Backlog for Skilled Immigrants Tops 1 Million: Over 200,000 Indians Could Die of Old Age While Awaiting Green Cards

  • The Employment-Based Green Card Backlog
  • Projection of Future Wait Times
  • The Indian Skilled Worker Backlog
  • Options for Reform
  • Conclusion
  • Appendix
  • Downloads
  • Endnotes
  • Related Content

To receive lawful permanent residence in the United States, an employment‐​based immigrant must first become the beneficiary of a petition usually submitted by an employer that requests that the government allow the immigrant to apply for a green card. Even with an approved petition, an immigrant cannot apply for a green card unless the green card cap is unfilled. Because demand has increased since Congress last updated the cap 30 years ago, the number of approved immigrants whom the cap is preventing from applying for green cards is skyrocketing.

As a result of the outdated green card limits, these immigrants are waiting in a backlog that has reached an unprecedented length. New data on beneficiaries of approved employment‐​based petitions for green cards from a Cato Institute Freedom of Information Act request show how much the system needs reform:

  • For the first time, the U.S. government has approved more than 1 million petitions for workers, investors, and their families who cannot receive legal permanent residence solely as a result of the low green card caps.
  • The government is approving nearly two petitions for employment‐​based immigrants for every green card it is issuing to them. At the current rate of increase, the backlog will exceed 2.4 million by 2030.
  • Skilled Indian workers make up 75 percent of the employment‐​based backlog, and recently backlogged Indian workers face an impossible wait of nine decades if they all could remain in the line.
  • More than 200,000 petitions filed for Indians could expire as a result of the workers dying of old age before they receive green cards.
  • Indian and Chinese immigrants had average wage offers two and a half times U.S. median wages—$30,000 higher than the average offer for other immigrants—yet they face waits that are decades longer.

Indians endure much longer waits because the law imposes limits on the number of green cards for immigrants from any single birthplace and because U.S. employers file far more petitions for Indians than the limits allow. With recent skilled Indian workers facing a de facto ban on legal permanent residence based solely on their place of birth, Congress should prioritize the removal of these limits. However, this reform alone would still leave unsustainable waits of more than a decade for every employer‐​sponsored immigrant and allow the backlog to escalate past 2.4 million by 2030. To avoid driving billions of dollars in investment and hundreds of thousands of skilled workers abroad, Congress must quickly increase the number of employment‐​based green cards before this problem worsens.

The Employment‐​Based Green Card Backlog

There are at least two steps that all immigrants must follow to receive an employment‐​based green card—which denotes lawful permanent residence. First, an immigrant must become the beneficiary of a petition, usually submitted by an employer, requesting that the immigrant receive permission to apply for a green card. Second, if the petition is approved, the immigrant must apply for a green card. However, a beneficiary of an approved green card petition may only proceed to the second step and submit a green card application if the green card cap—the limit on legal immigration set by Congress—has not been reached.

Immigrant beneficiaries of approved petitions enter the green card backlog—the wait list for immigrants—if they cannot apply for green cards due to insufficient cap space. Immigrants in the backlog are not waiting for their petitions or applications to be administratively processed. Their petitions have already been approved, but they cannot yet apply for green cards. They are only waiting for cap numbers to become available. Table 1 provides definitions of the main immigration terms in this brief.

This brief details the employment‐​based backlog for those immigrants delayed solely by the caps on legal immigration, not by inefficient bureaucracy. The Cato Institute obtained the data from U.S. Citizenship and Immigration Services (USCIS) through a Freedom of Information Act request.1

Figure 1 compares the total employment‐​based backlog on April 20, 2018, to the total employment‐​based backlog on November 12, 2019.2 In little more than a year and a half, 215,395 approved petitions entered the employment‐​based backlog, growing the total waiting list from 831,826 to 1,047,221—an average of 137,852 per year. Since the employment‐​based system has a cap of 140,000 green cards per year, this means that every year there are about twice as many petitions being filed for green cards for immigrants as there are green cards being issued to them.

The employment‐​based green card limits are preventing America’s highest‐​skilled immigrants from receiving legal permanent residence. The U.S. employment‐​based immigration system is divided into five “preference” categories (and one subcategory) based on the priority that they receive for green cards. Each category has a separate annual cap. The categories are

  • EB-1, employment‐​based first preference, priority workers: 40,040 annual cap for multinational executives and internationally acclaimed professors, scientists, artists, athletes, or businessmen, plus unused EB-4 and EB-5 green cards;3
  • EB-2, employment‐​based second preference, professionals: 40,040 workers with offers of employment in jobs requiring an advanced degree or higher and workers with expertise significantly above what is ordinary in their field, plus unused EB-1 green cards;4
  • EB-3, employment‐​based third preference, skilled workers: 35,040 workers with offers of employment in jobs requiring a bachelor’s degree and skilled workers with at least two years of experience, plus unused EB-2 green cards;5
    • EB-3O, employment‐​based third preference, other workers: 5,000 workers with offers of employment in jobs not requiring a bachelor’s degree;6
  • EB-4, employment‐​based fourth preference, special immigrants: 9,940 religious workers, broadcasters, U.S. government and military employees, and abandoned juveniles;7 and
  • EB-5, employment‐​based fifth preference, investors: 9,940 foreign investors who made investments of $500,000 to $1.8 million in a new commercial enterprise in the United States that will create 10 permanent full‐​time jobs for U.S. workers.8

Figure 2 breaks down the backlog in Figure 1 into each category. The largest growth in the backlog occurred in the EB-2 category. This is the primary green card for America’s foreign doctors and other highly skilled professionals. With 628,592 petitions, EB-2 now makes up 60 percent of the backlog. The oldest petitions were filed in the EB-2 and EB-3 categories in 2009, meaning many immigrants in the backlog have already waited more than a decade for their green cards.9 From April 2018 to November 2019, every category’s backlog increased except for EB-1.

Besides principal immigrants—workers or investors who are beneficiaries of petition—their spouses and minor children (age 20 or younger) are also eligible for green cards and wait in the backlog alongside the principal applicants. Because they are subject to the same caps, family members take away green card cap space from principal immigrants, which increases the backlog and wait times for everyone.10

Figure 3 breaks down the backlog in Figure 1 into three types of petition beneficiaries: principal applicants, dependent children, and dependent spouses. It shows that family members of workers and investors now make up the majority (over 51 percent) of the employment‐​based backlog. Principal petition beneficiaries are almost 49 percent, spouses are about 30 percent, and children are over 21 percent. The number of new principal petition beneficiaries entering the backlog in 2019 roughly equaled the number of green cards available in 2019, but because families took half the green card cap, the backlog continued to grow. If dependents were exempt from the cap, the backlog would stop growing.

Employment‐​based green cards are also limited by birthplace (per‐​country caps). With an exception only to avoid wasting green cards, the law prevents immigrants of any single birthplace from receiving more than 7 percent of green cards issued in a single year. As a result, the backlog grows almost exclusively for immigrants from countries with the largest number of petition beneficiaries: India and China. Indians have the largest backlog with 780,579 petitions, making up 75 percent of the total. As Figure 4 shows, the Indian backlog increased by almost 150,000 from April 2018 to November 2019. The backlogs for China and the rest of the world have also increased, though not as dramatically.

From April 2018 to November 2019, the overall employment‐​based green card backlog increased at an annualized rate of 137,852. At that rate of increase, by the start of 2030, the green card backlog will exceed 2.4 million petitions—nearly all for immigrants from India or China. The result would be a de facto end to legal permanent immigration for new Indian skilled workers and Chinese investors.

Projection of Future Wait Times

Table 2 provides three estimates of how long it could take to process all petitions in the queue if there is no reform.

The first column of Table 2 shows how long all petitions would take to process if every petition beneficiary remains in line forever at the current rate of green card issuances. Essentially, this is the theoretical maximum wait time that all new beneficiaries face. Eight groups face waits of more than a decade—if everyone remains in line—and the longest projected wait would be for new EB-2 and EB-3 Indian employees of U.S. businesses who face a lifetime wait of 89 years (see endnote for why these categories are combined).11

The second column of Table 2 shows how many petitions could expire due to deaths of beneficiaries (see the appendix for detailed methodology). Based on Social Security Administration estimates of the average rate of death by age for individuals, 209,116 petitions could expire as a result of workers dying before they receive their green cards, even if they managed to remain eligible in line forever. Nearly all the deaths—98 percent—would come from Indians in the EB-2 and EB-3 lines. Still, it would take 63 years to process the survivors.

The third column of Table 2 considers deaths and the likely number of abandoned petitions to show how many petitions will ultimately result in green cards through the employment‐​based system (see the appendix for detailed methodology). Abandonment includes “aging out,” when a child of a petition beneficiary loses eligibility as a result of turning 21. The rate of abandonment for principal beneficiaries comes from the number of approved petitions in past years, which show how large the backlog would be had no petitions been abandoned.12 Only about 56 percent of those currently waiting for green cards could receive one through the employment‐​based process. As a result of their exceptionally long waits, only 44 percent of Indian beneficiaries of EB-2 and EB-3 petitions could receive green cards.

Petitioners may abandon their petitions for several reasons: another employer could refile on the worker’s behalf, making the initial petition redundant; the immigrant could obtain a green card another way—such as by marrying a U.S. citizen—or could give up entirely, perhaps by leaving the country. From 2016 to 2019, for example, the number of Indians applying for permanent residence in Canada doubled.13 Some Indians are also already dying in line.14

The Indian Skilled Worker Backlog

The Indian EB-2 and EB-3 backlog is by far the longest, affecting the lives of hundreds of thousands of skilled employees of U.S. businesses with advanced (EB-2) or bachelor’s (EB-3) degrees. Everyone in the current EB-2/EB-3 backlog originally applied during or after fiscal year 2009.15 Table 3 estimates the backlog based on the years when petitions were filed and estimates the dates when immigrants could receive green cards if they don’t die or leave the line (see the appendix for detailed methodology). The share of Indian petition beneficiaries likely to receive employment‐​based green cards declines dramatically from 91 percent of those who applied during fiscal years 2009 and 2010 to 24 percent of those who applied from fiscal year 2019 to early 2020. In other words, three‐​quarters of recent Indian petition beneficiaries may not receive green cards.

Under the country cap system, the law guarantees Indians only 7 percent of total available green cards, but they can exceed this number if not all the green cards are used. Unfor­tunately for them, demand from the rest of the world is increasing. From 2018 to 2019, the share of green cards awarded to Indians fell from 13 percent to 10 percent, even as their share of petitions increased from 50 percent to 53 percent. This inequity between the share of petition beneficiaries and the share of approvals explains why Indians filled 89 percent of the employment‐​based backlog in 2020 (see Figure 5).16

This unintentional discrimination against Indian workers has distinct harms beyond the long waits. More than 92 percent of EB-2 and EB-3 Indians who receive green cards are already in the United States working in temporary status—mainly the H-1B work visa that workers can renew indefinitely so long as they are in the green card queue.17 H-1B workers must maintain a job with only certain approved H-1B employers. They cannot be unemployed at any time or start their own businesses. If H-1B employers close or downsize—an obvious possibility over decades—visa holders lose their status and places in the green card queue.

Minor children of H-1B workers are also eligible for status under employed parents’ applications, but they lose their status and their places in the green card backlog when they turn 21. This means the longer the wait, the more India‐​born children will lose both their status and eligibility for employment‐​based green cards. More than two‐​thirds of minor children of Indian EB-2 and EB-3 green card petition beneficiaries (almost 90,000) could age out and lose eligibility for green cards. These children would have to self‐​deport or find new temporary or permanent visas. Meanwhile, workers from other countries and their families will pass the Indians in line and continue to receive green cards.

The per‐​country limits’ de facto anti‐​Indian policy discriminates against higher‐​skilled immigrants. In 2019, EB-2 and EB-3 Indians and Chinese had significantly higher wage offers than immigrants from other countries. In effect, the policy ushers lower‐​paid immigrants from other countries ahead of Indians.

Figure 6 shows the average wage offer for beneficiaries of EB-2 and EB-3 petitions in 2019. Indians and Chinese had average wage offers of about $120,000—two and a half times the earnings of the median wage or salary worker in the United States and a third above the average offer of about $90,000 for other EB-2 and EB-3 petition beneficiaries.18 Despite the higher wage offers to Indians and Chinese, they are forced to wait much longer. The differences in wage offers are not simply an accident of where Indians and Chinese choose to live: the average wage offers for Indians and Chinese were higher than other nationalities in every state except three.

U.S. employers clearly demand workers across the skills spectrum, but there is no economic reason to adopt policies that intentionally make higher‐​paid immigrants wait longer—or forever—for green cards based on their birthplaces.

Options for Reform

The per‐​country limits cause the most significant waits in the employment‐​based system. The Fairness for High‐​Skilled Immigrants Act of 2019 (H.R. 1044), which the House of Representatives passed on a vote of 365 to 65 in July 2019, would phase out country caps over a four‐​year period.19 Because Indians and Chinese dominate the backlog, the act guarantees other nationalities at least 15 percent of the green cards in the first year and 10 percent in the two succeeding years. It also contains a provision assuring that immigrants from all other countries who apply before enactment will not lose their advantageous positions.

Table 4 shows how the Fairness for High‐​Skilled Immigrants Act would affect the number of EB-2 and EB-3 green cards received by immigrant birthplace as well as the size of the EB-2/EB-3 backlog by birthplace, again factoring in deaths and abandoned petitions (see the appendix for detailed methodology). Table 4 also projects the backlog increase moving forward based on the share of petitions by birthplace in 2019.20 Under the bill, all new petitions starting in 2022 would immediately enter the backlog regardless of birthplace.

Table 4 reveals that it would take less than a decade for the existing backlog to be processed without the per‐​country limits compared to several decades with them (see Table 2). Furthermore, the Fairness for High‐​Skilled Immigrants Act would move policy toward its stated purpose of fairness or equity between nationalities. The share of green cards issued would more closely resemble the share of the green card backlog in 2030 (highlighted in Table 4). Inequity would still exist in 2030, however, because the advantages of other nationalities would partially continue during the first three years. While not shown in Table 4, the share of Indian petitions in the 2020 backlog eventually receiving employment‐​based green cards would increase under the Fairness for High‐​Skilled Immigrants Act from 44 percent (as seen in Table 2) to 81 percent, as a result of fewer deaths and abandoned petitions.

Table 4 also shows that if the current rate of increase in the backlog continues, the legislation would still permit a massive increase in the EB-2/EB-3 backlog by 2030—skyrocketing past 2 million. Indians would improve their position relative to the status quo but would still have more than 1.2 million petitions pending, which would take decades to process.

Senators have proposed additional reforms to address this problem. Senators Richard Durbin (D-IL) and Rand Paul (R-KY) want to phase out the per‐​country limits and increase the availability of green cards, making the system more equitable and more efficient. Durbin’s Resolving Extended Limbo for Immigrant Employees and Families (RELIEF) Act (S. 2603) exempts spouses and minor children from the cap, effectively doubling the number of green cards, and it creates a new temporary pool of green cards equal to the current number of backlogged petitions as of 2020 to be dispensed over five years. Paul’s Backlog Elimination, Legal Immigration, and Employment Visa Enhancement (BELIEVE) Act (S. 2091) eliminates the per‐​country limits, increases the overall limit on employment‐​based green cards from 140,000 to 270,000, exempts spouses and minor children of workers and investors from the cap, and creates a new uncapped category for shortage occupations, currently nurses and physical therapists. These reforms would effectively quadruple the available green cards in the EB-2/EB-3 categories.

Figure 7 compares the potential effects of the Fairness for High‐​Skilled Immigrants Act and current law and the RELIEF and BELIEVE Acts on the total EB-2/EB-3 green card backlog, including existing petitions and future petitions (projected based on total increase). The Fairness for High‐​Skilled Immigrants Act would only redistribute the backlog more fairly between nationalities—it would not reduce it. Under the Fairness for High‐​Skilled Immigrants Act as well as current law, the backlog would increase 160 percent by 2030, while the backlog would decline by 74 percent under the RELIEF Act. Only Paul’s BELIEVE Act would eliminate the backlog completely in 2026.


America’s employment‐​based immigration system is broken. More than 1 million immigrants are waiting for lawful permanent residence solely from outdated caps, and the current rate of increase in the backlog predicts that it will total more than 2.4 million by 2030. Per‐​country limits force Indians to bear nearly the entire burden of a broken system with more recent immigrants facing lifetime waits for green cards. About 205,000 could die in line even if they stay in line as long as possible.

Congress needs to address this emergency before the country loses out on hundreds of thousands of skilled workers who could contribute to America’s economy. Abandoning per‐​country limits would be a good first step, but it would be insufficient to prevent unsustainable waits for all immigrants. Congress also needs to increase the number of green cards dramatically to resolve this crisis.


Methodology for Figures 1–4

U.S. Citizenship and Immigration Services (USCIS) reported the number of principal petition beneficiaries (i.e., workers or investors) by category and birthplace. It provided a “multiplier” to estimate the number of derivative beneficiaries (i.e., spouses and minor children eligible for green cards by virtue of their relationships to the principal applicants). USCIS multiplies the multiplier by the number of principal beneficiaries to calculate the number of derivative beneficiaries.

USCIS calculated the multiplier by taking the number of derivative beneficiaries who received green cards divided by the number of principal beneficiaries who received green cards, as recorded on Table 7 of the Department of Homeland Security’s (DHS’s) 2016 Yearbook of Immigration Statistics. Figures 1–4 estimate these multipliers separately for spouses and children in the same manner as USCIS using DHS’s 2018 Yearbook of Immigration Statistics. These multipliers were multiplied against the number of principal beneficiaries listed in USCIS’s disclosure.21

Methodology for Table 2

No Changes Column. The backlog estimates in Table 2 are estimated by multiplying the number of principal beneficiaries listed in USCIS’s Freedom of Information Act (FOIA) disclosure by the derivative multiplier described in the methodology for Figures 1–4. The number of annual green cards issued by birthplace and category are estimated using the numbers from 2019. The number of years to process with no changes was calculated by dividing the backlog by the number of green cards issued in 2018. EB-2 and EB-3 were combined because every EB-2 applicant can refile under EB-3 (i.e., “port”) when those waits become shorter than EB-2. EB-2 requires an advanced degree; EB-3 requires a bachelor’s degree. This explains why the current wait times for EB-2 and EB-3 are already so close. EB-2 and EB-3 beneficiaries both waited about 10 years for the chance to apply for green cards in 2019.22

Deaths Column. The number of future deaths was calculated as follows:

First, the ages of adult backlogged beneficiaries—principal beneficiaries and their spouses—were estimated for each category according to the age distribution of H-1B visa holders extending their status in 2017. Most employment‐​based green card applicants start on H‐​1Bs, and those applying for extensions are likely to have very similar age distribution as pending employment‐​based applicants. USCIS only reports the H-1B age distribution in five‐​year increments, so the population at individual ages was estimated by dividing by five. For children, the age distribution originated from Table 8 of DHS’s Yearbook of Immigration Statistics, showing the age distribution of new legal permanent residents.23

Second, the probability of dying at each age was taken from the Social Security Administration’s life tables.

Third, the probability of dying at each age was multiplied by the backlogged population at each age to produce the number of deaths in a year.

Fourth, the backlogged population at each age was aged one year for each year into the projection.

Fifth, the number of deaths in each age group in the prior year was subtracted from the prior year’s population.

Sixth, the population of each individual age was reduced by the number of green cards issued in the prior year in proportion to that population’s share of the total population, based on the number of green cards issued in fiscal year (FY) 2019.

This process was repeated for each year until the initial population was reduced to zero. Deaths for each year were summed to produce an estimated number of total deaths, assuming all beneficiaries remained in the backlog. The number of years to process is the number of years that it took to reduce the initial populations to zero.

Abandonment Column. The rate of abandonment was calculated based on the number of approved petitions for EB-2 and EB-3 Indian nationals from 2009 to 2019. FY 2009 was taken as the starting point because the Department of State (DOS) was currently processing petitions from 2009 in November 2019, and Indian EB-2 and EB-3 petition beneficiaries had the longest time sample to review.24

First, the number of approved I-140 petitions for Indians filed in 2009 was reduced by the percentage of that year that had been allowed to apply for a green card/​immigrant visa as of November 2019 based on DOS’s “Visa Bulletin for November 2019.”

Second, the number of approvals for each year was reduced by an annual rate necessary to obtain the actual number of backlogged petitions in November 2019. This rate was determined to be 2.4975 percent. This rate was used for EB-1 to EB-4 categories. Steps 1 and 2 were similarly performed for EB-5 petition approvals from 2013 to 2020 to produce an abandonment rate for EB-5 petitions.

Third, the rate of abandonment was then applied to the population of backlogged beneficiaries in 2020 by age (as calculated above). Each age group was aged forward one year, and the number of abandoned petitions in the prior year was then subtracted from the population for each age in the prior year in the same manner as deaths. Abandoned petitions as well as deaths and green cards as calculated above were subtracted from each age group until the initial population was zero.

Fourth, beneficiaries turning age 21 were dropped from the backlog each year of the projection because the law only entitles children under age 21 to derivative eligibility for green cards. The number of green cards received in Table 2 was calculated by summing the projected number of green cards issued under this calculation (i.e., minus deaths and abandoned petitions).

Methodology for Table 3

USCIS’s FOIA disclosure does not provide the date of the approved petitions, so Table 3 estimates the distribution of backlogged beneficiaries based on the distribution of I-140 approved petitions filed from FY 2009 to November 2019. FY 2009 is currently the oldest petition without a current priority date in the EB-2/EB-3 line, so all backlogged petitions entered after that date. The model that includes deaths and abandonments used in Table 2 was then applied to each to estimate the dates they would be processed.25

Methodology for Figure 6

The Office of Foreign Labor Certification (OFLC) publishes the individual wage offers for all EB-2 and EB-3 immigrants whose employers are required to receive approved labor certifications. Figure 6 includes only the labor certifications that were certified. The wage offers can be annual, monthly, weekly, or hourly and can have a high or low range, so Figure 6’s projection comes from annualizing each wage range and then averaging the range for each individual. The countries reflect the birth country of the immigrant, not their citizenship. The data needed mild cleaning to correct for errors that listed yearly wages as paid on a more regular schedule.26

Methodology for Table 4

Table 4 uses the same model as Table 2’s projection that includes deaths and abandonment. It assumes that in 2020, the status quo will prevail for green card issuances because the Fairness for High‐​Skilled Immigrants Act includes a provision that states that anyone who applied for the date of enactment will receive the same treatment as before, and at any one time, there is more than a year’s worth of new petition beneficiaries. For 2021 and 2020, non‐​Indian, non‐​Chinese nationalities are guaranteed 10 percent of the green cards, and for all years after that, the remaining 2020 Chinese backlog was equally distributed across the number of years it would take for the 2020 Indian backlog to be processed.

The annual rate of increase in the total backlog was estimated for all beneficiaries by taking the monthly rate of increase in the backlog from April 2018 to November 2019 and multiplying by 12. Then, the gross increase—that is, the difference between the total backlog and the backlog in 2020—was distributed based on the percentage of petitions issued by birthplace in 2019.27

Methodology for Figure 7

The estimated increase in the total backlog under the Fairness for High‐​Skilled Immigrants Act comes from the last row of Table 4, estimated by the annual rate of increase in the backlog from April 2018 to November 2019. The increase in the number of green cards under the Resolving Extended Limbo for Immigrant Employees and Families and Backlog Elimination, Legal Immigration, and Employment Visa Enhancement Acts was subtracted from the estimated backlog. Both cases assume that 50 percent of EB-1 numbers go unused and spill over to EB-2 and EB-3 after 2022.

David J. Bier

David J. Bier is the immigration policy analyst for the Cato Institute’s Center for Global Liberty and Prosperity.