New official numbers show that the U.S. government failed to issue a quarter of the available employment-based green cards for legal immigrants in fiscal year 2021. Despite about 1.4 million immigrants waiting for an employment-based green card in 2021, the government only used 195,507 of the 262,288 available green card numbers, wasting 66,781 green cards.

The official, final number of wasted green cards is somewhat more than the 62,000 number government officials told the press at the end of FY 2021. Table 1 shows the cap for each of the five employment-based categories and the number of green cards wasted in each category. With 84 percent unused, the EB‑5 investor category had the most green card waste in FY 2021 as a percentage of available green cards. In absolute terms, the greatest waste came from the EB‑3 category for (mostly) bachelor’s degree holders, which wasted 19,774 green cards last fiscal year.

The numbers in the wasted column of Table 1 overlook an important detail of how the employment-based immigration caps are supposed to work. If any EB‑4 and EB‑5 numbers are going to go unused, those numbers are supposed to be used by EB‑1, and if EB‑1 doesn’t use all of them, they can be used by EB‑2, and then any unused for EB‑2 can be used by EB‑3. This means that EB‑2 could have used as many as 107,000 green cards, and EB‑3 as many as 122,000 (assuming all the other categories failed to use the same amount).

This cap system that flows unused numbers from one category to the next is supposed to guarantee that all the numbers are used every year, but last fiscal year, the system broke down. For reasons that I explained in the Washington Post last year, the U.S. government simply failed to make the necessary changes fast enough to ensure that every green card was issued, so those green cards were wasted.

Green card waste affects Indian immigrants far more than any other group of immigrants because they make up 82 percent of the immigrants waiting as a result of the caps. This is because the law limits immigrants from any single birthplace to no more than 7 percent of the green cards available in a year, and for many years, Indians have made up about half of all the new applicants. New Indian applicants enter the backlog, while immigrants from other countries get to pass them in line and receive a green card as soon as their applications are adjudicated.

But there’s a very important exception to this “per-country cap”: if the rest of the world will not use all the green cards, Indians can receive more than 7 percent of the total. In 2021, the employment-based cap was increased by more than 120,000 because the law dictates that any unused family-based green card numbers should be reassigned to the employment-based categories in the following year. 

The influx meant that Indians could finally bypass the country caps and get green cards far in excess of the 7 percent limit, and they did. For the EB‑2 and EB‑3 categories in which most Indian applicants are waiting, issuances went from 5,793 in FY 2020 to 43,200 in FY 2021. The additional employment-based green cards in FY 2021 effectively provided Indians with six and a half years’ worth of issuances in a single fiscal year.

But that’s exactly what makes the waste of nearly 67,000 green cards so frustrating for anyone in the backlog. The failure to process these green cards basically means that the administration has (illegally) added 11.5 additional years of waiting for Indian immigrants. I say “illegally” because the government is required to implement the laws that Congress enacts, and it has violated the law. Unfortunately, courts failed to hold them accountable, so Congress must step in and force the agencies to follow the law.

The employment-based cap is even higher in FY 2022 than it was in FY 2021: 281,430. Based on numbers from the first quarter of FY 2022 as well as statements from agency officials and attorneys, it seems clear that the agencies will duplicate their poor performance in FY 2021. As of January 2022, the agency was on pace to waste more than 100,000 employment-based green cards in FY 2022. The Biden administration must show its commitment to legal immigration by making sure that every available green card is used this year.

Opponents of the H‑1B visa often claim that H‑1B employers “pay low wages.” This has never been true, but the latest wage data prove how ridiculous this claim is. H‑1B workers are highly paid: their wages are in the 90^th percentile of all wages in the United States, meaning that they have wages in the top 10 percent of U.S. wage earners. H‑1B workers are not low wage or “cheap” workers in any honest assessment of the meaning of those words.

The Bureau of Labor Statistics announced last week that the median wage for all U.S. workers in 2021 was $45,760 and that the 90^th percentile for all U.S. wages was $102,810. Meanwhile, according to the Department of Homeland Security (DHS), the median wage for H‑1B worker in 2021 was about $108,000—above the 90^th percentile and well over double the median wage for all U.S. workers. The majority of H‑1B workers are making more than all but about 8 percent of U.S. workers.

Figure 1 shows how H‑1B wages have surged above the 90^th percentile for the first time since DHS has reported H‑1B wages in 2003. It also shows that H‑1B wage growth has exceeded the growth for all U.S. workers. From 2003 to 2021, the nominal median H‑1B wage grew 52 percent, while the nominal median for all U.S. workers grew just 39 percent. If H‑1B employers could just pay whatever they want—as opponents claim—these increases would not be happening.

DHS held its annual lottery last week to allocate the 85,000 cap-subject H‑1B temporary work visas for high skilled foreign workers. The exact number of entrants is not clear, but once again, demand for H‑1B workers far exceeded the cap. In 2021, employers submitted more than 300,000 registrations. The wage data show that the U.S. government is turning away billions of dollars in production, decreasing tax revenue and harming U.S. competitiveness. There is no reason to impose an arbitrary limit on H‑1B visas. The numbers should reflect the demand for workers.

Of course, the H‑1B program has its problems. H‑1B workers should be able to receive promotions, change jobs, and start businesses without onerous government restrictions. DHS should automatically send them green cards shortly after their arrival without needing to undergo yet another bureaucratic and expensive effort that is often delayed by ridiculously low and outdated caps, both overall and on immigrants from specific countries.

But in reforming the green card process, Congress should not ignore the massive economic benefits from the H‑1B program—which is the initial on-ramp for the vast majority of skilled green card applicants. The latest wage data only further confirm the urgency of raising the H‑1B cap and making hiring foreign workers easier.

The State Department has slowly reopened most consulates under President Biden, and it is now cumulatively issuing about as many immigrant visas for prospective legal permanent residents across all consulates as it was issuing before the pandemic. Nonetheless, the reopening has not happened equally across the world. Consulates from North America and South America are issuing enough visas to reduce their backlogs of applicants, while immigrants from Africa, Asia, Europe, and Oceania are still seeing large declines in immigrant visa issuances. Africans have seen the largest decline.

As Figure 1 shows, the State Department has begun issuing immigrant visas at almost the same pace as fiscal year 2019 overall at all embassies and consulates around the world. Despite a massive backlog of nearly 469,017 applicants, it is still not recovered to any thing close to the monthly issuance rate from fiscal year 2016 when there was an average of 55,167 issuances per month. At the end of fiscal year 2021 (ending in October 2021), it appeared that the department would fully recover, but in fiscal year 2022, the rate has once against returned to a lower level.

As Figure 2 shows, there is significant disparity among regions within the overall recovery of immigrant visa issuances. As of February 2022, immigrants from the Americas have seen significant increases in the issuance of immigrant visas compared to FY 2019’s monthly average. South America in particular saw a 43 percent increase in immigrant visa approvals from FY 2019 to FY 2022. Meanwhile, Europe, Oceania, Asia, and Africa have seen decreases of about 20 percent or more. Africa saw the largest decline of 33 percent compared to FY 2019.

Figure 3 shows the change on a monthly basis compared to the FY 2019. As it shows, immigrant visa processing effectively stopped everywhere during 2020 (except for September 2020 in Europe). The administration appeared to rush to process more visas at the end of fiscal year 2021, but fiscal year has seen a return to a much lower rate—particularly for Africans.

One reason for the disparate outcome for Africans is that the Biden administration has deprioritized the diversity visa lottery program. As Figure 4 shows, the diversity lottery program is operating 84 percent below the monthly average for FY 2019, compared to 16 percent for family preference (or capped family cases). The immediate relatives, employment, and other categories are all operating above their FY 2019 levels. Before 2020, the diversity lottery program provided about a third of all immigrant visas that went to Africans. The rest of the world received just 6 percent of their immigrant visas from the diversity program. This means that the Biden administration’s decision to make the lottery program the lowest priority for immigrant visa issuances has caused Africans to lose out more visas than immigrants from other regions. It is only because federal courts have intervened to require the administration to process diversity visas that any significant numbers were issued at all in 2020 and 2021.

Unlike other categories, diversity lottery winners must receive their immigrant visas before the end of the fiscal year or they lose their chance to immigrate forever (absent a court intervening, and so far the Biden administration is trying to prevent diversity lottery winners from 2020 and 2021 from receiving the visas that courts have said they are entitled to by appealing those decisions).

The inequitable recovery in immigrant visa processing highlights how the Biden administration is resuming the operation of the immigration system that the Trump administration dismantled for only certain groups. It has not adopted the aggressive policies needed to fix the system for everyone.

The Centers for Disease Control (CDC) is supposed to reveal as soon as tonight whether it will renew an order under Title 42 of the U.S. Code that requires the Department of Homeland Security (DHS) to block and expel any immigrant who enters or seeks to enter without prior permission, including asylum seekers. If the CDC maintains the order until at least the end of May—as several outlets have reported—it should immediately suspend its application to those seeking asylum at ports of entry.

By applying Title 42 at ports, the CDC created a perverse incentive for immigrants to cross the border illegally, see if they could evade detection, and—if they could not—hope that DHS lacked the capacity to expel them. There was no reason to wait to apply legally once the ports were closed. This fateful and incomprehensible decision created absolute chaos at the border, causing border stations that CDC’s order was meant to keep empty to overflow. 

As I have explained in greater detail elsewhere, CDC’s decision effectively invented illegal immigration from Mexico for two nationalities in particular: Haitians and Cubans. Both groups had previously always been permitted by Mexico and the United States to apply for asylum at ports of entry. After the decision to end asylum at ports, they began to enter en masse illegally, resulting in the largest number of Cuban and Haitian illegal land border crossings ever.

The Cuban and Haitian experience proves that illegal immigration is choice made by the U.S. government (in collaboration with Mexico) to stop people from entering legally. If Mexican authorities do not block them, other nationalities could also be processed at ports of entry. Ukrainians and some Russians at ports are already receiving special exceptions from the order and entering the United States legally. DHS could unilaterally expand these exemptions using the discretionary authority already provided to DHS in CDC’s current order, but it would be even better for CDC to specifically order DHS to do this. 

By blocking asylum at ports, the CDC’s order is directly encouraging illegal immigration. There is absolutely no reason to extend this absurd and illogical aspect of Title 42. DHS should immediately start processing all immigrants seeking asylum at ports of entry. Asylum seekers would constitute a negligible part of the daily inflow of legal crossers from Mexico at ports of entry, and their admission would displace illegal crossings. The phase-out period until May 23 would give all asylum seekers time to learn about the new system, and this would end most illegal immigration from countries other than Mexico.

The Biden administration has repeatedly said that it wants a “humane and orderly” immigration system. The time to start creating that system is now. 

President Biden announced yesterday that his administration “is announcing plans” to “welcome up to 100,000 Ukrainians and others fleeing Russia’s aggression.” He doesn’t explain who the “others” are, but 100,000 would amount to less than 3 percent of the Ukrainian refugees outside of Ukraine and 1 percent of Ukraine’s displaced population including those inside Ukraine. It is welcome that the administration is finally showing interest in helping Ukrainians reach the United States, but this large-scale plan simply will not happen.

The agencies have not produced any logistical or operational plans to carry it out, and every indication is that the administration is still not serious about admitting many Ukrainians. Biden has repeatedly promised far more refugee resettlement than he has delivered, and all the signs indicate that this announcement is just an aspiration. 

The reasons for skepticism are plentiful. Biden did not immediately raise the refugee target to 125,000 as he promised during and after his campaign. When he finally did raise the cap to 62,500, he failed to actually increase admissions at all and resettled in 2021 the fewest refugees ever. Then, when he did raise the cap to 125,000 for FY 2022, he admitted just 8,000 through half the fiscal year. Separate from the formal refugee program, he promised to evacuate Afghan allies using “humanitarian parole,” but as soon as the initial evacuation ended, he abandoned nearly all of them.

There is every reason to believe that this latest target is just another in a list of immigration goals that Biden will fail to achieve. To begin with, there are zero operational details about this supposed effort. Officials are saying they are “developing” plans, but the immigration bureaucracy cannot even roll out the most basic immigration regulations quickly. By the time these plans are finalized, the war will either be over or at such a disastrous point that 100,000 will look pathetic. 

The historic backlogs at the State Department consulates and U.S. Citizenship and Immigration Services (USCIS) will not suddenly disappear because Biden said he wants to let in 100,000 Ukrainians. The administration cannot even process its normal flow of immigrant visa applicants. USCIS processed just 1,000 of 40,000 Afghan humanitarian parole applicants in 7 months, and it denied more than 80 percent of them on spurious grounds.

The idea that these logistical problems will not affect the Ukrainian processing is a fantasy. Indeed, the administration is reportedly planning to select Ukrainians for any hypothetical special new programs using narrow criteria. This will only make processing large numbers more difficult because it will require agencies to analyze more evidence.

Moving toward the 100,000 goal, the administration plans to count “full range of legal pathways to the United States,” including the formal refugee admissions program, parole, and immigrant and non-immigrant visas. In other words, it is going to count the immigrants, refugees, and Ukrainian visa holders who would have come to the United States anyway—that is, people who would have come whether the war occurred or not. It may even count the 75,000 Ukrainians who will be granted Temporary Protected Status who were already in the United States at the time of the invasion.

There is actually a way for the United States to admit Ukrainians by automating travel authorization by using the Visa Waiver Program screening system known as ESTA, and then granting humanitarian parole at ports of entry. Several House members of Congress led by Rep. Jason Crow (D‑CO) included this idea in a letter to the administration this week. But so far the administration has said that it opposes any “major” evacuation effort, which makes the idea of admitting 100,000 refugees seem even less likely.

The administration needs to stop responding to crises after the fact and start developing proactive solutions that would allow the system to quickly respond to crises as they emerge. But the current paper-based immigration system that incorporates outdated, redundant, and unnecessary vetting is ultimately far from the system that can do that. The United States can do better than its current broken system, and it should use this moment as an opportunity to fix it.

The United States hit a new record of about 7.7 million immigrants at various stages in its family-based permanent residence process in 2021—an increase of about 400,000 since 2019 and nearly a million since 2016. The staggering number is largely the consequence of outdated numerical limits on green cards last updated in 1990. As a result of the numerical limits on green cards, the big backlog will result in massive, decades-long wait times, and nearly 1.6 million family-sponsored immigrants will die before they have the chance to immigrate to the United States legally.

Figure 1 shows the growth of the backlog by stage in the process from 1992 to late-2021. When those limits first took effect in 1992, the family-based green card backlog was only 3.3 million. It is important to note that these estimates differ significantly from the most common source for information on the family-sponsored green card backlog: the State Department’s annual immigrant visa waiting list report. The numbers from that report are shown in orange, but that report does not include several groups of applicants. It excludes the “immediate relative” categories, anyone waiting to apply inside the United States, and—most importantly—anyone whose petition is yet to be adjudicated. As Figure 1 shows, 3 million had a sponsor’s petition pending.

Figure 2 breaks down the family-based backlog into its two main categories: immediate relatives (“uncapped”) and family preference immigrants (“capped”). Immediate relatives—spouses, minor children, and parents of adult U.S. citizens—have no direct numerical limits (though their admissions reduce the cap for family preference immigrants from 480,000 to not more than 226,000). Family preference immigrants are spouses and children of legal permanent residents, adult children of U.S. citizens, and siblings of adult U.S. citizens, as well as any spouses and minor children of those relatives. Family preference immigrants made up 89 percent of the family-based backlog in 2021.

The capped family preference categories’ backlog has grown from 3.3 million to 6.9 million. The family preference system has 5 sub-categories. Unmarried adult children of U.S. citizens (F‑1) constituted 6 percent of the backlog in 2021, spouses and minor children of legal permanent residents (F‑2A) 9 percent, adult children of legal permanent residents (F‑2B) 9 percent, married children of U.S. citizens (F‑3) 18 percent, and siblings of adult U.S. citizens 58 percent. Figure 3 shows the family preference backlog including both pending petitions, approved petitions, and pending adjustments of status.

Given these enormous backlogs, the wait times for recent and future applicants have grown into the decades. Based on the issuance rates in a typical year (2019), the average family-sponsored immigrant applying today would face a projected wait of about 19 years, but for several nationalities, the lines are longer than a person’s expected lifespan (224 years for a Mexican sibling of a U.S. citizen). Table 1 shows the projected likely number of deaths of family-sponsored immigrants (based on Social Security mortality tables, which may overstate the life expectancy for immigrants abroad). Effectively the sibling category where the shortest projected wait is 45 years is not even viable for most new applicants and nearly all older applicants would die.

Nearly 1.6 million family-sponsored immigrants currently in the backlog will die before they receive green cards. Mexicans will account for about 60 percent of the deaths in line. Mexicans are much more likely to die in line than other nationalities because green cards are not awarded proportionally to the number of applicants. Instead, each birthplace receives no more than 7 percent of the family preference green cards available under the caps each year.

There was never any reason to cap the number of family-sponsored green cards, and Congress should uncap them again, as was the case for the Western Hemisphere until the Immigration Act of 1965. Instead of uncapping green cards, Congress should permit backlogged family members to guarantee nonimmigrant visas for these close relatives of U.S. citizens, so that they are not permanently separated from their family while they wait. Even if it doesn’t uncap them entirely, Congress should take other measures to increase green cards—including exempting spouses and minor children of primary applicants—and recapturing previously unused green cards.

The United States had about 1.4 million employment-based immigration cases winding their way through its permanent residence process in 2021. The backlog of cases is nearly ten times the total number of green cards generally made available for employment-based immigrants in most years. Skilled Indian immigrants will suffer the most from this backlog, with more than 200,000 likely to die before they could conceivably receive a green card (absent a change in the law). Another roughly 90,000 children of immigrants—mainly Indians—will “age out” of green card eligibility during their waits. Only about half of the pending Indian immigrants will likely receive green cards under current law.

Figure 1 shows the number of employment-based immigrants by year since 2018 (the first year that the government reported the data) broken down by stage in the process. As of September 2021, there were nearly 320,000 petitions pending, about 875,000 approved petitions that were wait listed because of the annual green card limits, and about 245,000 who had filed an application for adjustment of status to receive a green card (i.e. legal permanent residence).

Figure 2 shows the backlog by country of birth for approved, wait listed petitions (i.e. those approved petitions without the ability to file for permanent residence because of the green card caps). As it shows, 82 percent of the approved petition employment-based backlog was born in India. Another 12 percent was born in China, and the rest (5.4 percent) were born in other countries. This disparity is caused by the per country limits, which prevents people from any single birthplace from receiving more than 7 percent of the employment-based green cards in any year (unless the green card numbers would otherwise go unused).

Figure 3 shows the backlog for approved, wait listed petitions by category. There is almost no wait list for EB‑1 (those with extraordinary ability, multinational executives and managers, and outstanding professors and researchers). Nearly three quarters (73 percent) of the wait listed petitions are in the EB‑2 category for employer-sponsored immigrants with advanced degrees or exceptional ability. Another 16 percent are in the EB‑3 category for employer-sponsored immigrants with bachelor’s or associate degrees (or the equivalent in experience). Less than 1 percent are immigrants with job offers requiring less than 2 years of training (EB-3O). EB‑4 special immigrants (religious workers, various U.S. government-affiliated immigrants, and abandoned children) and EB‑5 investors each constituted 5 percent. While all other categories saw significant declines from 2020 to 2021, the EB‑2 wait list only barely declined.

Table 1 shows the number of employment-based immigrants wait listed by country of origin and category, as well as the projected years to wait for a green card under normal rates of green card issuances (using 2019 as the baseline). EB‑2 and EB‑3 Indian applicants filing this year (those with a master’s or bachelor’s degree) face a wait of about 90 years. Obviously, this is an impossibly long wait. As Table 1 shows, about 215,000 petitions will expire as a result of the death of the immigrant before the immigrant receives a green card, and more than 99 percent of these deaths will be Indians.

Another roughly 90,000 children of employment-based immigrants will “age out” of eligibility when they turn 21 — because only children under the age of 21 may qualify based on their parents’ application — largely children of Indian applicants. This persistent separation of families’ applications is a brutal consequence of the low green card caps, and the fact that it afflicts mainly Indian immigrant families is the result of the discriminatory country caps.

Congress should completely eliminate the green card limits for employment-based immigrants. The caps serve no purpose because nearly all wait listed, employer-sponsored immigrants are already in the United States working in temporary statuses, and the caps on investors harmfully discourage investment in the United States. These immigrants have already proven their high merit based on their employers being willing to incur the costs of getting them permanent residence. The United States should not drive talented people to live in other countries.

If Congress cannot bring itself to eliminate the overall caps, it should at least remove the country caps which shift the full burden of the overall cap onto one group (Indians), and it should index the caps to economic growth so that they do not quickly become antiquated. It should stop counting spouses and minor children against the caps, because it makes no sense to reduce the number of available workers if one of them happens to be married or have children. These reforms would improve the problems of aging out and deaths in the backlog.