. The number of visas leftover for Chinese dropped significantly between FY17 (about 7,500) and FY18 (about 4,500), but remained about the same in FY19 (about 4,300). (That could change in FY21, if the visa availability approach succeeds in pushing a larger volume of rest-of-world applicants out of I-526 to the visa stage.)

I would like to see the film about the high-net-worth Chadians and North Koreans who managed to connect with EB-5 projects, document source of funds, and secure EB-5 visas in FY2019. And all those ones promise poignant stories – I’m curious about that one Croat, the one Kazahk, the one Surinamese, and the lone Kiwi who immigrated through EB-5 in FY19.

A few charts to highlight features of interest to me.

And finally, a reminder that visas can only be issued to people with active visa applications. The March 2020 visa bulletin ends with a reminder to Chinese with I-526 approval to get documentarily qualified at NVC, or risk losing place in line. The China Final Action date just jumped five months — not due to lack of Chinese with approved I-526, but due to lack of Chinese eligible to be called for a visa interview.

E. EMPLOYMENT-BASED FIFTH PREFERENCE VISA AVAILABILTY (note from March 2020 visa bulletin)
There has been a very rapid advancement of the China-mainland born fifth preference final action date for the month of March. This action has been taken in an effort to generate an increased level of demand. Despite the large amount of registered China fifth preference demand, currently there are not enough applicants who are actively pursuing final action on their case to fully utilize the amount of numbers which are expected to be available under the annual limit.
Once large numbers of applicants do begin to have their cases brought to final action, some type of corrective action may be required to control number use within the annual limit. It is important to remember that applicants who are entitled to immigrant status become documentarily qualified, and potentially eligible for interview, at their own initiative and convenience. By no means has every applicant with a priority date earlier than a prevailing final action date been processed for final visa action.

This brochure from DOS gives an overview of the NVC process and what it means to be documentarily qualified.

I-526 backlog by country and priority date

Last week USCIS announced a process change for I-526 adjudications, replacing the (nominal) first-in-first-out approach with a “visa availability approach” that gives priority to petitioners from countries who have visas immediately available, or soon available. This replaces a system that depends on I-526 priority dates with a system that depends on priority date plus country of origin.

To interpret the change, we need to know the composition of the I-526 backlog, by priority date and country of origin. IPO can, at any time, print out and publish a report of this data. IPO did this once, in October 2018, then deleted the report. IPO now refuses to disclose current data for the currently-pending I-526 itemized by petitioner country and priority date month/year. Why? There are obvious wicked reasons – intent to obscure processing, or gratuitous naysaying – and no good reason that I can think of. Please, reconsider, IPO.

In absence of answers direct from USCIS, here’s my estimate of the I-526 inventory as of October 1, 2019. (I can’t guess for a more recent date, not knowing anything about I-526 receipt or adjudication numbers since October.)

The total inventory estimate (Column A) should be nearly accurate, because it’s based on a Department of State applicant estimate that we can deconstruct into a pending petition estimate, with a few given clues. (Calculation and sources at the base of this post.)

Columns B, C, and D would be nearly accurate, if processing since October 2018 had been FIFO as USCIS claims. (According to the October 2018 pending petition inventory report, 4,630 petitions had priority dates earlier than March 2017, and 9,583 petitions had priority dates from March 2017 to September 2018. USCIS processed 4,673 I-526 from October 2018 to September 2019. So if USCIS processed the oldest petitions first, the remaining inventory as of October 1, 2019 should have no pending I-526 left from before March 2017, while all priority dates since March 2017 would still be pending. Reality contradicts the FIFO theory, judging by the I-526 processing times report and individual experience.  But I leave the theoretical calculation in Table II as a reference, until USCIS finally consents to give us the real numbers.)

Look at my estimated I-526 inventory above, and imagine the difference a “visa availability approach” will make once it starts being implemented in April.

If IPO is guided by the current Visa Bulletin Chart B (where China is in 2014, and all other countries current), then it would take only China I-526 off the table for now. That would cut about 6,000 petitions from the current workload.  At the other extreme, If IPO is guided by Charles’ Oppenheim’s long-range estimates for visa availability, then it could remove the following from the table for now: China priority dates from 2015 on; Vietnam priority dates from 2017 on; India priority dates from 2018 on; and South Korea, Taiwan, and Brazil priority dates from 2019 on. That would cut at least 9,800 petitions from the active inventory.

The impact on I-526 wait times, for majority and minority countries, will depend on how IPO interprets visa availability – but even more, on IPO’s decisions about volume of adjudications. If IPO exercises its ability (proven in 2018) to adjudicate over 15,000 I-526 per year, then minority-country I-526 can all be processed in a few months, and majority country petitions not far behind. But last year, IPO approved only 3,660 I-526. If that low performance level continues into the future, then IPO could use up its entire capacity for the year just working on minority-country petitions. That would significantly delay majority country-petitions. (Perhaps not coincidentally, 3,660 is almost exactly the minimum number of annual I-526 approvals needed to produce enough applicants for the 10,000 annual visa quota, given the historical average visas per principal. I hope IPO isn’t taking the visa quota as an excuse to shirk I-526 adjudication responsibilities, take resources away from I-526, and let the backlog of fee-paid petitions gather dust as long as can possibly be excused by visa waits.)

Theoretically, the I-526 wait time ( w ) for a petition equals inventory (i) of petitions with priority to get adjudicated earlier than that petition, divided by number of petitions that can get adjudicated per period (t). w=i/t  The visa availability approach changes i – making it smaller for minority countries and larger for majority countries, with obvious results for w. For what it’s worth, I can calculate i for different countries and priority dates, and quantify the effect of FIFO vs. visa availability approach on i.  But my EB-5 timing estimates page is still empty for the moment, as I don’t know what to make of t. Was last year’s very low volume an aberration, as IPO Chief Sarah Kendall suggested when asked about it last year, or a new strategy to be continued? I hope we can get some sense at the stakeholder meeting next month.

And finally FYI, the detail of my I-526 inventory estimate above. The sources: Department of State report of EB-5 Applicants with Petitions on file at NVC and Estimated USCIS Applicant Data as of 10/1/2019 (slide 8 in this IIUSA presentation), Department of State report of average percentage of EB-5 principal investors in visa applications (slide 15 in this IIUSA presentation), USCIS’s FY2019 quarterly I-526 data report (linked here), and the log of pending petitions as of October 2018 (my version of the file formerly posted on the USCIS website saved here).

Are these posts helpful? If so, please consider making a contribution to support the work.

2/2020 Visa Bulletin India FAD

The February 2020 Visa Bulletin has a Final Action Date of September 1, 2018 for India.  This represents another dramatic jump (with the January visa bulletin having a May 1, 2018 final action date). I previously wrote about overall EB-5 visa timing in 2020. This post attempts additional clarifications specific to EB-5 petitioners and applicants born in India.

February’s visa bulletin does mean:

• that visas can be issued in February to qualified Indian applicants with priority dates before 9/1/2018 (Qualified applicants are at the visa stage, having already received I-526 approval)
• that there are few Indian applicants at the visa stage (A note at the end of the January 2020 visa bulletin explains that the India final action date will proceed rapidly in 2020 so long as demand is low — i.e. so long as few Indians are in a position to apply for a visa)

February’s visa bulletin does not mean:

• that India demand is low at the I-526 stage
• that visas are available to people with pre-9/1/2018 priority dates whose I-526 are still pending
• that USCIS has processed or will soon process most India I-526 with pre-9/1/2018 priority dates
• that Department of State has already finished issuing visas to all Indians with priority dates before May 1, 2018 (the final action date in January 2020’s visa bulletin)
• that having reached September 2018, India’s final action date will continue to advance in the future, and will not turn back

In fact, USCIS indicates that it is currently working on I-526 filed 32.5 to 49.5 months ago (i.e. 2015 to 2017 priority dates). I-526 adjudications are not guided by the visa bulletin. Ideally, USCIS would process petitions in the order received, so that people reached the visa stage more or less in order by priority date. But evidently, this is not the case. Some 2018 priority dates are getting visas now even as some earlier priority dates remain at the I-526 stage.

We know that Indians filed at least 525 I-526 petitions from January through August 2018. (That’s how many I-526 were pending from those priority dates as of October 2018, when USCIS published this I526list report.) Over 500 I-526 would conservatively result in over 1,000 visa applications. Over 1,000 visas would take over one year to issue. So when the Visa Bulletin jumps from January to September 2018 in just three months, it’s evident that most people with those priority dates just haven’t even reached the visa stage. If they had, the visa bulletin would slow down to the time needed to issue over 1,000 visas.  And recall the 781 Indian I-526 with pre-2018 priority dates that were still pending at USCIS at the end of 2018 — some of those must still be in the system as well.

For Indians with priority dates before 2019, I suggest gazing at the I526list report from October 2018. Add up the number of pending I-526 with priority dates before yours. Estimate the number of visa applicants associated with those I-526. Subtract the number who could possibly have received visas since then, considering the number of I-526 adjudicated worldwide and number of visas issued to Indians since October 2018. The result estimates the number of people still with priority to get a visa before you do — assuming they can reach the visa stage. That result divided by 700 is your approximate wait time, in a FIFO world. But we’re not in a FIFO world, as the disconnect between USCIS processing time reports and the visa bulletin shows. As things stand, some people from India will get a visa unexpectedly early, and some unfairly late.  The visa bulletin will jump around, depending on who reaches the visa stage when. There’s room for hope, and room for fear — just not much predictability at this stage.

FY2019 Q4 Petition Processing Statistics

USCIS has published processing data on the Immigration & Citizenship Data page for July through September 2019 (FY2019 Q4).

I eagerly awaited this update, with three questions in mind. Would IPO start to recover processing volume in Q4, considering that previous reductions were credited to relatively minor and temporary factors (the RC authorization lapse in Q2 and I-526 training in Q3)? Would denial rates remain high? How large was the I-526 filing surge ahead of the regulations?

Now we know that IPO performance did not recover — yet —  in Q4. I-526 and I-829 denial numbers have not increased significantly. The I-526 denial percentage rose in Q3 and remained elevated in Q4 because approval numbers were so much lower than before.  People who filed I-526 by the end of September apparently got in before any significant pre-regs I-526 surge.

I-924 was a minor part of IPO’s FY19 workload, with few receipts and few pending forms. Denial numbers are remarkably high, but I suspect that many are actually withdrawals due to delayed processing. I-924 requests for exemplar approval lose value to the applicant if not adjudicated quickly.

I have a few questions for IPO Chief Sarah Kendall.

1. IPO adjudicated 2.8x more forms in FY2018 than FY2019. Please explain.
2. In one year under your leadership, IPO reversed five years of processing improvements, regressing to 2013 performance levels. Do you consider this reduced processing volume a problem that you plan to fix, or an expected outcome in your overall strategy?
3. IPO had almost twice as many staff in FY19 as it had in FY15 (214 vs 110), yet adjudicated 41% fewer forms in FY19 than FY15. Productivity per staff member was 69% lower in FY19 than it was in FY15. Do you consider this productivity loss a problem that you plan to fix, or an expected outcome in your overall strategy?
4. Looking at service-wide processing data between FY18 and FY19, EB-5 forms are the only EB forms that fell behind – other EB forms show increased approval numbers year-on-year. In fact, Form I-526 and I-924 rank #3 and #1 for worst performance in the entire service (with, respectively, 74% and 88% fewer approvals in FY19 and FY18). Only Form I-821 for Temporary Protected Status can compete, with approvals falling by 87%. Should we take a political message from these facts? Or does DHS see a problem and plan improvements in 2020?
5. Do you recognize the connection between efficiency and integrity? Do you see the problem in denials that come too late to stem bad deals, and approvals that come too late to save good deals? Do you have a plan to strengthen program integrity by improving efficiency?
6. If IPO continued FY19 processing volume into the future, then the current I-526 backlog would take three years to process, and the current I-829 backlog would take six years to process. How does IPO plan to improve going forward, to avoid such long times becoming the reality?
7. What is your goal for processing volume in FY2020? How do you plan to reach that goal?

EB-5 Visa Timing in 2020

At least 70,000 people are currently in the stage between I-526 filing and receiving visas (conditional permanent residence). Who will receive EB-5 visas in 2020? The answer depends on visa availability, I-526 timing, and the role of country caps.

Visa Availability

Thanks to roll-over of unused numbers from other categories, EB-5 has a few more visa numbers than usual to work with this year.  Department of State has allocated 11,111 visas to EB-5 in FY2020, of which any one country can get up to 778 visas (7%) under the country caps.  (China is the exceptional case, having access to 288 visas by right plus – in practice – all the numbers unclaimed by low-volume countries. About 5,200 EB-5 visas will be leftover and available to China in FY2020, DOS estimated in October – almost 1,000 more than last year.)

I-526 Timing

For many people, I-526 processing time is the major factor determining the visa wait time. We have three primary sources of information about petition processing times: the USCIS Check Case Processing Times page (which has updates about twice monthly that I log here), the Historical Average Processing Times Page (with annual averages), and the Immigration & Citizenship Data page (which has quarterly updates on processing volume). The sources present this puzzle about I-526 processing times to date:

• the Check Case Processing page currently reports an estimated time range for I-526 processing of 32.5 to 49.5 months, and has reported times in the range of 27 to 52 months since June 2019;
• the Historical Average Processing page says that I-526 were pending an average of 19.8 months in FY2019;
• the January 2020 Visa Bulletin has a final action date of May 1, 2018 for India, suggesting that there are Indians who filed I-526 21 months ago and already at the visa stage
• the Immigration & Citizenship Data page last reported 13,000 I-526 petitions pending as of June 2019. That workload could possibly take 3-5 years to process only if USCIS processed fewer than 4,000 petitions a year going forward. But as recently as 2018, USCIS was processing that many I-526 every quarter.

Looking at the puzzle (and at my charts below, which plot the bi-monthly processing time report updates against the reported annual average), I incline toward the theory that USCIS started, mid-FY2019, to inflate the months reported on the Check Case Processing page to discourage complaints and inquiries, not because the average I-526 petition has been or will be pending 3-5 years. But we’ll have a better sense of the processing reality when more recent volume numbers get published. I hope that USCIS processing improvements will be a major focus of efforts in 2020 to improve EB-5 program integrity and viability.

Visa Timing by Country:

EB-5 investors today can expect to apply for a visa promptly upon I-526 approval, unless they are from China, Vietnam, or India – countries whose demand currently exceeds the 7% of annual EB-5 visas available to each country under the country cap. The country cap adds a constraint to the EB-5 process, creating backlogs and wait times for high-demand countries while keeping the path clear for new applicants low-demand countries. This status quo has been challenged by the Fairness for High-Skilled Immigrants Act. H.R.1044 passed the House in July 2019 (365 – 65), and Senator Mike Lee has worked tirelessly since then to push the companion S.386 in the Senate, negotiating his way progressively through blocks from Senator Paul, Senator Purdue, and most recently, Senator Durban. The Fairness for High-Skilled Immigrants Act bill has been around for five Congresses without ever becoming law (with a previous version also passing the House in 2011 before dying in the Senate), so I do not really expect it to pass now. But I also do not ignore the possibility, considering the intense lobbying push in 2019, and the bill’s major consequences. If by chance it passed, the bill would remove the country cap from all EB visas, thus decreasing the EB-5 visa wait for the backlog of China applicants while increasing the wait for people with more recent priority dates. If passed, new EB-5 investors from any country could expect to wait about seven years for conditional permanent residence, based on the size of the total EB-5 backlog. Most people currently awaiting I-526 approval would face a 3-7 year wait for visa availability, depending on priority dates, if the bill passed.

Visa Bulletin

Assuming no statutory change, the January 2020 Visa Bulletin provides the following projections for EB-5 visa availability through April/May 2020:

• The EB-5 category will remain “current” for most countries, meaning that people from most countries can file visa applications as soon as they qualify, regardless of priority date.
• The final action date for China “may be possible to advance at a slightly faster pace.” (The China FAD advanced an average of just one week per month in 2019, and is at November 22, 2014 as of January 2020.)
• The final action date for India will “likely to advance at a very rapid pace until the level of demand increases.” (The India FAD has advanced a year since being set in July 2019, and jumped from 1/1/2018 in December 2019 to 5/1/2018 for January 2020.)
• The final action date for Vietnam will likely have “limited forward movement.” (The Vietnam FAD advanced 26 weeks in 2019, and has reached 12/8/2016 for January 2020.)

Here are the stories I see behind the Visa Bulletin projections:

• EB-5 will remain current for most countries, because most countries have fewer than 778 people who could possibly reach the EB-5 visa application stage this year.
• The China final action date may advance due to an increase in visa supply, but the advance will likely be slight due to concurrent increase in visa demand. DOS expects to have nearly 1,000 more EB-5 visas to give China in FY20 than in FY19, as discussed above. Greater capacity would help the queue advance more quickly. However, Chinese filed over 3,000 more I-526 in FY15 than in FY14 (13,530 vs 9,722, to be precise). The Visa Bulletin is moving into that filing surge as it starts to accept visa applicants for FY15 priority dates. Greater volume of visa applicants for FY15 priority dates would tend make the line advance more slowly.
• The India final action date has advanced rapidly because “the level of demand” has been low, and expected to remain so. “Demand” means the number of Indians documentarily qualified for a visa – i.e. with I-526 approval and visa application or I-485 in order. That number is low only because USCIS is slow and erratic with I-526 adjudication. Charles Oppenheim estimates that there were 4,000 to 5,000 Indians in line for an EB-5 visa in 2019 – enough to claim many years of visas.  But about 85% of that queue was stuck in the I-526 stage, and thus not yet qualified to demand a visa. Department of State has 788 visas to give Indians this year. Which of those 4-5K Indians in the queue gets one of the 788 visas depends on which I-526 USCIS can approve this year. So long as USCIS is very slow, only approving a few petitions, and particularly delaying old petitions, the Visa Bulletin will continue to advance the India FAD to open the door for those few who have reached the visa application stage. If USCIS increases volume of I-526 approvals, then more Indians will be able to compete for this year’s visas. In that event, the India FAD will slow its advance, or even move back in time to accommodate an influx of applications with older priority dates. Personally, I expect the India FAD to retrogress this year. Indians filed at least 330 I-526 petitions from January to May 2018, enough to absorb at least a year of visa numbers. Indians filed at least 806 I-526 petitions between April 2017 and May 2018 (April 2017 being the near end of the estimated time range for I-526 processing, according to the current USCIS processing times information page). That’s enough to absorb more than two years of available visas. All that demand will pull the India FAD back, assuming it can ever emerge from I-526 processing.  The Visa Bulletin FAD for India advanced from May 2017 to May 2018 in just seven months, and from January 2018 to May 2018 in just one month – which tells us that DOS simply hasn’t yet received the surge of visa demand that’s on the way from those early priority dates. The 300 or so Indian investors plus family from early 2018 could not possibly have all received visas in December 2019. When the petitions stuck in I-526 processing finally arrive at the visa stage, the visa bulletin will have to recalculate and may retrogress the India FAD.
• The Vietnam FAD is expected to advance slowly from the current date of December 8, 2016. One would expect the movement to be slow because Vietnamese filed many I-526 in FY2017 – 523, to be precise, as compared with 404 petitions filed the previous year. This higher volume means that FY17 priority dates will take longer to move past the visa window than FY16 priority dates. Meanwhile, many FY17 petitions have completed I-526 processing, and thus the applicants are qualified and ready to claim available visas. However, we can see that the full surge of FY17 demand from Vietnam has not yet hit the visa stage, since the Visa Bulletin Chart B is still current for Vietnam, and USCIS is still allowing applicants to file I-485 using Chart B. That window could close as the pool of qualified Vietnamese applicants grows.

Visa Retrogression

People from China, Vietnam, and India who apply for visas through adjustment of status will be interested in this document, which gives helpful Q&A on visa availability and the I-485 process: USCIS Responses to Questions from the American Immigration Lawyers Association (AILA)

RC List Updates

And to close the year, an update on changes to the USCIS lists of approved regional centers. Note that I update approvals on my blog RC List page (together which such contact info as I can find) and terminations in my Excel Terminations log (together with a log of termination reasons and links to all termination letters posted so far by USCIS).

Additions to the USCIS Regional Center List, 08/27/19 to 12/30/19.

• EB5 Affiliate Network Southeast Regional Center, LLC (Alabama, Georgia, Mississippi, South Carolina, Tennessee)
• Plymouth Park Regional Center, LP (Texas)
• Protogroup, Inc. (Florida) — Reinstated after termination in 2018
• Southeast Regional Center LLC (Alabama, Georgia)

October 2019 Oppenheim EB-5 wait time estimates

Department of State Visa Control Office Chief Charles Oppenheim presented about EB-5 visa availability at the IIUSA conference on October 29, 2019. Here are his presentation slides and my recording. (Update: Lee Li of IIUSA has written a helpful slide-by-slide commentary on Oppenheim’s presentation in his article Data Analysis on Fiscal Year 2019 EB-5 Visa Number Usage & Estimated Visa Waiting Lines.)

The October 2019 presentation suggested encouraging headlines: shorter wait times and faster-moving visa bulletin dates than previously predicted. Behind the headlines lies a vexed story that I shall tackle in this difficult post.

Post Agenda:

1. Put Oppenheim’s October 2019 EB-5 timing estimate in context of past estimates
2. Discuss how to read the “EB-5 Applicants with Petitions on file at NVC and Estimated USCIS Applicant Data” slide in Oppenheim’s presentations
3. Review the factors that can cause the EB-5 backlog to grow and shrink
4. Collect available data relevant to interpreting Oppenheim’s estimates for China, Vietnam, and India
5. Interpret Oppenheim’s estimates for India
6. Discuss how backlog data relates to estimated Visa Bulletin final action dates

1. October 2019 Presentation in Context

Oppenheim estimates total EB-5 backlog size (actual applicants at the National Visa Center plus estimated applicants associated with pending I-526) and then calculates wait times as a function of backlog divided by annual visas available.  The following table compares key results from Oppenheim’s October 2019 presentation with previous presentations on October 30, 2018 and May 6, 2019.

Note that Oppenheim’s backlog and wait time estimates fell between May and October this year for all countries except South Korea, with particularly significant  reduction in the wait time estimate for India. I didn’t expect that, considering reports of a flood of I-526 filings ahead of the November 21 regulations deadline. What’s the story? Are there indeed fewer people in line for an EB-5 visa now than there were back in May, or has there been a change or omission in Oppenheim’s calculation? If fewer people in line, how did that happen? If a change or omission in the calculation, what is it, and should that cause us to rethink Oppenheim’s past or current wait time estimates? Read on…

Interpreting the “EB-5 Applicants with Petitions on file at NVC and Estimated USCIS Applicant Data” slide

The data quoted above comes from this key slide, a version of which is included in each of Oppenheim’s IIUSA presentations since 2018.

This slide is important because Oppenheim’s wait time estimates are calculated from the orange column. For example 977/700=1.4 year estimated wait for Brazil. 4,707/700=6.7 year estimated wait for India. The calculations assume average 700 visas available per year under the country cap, though the total can vary by year. The denominator for China is less predictable. Oppenheim estimated a 16.2-year wait for China in October 2019, which means that he must have been assuming 48,589/16.2 = 3,000 annual visas available on average to China going forward. (Aside: Oppenheim did not explain why he chose the 3,000-visa assumption for China. China received 4,326 visas in FY2019, and Oppenheim estimates that 5,270 visas will be available to China in FY2020. Average visas available to China going forward will only average as low as 3,000 if rest-of-the-world demand continues to rise going forward, which seems unlikely considering impending investment amount increases. Note also that the wait time estimates should have been tagged for petitions filed as of October 1, 2019–since that’s the date of the data upon which they are based–not for petitions filed as of October 29, 2019. Considering the likelihood of a filing surge in October 2019, this distinction could be significant.)

In his October 2019 presentation, and in follow-up discussion in person, Oppenheim clarified these points about how to read the EB-5 Applicants slide:

• The “Actual Number of Applicants at NVC” column is just what it says: an accounting of the actual applicants with petitions on file at the National Visa Center as of October 1, 2019. This column does not include people who have I-526 approval but without active petitions on file at NVC for one reason or another. It does not include applicants currently seeking a visa through adjustment of status with USCIS. The column includes no assumption about the number of actual applicants at NVC who may eventually get denied. The NVC column is most significant to Oppenheim’s visa bulletin calculations, because it indicates how many people are ready to claim a visa. (The I-526 column estimates potential future demand. But no one in that column is currently qualified to claim a visa, and Oppenheim does not know for sure when and if those potential applicants will emerge from the I-526 process and become qualified.)
• “DoS ESTIMATED Number of Applicants with Petitions on File at USCIS” refers specifically to I-526 petitions, and does not include I-485 petitions. This column estimates the future visa applicants associated with pending I-526 using this formula: actual I-526 pending at USCIS * assumption about approval rate for these I-526 * assumption about average visas per approved I-526. The pending I-526 data and approval rate assumption come from USCIS. The visas-to-I-526 assumption uses the “average percentage of EB-5 principal investors” Department of State data point that divides EB-5 visas issued to investors by total EB-5 visas issued to investors plus family. Oppenheim could not disclose the specific numbers used to calculate this column for October 2019. I’m particularly sad that he couldn’t disclose what USCIS told him about pending I-526 by country. He did volunteer that the I-526 approval rate assumption in the 10/2019 DoS estimate is the same for all countries, and lower than the approval rate assumption used for previous estimates. He further indicated that the “percentage of principals” assumption in the 10/2019 DoS estimate varies by country, is based on averages for visas issued in FY2019 and FY2018.
• “Estimated Grand Total” equals the blue column plus the green column. Oppenheim has not been counting I-485 applicants anywhere in the table because historically a small percentage of EB-5 visas have gone through adjustment of status. He agreed that it would be a good idea to count pending I-485 applicants in future backlog estimates. (Another hint that the “Estimated Grand Total” might be missing something: Oppenheim estimates about 48,600 total applicants for China. This seems unexpectedly low considering that at least 35,500 Chinese filed I-526 since the start of FY2015, per USCIS data, and few of those Chinese could’ve received visas yet considering that the visa bulletin still has a November 1, 2014 final action date for China. So either the China backlog has in fact experienced major attrition along the way — plausible, considering sentiment among past Chinese investors — or some category of Chinese who still could apply for a visa are not being counted in the NVC or I-526 columns.)

Potential Factors in Backlog Total Change

The size of the EB-5 backlog is constantly changing, as people enter the line by filing I-526 and bringing family, and leave it by losing eligibility or receiving visas. To review specific factors that can cause change over time to the numbers in Oppenheim’s backlog calculations, and/or the actual backlog:

Number of “Actual Number of Applicants at NVC”

• Decreased by applicants receiving visas
• Decreased by applicants being denied visas or losing eligibility (e.g. aging out, I-526 revoked)
• Decreased by applicants abandoning their petitions (need to contact NVC annually to avoid this)
• Increased by more investors receiving I-526 approval and filing visa applications (thus moving from the green column to the blue column)

“DoS ESTIMATED Number of Applicants with Petition on File at USCIS”

• Increased by I-526 filings
• Decreased by I-526 approvals and denials
• Increased or decreased by changes to the DoS assumption about number of pending I-526 that will be approved
• Increased or decreased by changes to the DoS assumption about how many family members will be associated with each principal applicant

“Estimated Grand Total”

• Increases if increases from incoming I-526 filings plus approval rate and family member assumptions exceed decreases from outgoing applicants who received visas or lost eligibility.
• Decreases if the opposite data and assumptions prevail.
• Could increase if Oppenheim started to count populations not included in the “Actual at NVC” and “Pending at USCIS” columns. This includes EB-5 applicants on pending I-485, and possibly other people with potential eligibility (I-526 approval) who do not currently have active petitions at NVC.

Data

Here is my spreadsheet that collects data particularly relevant to questions about Oppenheim’s wait time estimates in 2018 and 2019 – basically, available data related to the above bullet points.  I gaze at and play with these numbers as I to try to back calculate Oppenheim’s estimates, answer questions, and interpret a story. I’m not showing my messy calculations, but present the inputs for the convenience of others working with the similar questions. Curating this spreadsheet was not easy.

Example Interpretation and Application

Take India as an example of the challenge to interpret Oppenheim’s estimates.

Oppenheim’s estimated India wait time fell by 1.7 years between April 1, 2019 and October 1, 2019 because Oppenheim estimated that the India backlog fell by 1,078 applicants during that period —  Q3 and Q4 of FY2019. This backlog reduction is the net of 66 additional applicants at NVC and 1,144 fewer estimated applicants associated with pending I-526.

DOS issued 252 visas to Indians in Q3-Q4 of FY2019. (We don’t know how many visa applications were denied.) 252+66=318, so apparently the 1,144+ applicants who left the I-526 column between May and October did not all transfer over to the NVC column.

It could be that many applicants were indeed approved out of the I-526 column but then disappeared into uncounted categories—ie the pending I-485 pool and the still-preparing-a-visa-application pool. If that were true, then those people are out of Oppenheim’s calculation but not out of the queue in reality. In that case Oppenheim’s latest wait time would be an underestimate.

Or, maybe few applicants were actually approved out of the I-526 column, but the I-526 column slimmed nevertheless thanks to downgraded assumptions about I-526 approval rates and family size. Oppenheim confirmed in follow-up conversation that he did indeed change assumptions about future approval rates (significantly) and family sizes (insignificantly) for the October 2019 calculation. If I-526 receipts and adjudications were about equal in Q4 (as they were in Q3), then a changed visas-per-pending-I-526 assumption could explain the entire 20% change in estimated  applicants associated with India I-526. If I-526 receipts in fact exceeded adjudications in Q4 – as I would’ve thought considering the expected pre-regs filing surge and continually lengthening processing times reports – then the visas-per-pending-I-526 assumption must have fallen by even more than 20%. If Oppenheim’s revised visas-to-investor assumptions are more accurate than his previous assumptions, then the wait time estimates from May 2019 and October 2018 were overestimates. If not, the October 2019 estimate is an underestimate.

Oppenheim’s backlog estimate does not count applicants on pending I-485. In FY2018, consular processing accounted for over 90% of visas issued to China and Vietnam, and 67% of visas issued to Indians, according to the Annual Report of the Visa Office. If there continue to be a significant number of Indian EB-5 applicants on I-485, then Oppenheim is undercounting the India backlog. For China and Vietnam, it appears relatively safe to only look at NVC numbers.

Oppenheim’s backlog estimate does not make an assumption about the number of applicants pending at NVC who will not end up claiming visas. However, this factor might be significant in reality, as suggested by Oppenheim’s commentary on the visa bulletin. The India final action date jumped in August and September 2019 thanks to an unexpectedly large return of visa numbers. Those returned numbers represent people who had been at the head of the NVC queue but then were denied at the visa interview, or missed the interview. Their disappearance resulted in visas that had been marked out for them returning to NVC and becoming available to people who had expected a longer wait time.

Overall, contemplating the numbers for India, I conjecture:

• That there can’t after all have been much of an Indian I-526 filing surge at least up to September 30, 2019 (Indeed, only South Korea clearly experienced a major filing surge in FY2019 Q4)
• That Oppenheim must now be estimating an I-526 approval rate well under 75%
• That a relatively low approval rate going forward is plausible, given trends at USCIS, and would mean that previous wait time estimates assuming higher future approval rates were overestimates
• That the current India wait time estimate is likely still an underestimate because it does not count I-485

But such conjectures are exhausting and unsatisfying. I’ve temporarily suspended my EB-5 timing estimate service, because it’s so tedious to try to navigate and quantify all the “if/thens.” And then any estimate must be so laboriously and frustratingly qualified. Until now, I have generally used Oppenheim’s point-in-time estimates as anchors for priority-date-specific timing estimates. But that doesn’t work as well when Oppenheim’s assumptions change between the points in unknown ways.  When USCIS finally publishes I-526 data for FY2019 Q4 (and even better, FY2020 Q1), we’ll at least have a few more facts to anchor estimates and to help interpret Oppenheim’s estimates. And please please please USCIS, why can’t you continue to publish data on pending I-526 by country and month of priority date? This is so important to program integrity, and not justifiable as a state secret.

Backlog Estimates and the Visa Bulletin

Oppenheim’s IIUSA presentation gave predictions for Visa Bulletin final action dates.

Again, the India case is a challenge. How could the October 2020 Visa Bulletin possibly become “current” for India in one year (meaning visas available to qualified applicants for all priority dates) if Oppenheim doesn’t expect October 2019 priority dates to have visas available for another 6+ years? This becomes possible if the pool of qualified applicants remains small despite the large total backlog. In other words, if most of the 6+-year India backlog remains bogged down in slow I-526 processing, and thus unable to claim available visas. Oppenheim apparently foresees that Department of State could find itself in October 2020 with 700 visas to give India and well under 700 Indian applicants pending at NVC. That could happen if USCIS keeps up its low volume of approvals. This situation is less likely for Vietnam and China, because there are already significant NVC backlogs for those countries from back when USCIS adjudicated more petitions.

Among the many bad consequences of slow and chaotic I-526 processing: it devalues priority dates. In December 2019, Department of State offers visas to Indians with priority dates up to January 1, 1018, according to the visa bulletin.  Meanwhile, USCIS is processing investor petitions filed 29 to 50 months ago,  according to its processing times report. That means that Indians with late 2017 priority dates can be claiming visas now, ahead of Indians with 2015, 2016, and 2017 priority dates who are still stuck in I-526 processing. Obviously, the backlog is not moving in order by priority date.  In a queue system, a person’s wait time should be a function of the number of other people already in line at the time he or she entered the queue. That would allow for fairness and predictability. But the EB-5 queue is falling into disorder thanks to the two-step process. When USCIS is slow to adjudicate I-526 petitions, and apparently advances them out of date order, then priority dates lose their predictive value. It’s not fair that an Indian with a November 2017 priority date can claim a visa today, while an Indian with a November 2015 priority date isn’t even outside of normal I-526 processing times according to USCIS. It’s not fair when wait time estimates have to ask not only “how many people were in line before me” but “how many people will be able cut in line before me thanks to disordered USCIS processing?” But that’s the fact that we face today, thanks to USCIS processing failures.

Ironically, the “best case” scenario for the October 2020 visa bulletin assumes a worst case scenario for I-526 processing. If USCIS speeds up after all, approving more I-526 and thus advancing more applicants to the visa stage, than future visa bulletin final action dates will move further back.