Today, Cato published a new compilation of 30 short essays by 15 authors including some of the leading immigration law experts in the United States that explains executive actions that could improve and expand the legal immigration system. It is titled: Deregulating Legal Immigration: A Blueprint for Agency Action. Going into this project, we had the following goals:

1. Each proposal goes beyond reversing Trump policies. Obviously, we support the wholesale and immediate reversal of nearly every Trump immigration policy since nearly all have been harmful, but we felt that the Biden officials will already have marching orders to reverse Trump policies, so there was more of a need for ideas to make the system better than it was before Trump. We hope that Trump policies get swept away quickly so officials can turn their attention to our ideas.
2. Each proposal helps legal immigrants navigate our bureaucratic, outdated, and restrictive system. We adopted this focus mainly because we feel that the Biden team has already fully thought-out proposals to address immigrants without legal status, and others were better equipped to flesh out those ideas. Thus, this compilation was about filling niche in the market of ideas for reforming the system, not giving every important idea for reform.
3. Each proposal provides the legal authorities for action. We have many ideas for improving legal immigration (I have laid them out briefly here and here), but it was imperative for us that these executive actions have strong legal foundations under current law. These ideas are all legally actionable based on existing law and precedent, and this volume gives the administration almost everything that they need to implement them.
4. Each proposal is concise and easily read by nonexperts. This last one was exceptionally important to us. Legal experts have spilled thousands of pages about problems with the legal system, but our goal was to provide a product that could be consumed by administration officials, advocates, and others who may not be experts on these matters but could take an interest in seeing these ideas across the finish line.
5. Each proposal is authored by a legal expert: It was important to us that this was not simply a Cato-only project. We wanted to introduce ideas from many of the leading immigration authorities in this country (as well as a few of my own). These included multiple past presidents of the American Immigration Lawyers Association, a former administration official, authors of the leading immigration law reference texts in the United States, and other well-known immigration experts.

I hope that we accomplished these goals. I believe this is the most comprehensive and concrete, but also most consumable and concise compendium of legal immigration ideas for President Biden. The Biden administration should act quickly to undo as much of the damage of the last four years as possible. But he should go further and deregulate the legal immigration system to the greatest extent that the laws allow. Table 1 lists the ideas we included and their authors. The entire paper is available here.

About 15 years ago, Benjamin Powell and I started working together on an academic paper about Somalia that the Journal of Economic Behavior and Organization eventually published. We corresponded entirely through email on the project while our other co-author Ryan Ford worked with Powell in person. It wasn’t until years later that we even met in person. But little did either of us know that our collaboration together would eventually result in a co-authored book. Today is the release date for our book Wretched Refuse? The Political Economy of Immigration and Institutions. You could be reading the kindle version within a minute of purchasing it.

This book isn’t our first research collaboration on immigration and the topic has only grown in importance since then. It’s the most hotly contested policy topic in political debate in the United States and most of the developed world although it’s been temporarily knocked aside by the pandemic. After COVID-19 subsides, immigration will reemerge as one of the big topics of debate again. Our book will help you understand one important, widely felt, and differently expressed argument against immigration.

First, let me throw in some backstory. In 2013, I wrote a blog post entitled “Immigration Does Not Decrease Economic Freedom” that very simply examined the relationship between stocks of immigrants and economic freedom in the United States. At that time, I was starting to hear different arguments against liberalized immigration by a few libertarians and many conservatives who were traditionally pro-market and pro-immigration. Rather than the standard “immigrants vote Democratic so they’re bad” argument that had been extant for years, a less-partisan version began to circulate that immigrants from countries with worse economic institutions would bring them with them and eventually degrade American economic institutions and thus reduce economic growth in the United States and kill the economic goose that lays the golden eggs of productivity that attract immigrants in the first place. In other words, government control of the economy was a disease that could spread from the developing world to the United States via the vector of immigrants, and the patient would be permanently impaired as a result.

This potential outcome was a real worry to me. There are many different theories in economics that try to explain why some places are richer. While most of them have some merit, the difference in economic institutions across countries is the best. By economic institutions, I mean the rules of economic exchange and their enforcement mechanisms. They include formal government rules like property rights and the courts but also informal cultural rules like sealing a deal on a handshake (or an elbow-bump) and that we’re reasonably confident that handing over a credit card to a merchant probably won’t result in theft.

Informal economic institutions are difficult to measure and I don’t have much confidence in big cross-country surveys that provide data to potentially do so, but Cato and the Fraser Institute have published a big index on economic freedom that allowed me to look at the formal economic institutions that really matter. I found a slight positive relationship between immigration and economic freedom or, at worst, no meaningful relationship. This was the first step toward writing Wretched Refuse? The Political Economy of Immigration and Institutions.

After I published that blog post in 2013, Benjamin Powell and many other excellent economists contacted me to discuss how to study whether immigrants impact economic institutions. The results were multiple peer-reviewed academic papers that began examining this topic in a deep way. And we weren’t alone as other economists started openly discussing how immigrants affect institutions and to begin to model it. Then Powell got the idea to write a book on this project, asked me to coauthor it with him, and the result is available today for purchase. To our knowledge, our book is the first of its kind on this specific topic.

Our book’s topic may seem obscure to some people. After all, with every debate over immigration raging why write a book about a niche issue in immigration economics? There are several answers to that, but the first is that we wanted to know if we’d kill economic productivity in the developed world if we ever succeeded. The standard economic case for free immigration is strong, but some very smart people were telling us that we were missing a very large negative externality. That seemed worthy of a book length examination.

Our book has three major sections. The first shortly summarizes the economic reasons for expanding legal immigration and introduces the argument that liberalized immigration would undermine American institutions. The second section analyzes cross-country evidence on how immigration affects economic institutions, corruption, terrorism, and culture more broadly. The chapter on culture focuses on whether immigrants lower general social trust and trust in specific institutions. It’s the chapter where I learned the most, especially about the economics of culture and immigration, and it influenced other independent research. The third section of the book looks at individual case studies on how sudden exogenous immigration shocks in Israel and Jordan improved economic institutions there as well as how more liberal immigration laws in the United States delayed the growth of government.

The book was a lot of fun to write, I was blessed with a fantastic coauthor, and I hope you enjoy reading it half as much as we enjoyed writing it. Please check out this interview that we had with Nick Gillespie at Reason to get a more detailed glimpse of what our book is about and this blog post by Bryan Caplan about it.

With President-elect Joe Biden set to assume the presidency on January 20, he has set the ambitious goal of reversing his predecessor’s immigration policies. The scope of President Donald Trump’s influence can be measured in terms of policies—the Migration Policy Institute has identified more than 400 changes since 2017—but it can also be measured in terms of people—immigrants not coming to America, not receiving permanent residence, or being denied employment authorization and status.

This post explores the trends in immigration approvals and denials that the new administration must reverse to fulfill its campaign promises.

Related Study

Deregulating Legal Immigration: A Blueprint for Agency Action

President Trump went to unprecedented lengths to restrict legal immigration in his one term—but the system was overly burdensome and expensive long before him. President-elect Joe Biden will have the opportunity to reverse these actions and deregulate this system. Here is a compendium of 30 proposals by 15 authors—including several of America’s leading immigration law experts—to help the Biden administration operate the immigration system as openly and efficiently as the law allows.

Read the Reforms

738,000 Fewer Immigrants From Abroad

Immigrant visas are issued to prospective legal permanent residents by the Department of State (DOS), usually following a petition by a U.S. sponsor approved by U.S. Citizenship and Immigration Services (USCIS). Before President Trump, 86 percent of immigrant visas were issued to the immediate family members of U.S. citizens and legal permanent residents—spouses, children, parents, and siblings (and their spouses and minor children). Fiancés and their children also receive a K nonimmigrant visa that provides a path to permanent residence, so they are also included in Figure 1.

Figure 1 shows immigrant visas issued by month under the Trump presidency compared to the average for the last fiscal year of the Obama administration (2016). Immigrant visa issuances were down 83 percent in October 2020 compared to FY 2016. Even before the pandemic, permanent immigration from abroad had declined by about 24 percent. Overall, there were 738,857 fewer immigrant visas issued as of October 2020 than there would have been if the rate of issuance under President Obama had continued under President Trump.

The main cause of the drop in FY 2020 was DOS’s closure of consulates and refusal to waive in-person visa interviews. But another reason is that USCIS started denying petitions for immigrant visas at a much higher rate. Figure 2 shows the USCIS denial rate for immigrant petitions (which can also be the basis of an adjustment of status application). The denial rate was higher in every quarter under the Trump administration than the annual average in FY 2016. The denial rate basically doubled from about 8 percent in FY 2016 to 16 percent in the third quarter of FY 2020. This translates into 72,856 more denials for families, employers, and others seeking to start the immigrant visa process.

Even if USCIS approves their family member’s petition, immigrant visa applicants must still receive approvals from DOS’s consular affairs. Unfortunately, DOS has not provided any data for denials in FY 2020, but even by FY 2019, the denial rate had again almost doubled from 13 percent to 21 percent. Assuming the denial rate in 2020 remains at least what it was in 2019 (and it will likely continue to climb), the Trump administration will have produced about 101,120 additional immigrant visa denials than would have occurred if the denial rate remained the same. These denials undoubtedly intimidated other applicants into choosing not to apply at all.

291,000 Fewer Refugees

The U.S. Refugee Admissions Program resettles refugees to the United States if they possess a well-founded fear of persecution based on a protected ground. Unlike immigrant visas, the most dramatic drop in refugee admissions occurred almost immediately upon President Trump taking office. As of September 2020, the number of refugee admissions had fallen by 67.5 percent compared to the average for 2016. We can be sure this low admissions rate will continue through January, indicating the United States will have admitted about 291,393 fewer refugees than would have come had the admissions stayed at 2016 levels.

246,000 Fewer Adjustments of Status to Permanent Residence

An adjustment of status application requests that USCIS adjust the applicant’s status legal permanent residence. It is an application filed by immigrants inside the United States in lieu of traveling abroad and obtaining an immigrant visa at a consulate. In normal years, about half of new permanent residents adjust their status in the United States. Unlike immigrant visas, only half of adjustments are family-based, while the rest are employment-based as well as refugees, asylees, and applicants in other humanitarian categories.

Figure 5 shows the number of approved adjustment of status applications by quarter. As of the third quarter of FY 2020 (March to June), the number of approvals had fallen by 56.2 percent compared to the average in FY 2016. Unlike immigrant visas and refugees, the drop started late in Trump’s term in the 4^th quarter of FY 2019 and the major decline occurred in the third quarter of FY 2020. If the rate of approvals remains the same until January, about 246,313 fewer applicants will have received permanent residence under the Trump administration than would have received it had approvals remained at FY 2016 levels.

Like immigrant visas, the main cause of the decline in adjustments is that USCIS closed field offices and generally refused to waive in-person interviews during the pandemic, bottlenecking the process. However, another major contributor is increased USCIS denials—both of petitions (see Figure 2) and adjustment applications themselves (Figure 6). Despite adjustment approvals remaining as high as before, the denial rate has been higher in every quarter of the Trump administration than the average in FY 2016. The denial rate has increased almost threefold during the last four years. By January, there will likely end up being about 70,256 more denials of adjustments of status during the Trump years as there would have been had the denial rate remained at the 2016 rate.

287,000 Fewer Nonimmigrant Work and Cultural Exchange Visas

The United States permits foreign workers to enter the United States on nonimmigrant (temporary) visas—either work visas or cultural exchange visas—issued by DOS. Exchange visas allow the visa holder to work as part of the cultural experience, and nearly all choose to do so. The foreign workers’ spouses and minor children may join them in the United States as well. While temporary, nonimmigrants can often renew their statuses in the United States. In most cases, temporary workers (e.g., H‑1B, H‑2A, H‑2B, L‑1, etc.) may not enter the United States without USCIS approving an employer petition on their behalf.

Figure 7 shows the number of nonimmigrant work and cultural exchange visas issued by month. In general, work visa issuances were higher than they were in FY 2016 (monthly data are not available), but in October 2020, work and cultural exchange visa issuances were down 69 percent from October 2019, largely thanks to consulate closures and refusals to waive interviews by DOS. There would have been 287,003 more work and cultural exchange visas issued during the Trump presidency if the FY 2016 visa issuance rate continued.

Work visas grew before the pandemic because labor demand grew, not because USCIS or DOS adopted more favorable processing standards. From FY 2016 to FY 2019, DOS’s visa denial rate grew from 4.4 percent to 6.5 percent, causing 34,389 more denials than would have occurred without the increase. More dramatically, Figure 8 shows that the denial rate for employer petitions for nonimmigrant workers by USCIS more than doubled from 7 percent in FY 2016 to almost 15 percent at the midpoint of FY 2020. This translates into 113,000 additional denials of nonimmigrant workers.

These denials reflect a concerted effort by USCIS to make hiring foreign workers much more difficult. USCIS has also nearly doubled the rate at which issues “requests for evidence” (RFEs) in response to employers’ petitions. RFEs amount to an initial denial in which USCIS demands more evidence to support the petition. The USCIS Ombudsman has found on multiple occasions that USCIS commonly issues RFEs for information already provided or based on erroneous understandings of the law. Many lawyers assume that USCIS uses RFEs to delay having to adjudicate the application. These delays are always costly for employers but are particularly damaging for seasonal H‑2 employers because the jobs are of a short, defined period.

96,000 More Denials of Asylum

Asylum provides refugee status to people who meet the refugee definition (well-founded fear of persecution based on protected class). Under the asylum statute, asylum applicants may apply for asylum anywhere in the United States. Asylum applicants are processed either as “affirmative” applications through USCIS or as “defensive” applications made as part of removal proceedings through the Justice Department. Defensive asylum applicants include those who crossed illegally to request asylum as well as those who applied at ports of entry—either at airports or the land border crossing points.

Defensive asylum grants increased under the Trump administration from 8,648 in FY 2016 to 14,507 in FY 2020, and affirmative asylum grants also initially increased from to 9,538 to 19,945 from FY 2016 to FY 2019, but so far in FY 2020, asylum grants have fallen back to almost to where they were in 2016 on pace for about 9,700. Once again, these increases reflect rising requests for asylum and greater processing of applications, not more favorable processing.

Figure 10 shows affirmative and defensive asylum denial rates by quarter. The defensive denial rate has more than doubled the average for FY 2016, and the affirmative denial rate has grown significantly as well. Three-quarters of affirmative asylum applications adjudicated by USCIS in FY 2020 were denied, compared to 57 percent in FY 2016. These increases produced 25,687 more affirmative asylum denials and 70,774 more defensive denials than would otherwise have occurred had the FY 2016 rates prevailed.

These statistics only reflect those whose applications made it to a final adjudication. Most defensive asylum applicants must first clear an initial screening by USCIS to identify if their stated fear of return is “credible.”

Figure 11 shows the credible fear denial rate for 2017 to 2020. Credible fear denials exploded when USCIS implemented its categorical ban on asylum for almost anyone who transited through another country before reaching the United States. A district court enjoined the policy on June 30, 2020 after which time credible fear was established at a higher rate (though a separate policy still allowed Border Patrol to remove those who crossed illegally). Overall, USCIS found credible fear in 28,545 fewer asylum cases than if it had maintained its 2016 average rate throughout Trump’s 4 years. Many more were persuaded to not even bother to apply.

698,000 Fewer International Students

Foreign students can travel to the United States to study and may also work before or after graduation under the Curricular Practical Training and Optional Practical Training programs. Their spouses and minor children also can travel and live with them. The number of international students on F or M visas crashed 75.5 percent from FY 2016 to FY 2020. The largest drop occurred in FY 2020 during the pandemic when the administration hesitated over allowing students to enter for online-only classes, and DOS closed consulates and refused to waive interviews. The United States saw 697,625 fewer student visa holders as a result of the declining visa issuances.

Nine Million Fewer Tourist and Business Traveler Visas

Foreigners can also come to the United States as tourists or business travelers on a B visa. Even before COVID-19, every month but one saw fewer tourist or business traveler visas issued than the monthly average in FY 2016. The average in the 12 months before the national emergency in March 2020 saw 21 percent fewer tourist and business traveler visas issued than in FY 2016 (Figure 13). As of October 2020, business and tourist visa issuances were down 92 percent compared with FY 2016. Overall, nearly 9.1 million fewer tourist and business traveler visas were issued during Trump’s four years than would have occurred if the FY 2016 rate had continued. The drop in tourism and travel in 2020 was caused both by traveler avoidance, other countries’ exit policies, President Trump’s entry bans, DOS’s consulate closures, and DOS’s refusal to issue interview waivers.

7 Million Pending Applicants

Beyond denials and fewer approvals stands an almost unfathomable backlog of 5.7 million applicants for immigration benefits at USCIS—everything from travel and work authorizations to adjustment to permanent residence to citizenship. Until 2017, the backlog of pending applicants was a little less than 2 times the quarterly number of new applications—with changes in the receipts reflected in changes in the backlog (Figure 14). But in 2017, the backlog spiraled out of control despite declining applications. As of the third quarter of 2020, the backlog is now 3.5 times the number of new applications. The Justice Department’s immigration courts also have another 1.3 million pending cases, more than double the court backlog in FY 2016.

The USCIS and DOJ backlogs effectively deny immigrants and Americans their legal rights by delaying their ability to receive status, work, or travel authorization or preventing them from sponsoring their employees or family. According to USCIS, the average time that all pending USCIS applications have waited has grown from 4.7 to 7.3 months from FY 2016 to FY 2020—with the share waiting more than 7 months shooting up from 4.7 percent in FY 2016 to 50.2 percent in FY 2020. For example, adjustment of status (green card) applications increased from 6.8 to 11.4 months. This is just the average time that applications—including those recently filed have spent waiting—but the average time to actually finish adjudicating an application is much longer.

Conclusion

By taking FY 2016 as its baseline, this analysis fails to capture the lost opportunity that increased legal immigration would have brought over the last four years. Nonetheless, the numbers are clear that President Trump’s immigration policies have irreparably harmed millions of immigrants and Americans. But a new administration will need to go beyond simply reversing his policies to avoid having their effects felt for years to come. New policies would need to allow immigrants to reopen cases wrongfully denied and issue visas and status to them. They should streamline processing these applications to avoid further delaying others. If President Biden’s goal is to rectify the harm of the last four years, he must act boldly and aggressively to fix the numerous problems of his predecessor.

In the wake of the 2020 election, some Republicans took heart in the apparent shift of some Hispanic voters toward Trump relative to 2016. As I wrote elsewhere, the size of this shift is probably not large (bad exit poll data obscures this) and it still resulted in less Hispanic support than was earned by George W. Bush in 2004. Ethnic attrition also makes it difficult to judge the ethnic and racial support for presidential candidates as the descendants of immigrants, especially if born to an inter-ethnic or inter-racial couple, are much less likely to self-identify as Hispanic, Asian, or with other racial or ethnic origins. Thus, Trump’s support among second, third, or higher generation Hispanics could be even higher than the ethnic responses to exit polls suggest.

Another way of looking at the shift in support for Trump is to examine county level voting outcomes. Trump did win some counties in the Rio Grande Valley of Texas that he would not have won without a significant shift in the Hispanic vote in his direction. But Hispanics are only part of the immigrant population in the United States and, surprising to many, the majority of Hispanics living here are not immigrants, but native-born Americans. We should therefore expect that, over time, nativist policies and talking points would negatively affect Hispanic voters less and less as they and their descendants more fully assimilate into American culture and think of themselves primarily as patriotic Americans. There are also many immigrants from Asian, Africa, and Europe who were also affected by Trump’s immigration policies who get lost in the discussion over Hispanic voting patterns.

This blog post looks at the change in the county level vote share for President Trump from 2016 to the 2020 election, with a particular focus on the share of the county population who are naturalized citizens in all counties in the United States. Naturalized citizens are immigrants who have become American citizens and who can therefore vote in American elections. To the extent that immigrants would or could directly affect American elections, naturalized citizens are the voters who would do so. The 2020 county level election results data come from Fox News, the 2016 election results come from the MIT Election Data Lab, and the demographic data are from the 2014–2018 ACS 5‑year estimates.

Figure 1 shows the relationship between the 2016 and 2020 vote share for Trump on the county level. The 2016 and 2020 Trump vote share is correlated with a coefficient of +0.96 that is significant at the 1 percent level. In other words, a county’s vote share for Trump in 2016 is highly correlated to a statistically significant extent with the county’s vote share for Trump in 2020. Data points further to the right of the dashed line show more of a shift toward Trump, data points to the left show more of a shift toward the Democrats. It’s also important to realize that this shows a shift in the share of the vote, so although it looks like the shift for counties toward Trump was very large, those shifts were in some smaller population counties and obscure the overall shift toward Biden. For instance, the biggest county level vote share shift was in Graham County, Arizona where Trump’s vote share increased from 40 percent in 2016 to 72 percent in 2020. But Graham County is small, with a population slightly under 38,000 in our sample, so large shifts in small counties didn’t make the difference state-wide in Arizona. So far, Figure 1 is not surprising.

Figure 1

County Level Vote Shares for President Trump, 2016 and 2020

Sources: Fox News, MIT, American Community Survey.

Figure 2 shows the relationship between the shift toward Trump from 2016 to 2020 and the naturalized immigrant share of a county’s population. The correlation coefficient is a statistically insignificant ‑0.038. In other words, there was no meaningful shift in the county-level immigrant citizen vote share. Trump did neither worse nor better than he did in 2016. Interestingly, this statistically insignificant results can be entirely explained by Miami-Dade County, Florida. Excluding Miami-Dade County from the regression shows that counties with more naturalized citizens were more likely to be less supporting of him in 2020 everywhere else in the country with a correlation coefficient of ‑0.067 that is significant at the 5 percent level. Without the naturalized American voters in Miami-Dade county, who account for 30 percent of that county’s population, counties with new Americans would have been more likely to vote against Trump in 2020 than they did in 2016.

Figure 2

Naturalized Share of the Population and Vote Share Shift Toward Trump

Sources: Fox News, MIT, American Community Survey.

The naturalized immigrant population in the entire country is just over 7 percent. Looking entirely at the 121 counties where the naturalized immigrant share of the population is at or above the nationwide level shows that they shifted slightly toward Trump in 2020, to 39 percent up from 37 percent in 2016 (Figure 3).

Figure 3

County Level Vote Shares for President Trump in Counties with High Naturalized Immigrant Population Shares, 2016 and 2020

Sources: Fox News, MIT, American Community Survey.

There are a few lessons here. First, counties that voted for a candidate in one year are likely to vote for that candidate the next year and to a similar extent. No surprise there. Second, counties with naturalized American citizens didn’t become more anti-Trump if Miami-Dade is included. Third, the big shift toward Trump in Miami masks the nationwide shift against Trump in counties with more naturalized Americans. Excluding Miami-Dade County shows a general shift against Trump in counties with more naturalized Americans that is significant at the 5 percent level.

Caution is always warranted in interpreting county-level electoral results like these, so please don’t confuse this analysis with one that uses microdata, but it sure looks like the anti-socialist naturalized American voters in Miami-Dade voted differently in 2020 than other naturalized Americans. They lived in a county that still voted for Biden but that swung hard toward Trump by over 12 points.

The Senate passed an amended version of the Fairness for High Skilled Immigrants Act (S. 386/H.R. 1044). The bill has already passed the House of Representatives on a massive 365–65 vote. Since then, Sen. Mike Lee (R‑UT) repeatedly attempted to pass the bill on “unanimous consent” under which any member can object—which led to deals with Senators Chuck Grassley (R‑IA), Rand Paul (R‑KY), David Purdue (R‑GA), and Dick Durbin (D‑IL) that amended the House bill. Sen. Rick Scott (R‑FL) who was the latest member to object lifted his hold last night, allowing final passage.

E-Publication

Reforming the Immigration System: A Brief Outline

With the election of a new president, Congress has the opportunity to correct the two fundamental flaws that plague the current U.S. system. First, the system is too restrictive, and second, the system is too inflexible to adapt to new economic or social conditions, allowing small problems to build into national crises. Here is a compendium of 52 reform proposals, both for enforcing the law, and for a less restrictive and more flexible immigration system.

Read the Reforms

The Senate version is now substantially different from the House version with deleterious provisions to which the House Judiciary Committee has already voiced opposition, but it has also committed to finding common ground to resolve.

What’s in the Fairness for High Skilled Immigrants Act, December 2020 version?

Green card reforms:

1. Phases out employment-based per county limits on green cards: The main purpose of the legislation is to treat all employment-based immigrant visa applicants on a first-come, first-served basis without regard to birthplace. Under current law, immigrants from no single birthplace can receive more than 7% of the total number of immigrant visas or green cards issued in a year unless they would otherwise go unused. The effect of this provision is that while Indians are half the skilled employer-sponsored applicants, they receive just 10 percent of those green cards and—as a result—are nearly 90 percent of the backlogged applicants.
   • Comments: The House bill is the same. Basically, this provision is the only reason the bill has made it as far as it has. The discrimination against Indian skilled immigrants mean, as I’ve estimated before, that new Indian green card applicants will almost certainly never receive green cards in their lifetime. More than 200,000 of the 700,000 Indians in line will likely die before they receive their green cards. The fact that other immigrants almost immediately receive their green cards makes the system massively unfair and is already causing skilled workers to leave the country. Indians and Chinese—the only two significantly backlogged applicants—also receive wage offers significantly higher than those for the average applicant from other countries.
2. Provides for an 11-year phase out period: The bill’s green card changes will only take effect on October 1, 2022. For the EB‑2 and EB‑3 categories for non-executive level employees of U.S. businesses, the bill guarantees immigrants which are not from the top two origin countries (India and China) a certain percentage of the green cards for 9 years: year 1 (30%), year 2 (25%), year 3 (20%), year 4 (15%), years 5 and 6 (10%), and years 7 through 9 (5%). No more than 25 percent of these “reserved” green cards can go to immigrants from any single country. No more than 85 percent of the other “unreserved” green cards can go to a single country (India). In addition, a minimum of 5.75% of all EB‑2 or EB‑3 green cards will go to immigrants from these non-top 2 countries for 9 years prioritizing spouses and minor children of immigrants already in the United States and immigrants awaiting visas abroad. It’s unclear if the 5.75% counts toward the prior reservation or must be in addition to it. This ambiguity means it is not possible to say with certainty how long it will take for the current backlog to be processed under the bill.
   • Comments: The House bill would have taken effect immediately and only contained a 3‑year phaseout with set asides for non-Indian or Chinese applicants of 15% in the first year and 10% in the next two years. Eleven years is an incredibly long time to continue a system based on birthplace discrimination. It will probably take about 13 years to process all existing Indian applicants under this system, while new applicants continue to take priority. If only applicants who are not currently in the United States received priority, that would be a rational basis for discrimination since those already in the United States are already benefiting more from the U.S. immigration system than those abroad. But these provisions continue the discrimination that the bill is designed to eliminate for a decade (albeit to a lesser degree).
3. Guarantee for nurses and physical therapists: The bill carves out 4,400 EB‑3 green cards (11% of the category) for nurses and physical therapists—defined as “shortage occupations”—for 7 years. The spouses and minor children would not count against this limit but would still receive green cards at the same time.
   • Comments: The House bill has no similar provision. This essentially creates a temporary new category for legal immigrants who DOL deems to be in short supply. I have no problem with this. It is as arbitrary as the rest of the employment-based categorization scheme and does not involve birthplace discrimination.
4. Caps H‑1B visa holders and H‑4 visa holders (or those who held H‑4 status in the last 2 years) to no more than 70% of all employment-based green cards during the first 9 years after implementation and 50% for all subsequent years.
   • Comments: This is a way of continuing discrimination against Indians indirectly because Indians are 70 percent of H‑1B visa holders. Moreover, nearly all of the backlogged immigrants in the green card queue from India are working on the H‑1B visa. This provision undermines the purpose of the bill and makes little sense as an anti-H-1B measure because it forces H‑1B holders to remain on the temporary status longer than they would otherwise. We don’t know the exact breakdown of the status of those in the green card backlog from India, but it’s likely at least 90 percent H‑1B visa holders, so this will delay the receipt of green cards to backlogged Indians. If not for the backlog, the effect would not be very great. About 70 percent of EB‑2 and EB‑3 green card applicants in 2019 were on the H‑1B visa based on Department of Labor’s labor certification data. It is likely less than this in the EB‑1 category (multinational executives, those with extraordinary ability, etc.) who have other temporary visa options available or come from abroad, and there probably are very few in the EB‑4 special immigrant and EB‑5 investor categories, so assuming only 50 percent of EB‑1 is on H‑1B visas, the number of new H‑1B green card applicants is probably only slightly higher than 50 percent. That said, it would hamstring any increase in that program.
5. Raises the family-sponsored per country limit on green cards to 15 percent: Family-sponsored preference categories also currently have a 7% limit on green cards for immigrants from individual birthplaces. The bill more than doubles that limit to 15% increasing in absolute terms from about 15,820 to 33,900. This will primarily benefit long-backlogged immigrants from Mexico and the Philippines, but also India and China.
   • Comments: The House bill is the same. This provision does not go as far as the employment-based provision in ending discrimination based on birthplace in the family-sponsored system, but it is a significant benefit to long-backlogged, family-sponsored immigrants who are mostly waiting for immigrant visas abroad.
6. Bars adjustment of status to all Chinese “affiliated” with the Chinese Communist Party: The new language (presumably proposed by Sen. Scott) requires DHS to “not adjust status of any alien affiliated with the military forces of the People’s Republic of China or the Chinese Communist Party” (CCP). This is similar to the existing ground of inadmissibility in 8 U.S.C. 1182(a)(3)(D) for members or those “affiliated with” any communist party anywhere. However, the existing ground of inadmissibility has exceptions for involuntary membership, past membership, or close family members. However, the current ground applies to both adjustment of status in the United States as well as consular processing abroad. Effectively, this provision requires all Chinese immigrants to apply for immigrant visas at consulates abroad.
   • Comments: This is just more de facto national origin discrimination. Most Chinese have no ideological connection to the CCP even if they join it. The main reason to have joined the party is that it facilitates promotions, especially within government or state-owned enterprises but overall. Lotus Yuen of The Atlantic has called membership the “ultimate resumé booster” in China. It can also allow Chinese to avoid direct state persecution. While the government obviously has an interest in stopping actual espionage, this ban is overbroad. The United States should want communists to experience the superiority of the U.S. system and encourage defectors from communist China. The United States has benefited greatly from Chinese immigrant innovators in science, technology, and medicine, and this ban would push inventors back toward the communist regime. That said, because it only applies to adjustment in the United States, the practical effect amounts to an expensive inconvenience rather than an outright ban.

Adjustment of status from temporary visa to green card

7. “Early filing” (H‑1B lite status): Allows backlogged temporary workers to receive a separate, limited, 3‑year, renewable employment authorization (apart from their underlying status) 2 years after their employer petition was approved by filing an adjustment of status to legal permanent residence application (i.e. green card application) prior to a green card number being available under the caps. You can credit Sen. Durbin with this provision. Currently, anyone whose adjustment of status application is pending for at least 180 days can receive an employment authorization document (EAD). This EAD allows them to work for any employer that they want while remaining in line based on the original employers’ petition. The original employer’s petition remains valid so long as they work in “the same or a similar occupational classification.” However, you currently cannot apply for legal permanent residence prior to a green card or immigrant visa number being available. The bill states that these new “early filers” could also receive this same authorization. However, the bill adds new requirements for this authorization. The job would have to have wages “commensurate with” those for the employer’s similarly situated U.S. workers in the area.
   
   
   If the employer had fewer than 2 such employees, it would have to attest that they were similar to the wages for similar U.S. workers in the area. The worker would have to file a Confirmation of Bona Fide Job Offer or Portability with a request for employment authorization. The employment authorization would last for three year increments with renewals, which is better than the adjustment of status EAD available now (which is only a 1 or 2 years). The worker would also have to provide a signed letter from their employer with the required attestations. A Confirmation of Bona Fide Job Offer or Portability would need to be filed (again if necessary) within 12 months of the green card application being adjudicated. If the Confirmation of Bona Fide Job Offer or Portability was deemed not to meet the requirements, the green card application would be denied. The minor children and certain spouses of temporary workers would also not benefit from this provision. The cost would be $2,000 for each Confirmation of Bona Fide Job Offer or Portability, in addition to the cost of the adjustment of status green card application. Half the fees would go to immigration adjudications and half into the general fund of the U.S. Treasury.
   • Comments: The senators have made this provision about as watered down as they can get it, but it is still the most important unequivocally positive change from the language that the senators have added. It would make it easier for H‑1B workers to change jobs. Currently, H‑1Bs stuck in the backlog have to renew every single year, which is costly and problematic if the government decides to readjudicate the underlying H‑1B petition. It would also allow other temporary workers, such as those on Optional Practical Training, to extend their status when they otherwise would not be able to, potentially enabling them to avoid having to obtain an H‑1B at all.
8. Prevents “aging out” of children of temporary workers in the backlog who have filed an adjustment of status application under the early filing provision. Currently, a dependent child of an H‑1B worker loses their status on their 21^st birthday. They also lose their eligibility for a green card at the same time. The bill would provide them both a status past their 21^st birthday.
   • Comments: This is an unambiguously positive provision.

H‑1B high skilled temporary worker reforms (none in House bill)

9. Requires the posting of H‑1B jobs for new H‑1Bs on government website for 30 days. If the Department of Labor (DOL) cannot get the website up and running within 180 days, the bill allows just a 30-day extension. If the website still cannot work, the H‑1B program could not permit additional H‑1B applications.
   • Comments: This seems like it is risking a lot for DOL to create a working website in less than a year. Forcing employers to advertise positions that may or may not actually be available makes little sense.
10. Bans advertising only to H‑1Bs.
    • Comments: This provision is not unreasonable.
11. Bans recruiting primarily H‑1Bs.
    • Comments: Unlike the prior ban on advertising, this provision undermines a major purpose of the H‑1B visa, which is to allow employers to hire workers for specialty positions. This tells businesses that they cannot simply recruit and hire a specific foreign worker or workers who they believe will fill whatever niche they need. Moreover, the “primarily” implies that companies would have to spend at least 50 percent of their time recruiting U.S. workers, even if they had already decided that a specific noncitizen was the person that they wanted. It treats the H‑1B program like the lesser-skilled H‑2 programs where the main purpose of the hire is labor, not skills. Moreover, it would greatly harm businesses that already employ the worker under a different visa category (L‑1, F‑1, J‑1, etc.).
      
      
      Employers could probably evade this requirement by making the job descriptions so demanding that only an existing employee could fill the position, as they commonly do under the permanent labor certification process. But why do the authors think it helps U.S. workers to create a bunch of sham job advertisements? The Justice Department’s recent Facebook lawsuit highlights the perils of handing such vague language to any administration. This concern is compounded because the bill also allows DOL to troll through companies’ files whenever it wants without any underlying complaint from an employee or U.S. worker.
12. Requires providing every W‑2 for every H‑1B worker employed by the company over whatever period DOL wants.
    • Comments: This provision would impose a significant administrative burden for no upside. The H‑1B process is already too time-consuming and expensive.
13. Bans hiring new H‑1Bs if an employer has more than 50% of its workforce on H‑1Bs or L‑1s (for skilled intracompany transfers from abroad).
    • Comments: As far as I can tell, only two large companies come close to fitting this profile: Cognizant (49.999%) and Tech Mahindra (50.3%), though a few others may be close if H‑1B visas were more readily available. The provision sets a dangerous precedent that H‑1B-heavy companies should be legally discouraged. But it’s unclear what the purpose of this restriction is, except to target certain companies that specialize in certain tech services to the benefit of others who are more widely diversified. Even if the requirement was only 25 percent, it would only force Cognizant and other specialized companies to sell or merge with a larger company with more employees, not change any business practices or cease hiring H‑1B workers.
14. Requires DOL to charge a fee for H‑1B labor condition applications (LCAs) to cover the cost of processing.
    • Comments: The bill also says that the fee could be used for “administration, oversight, investigation, and enforcement.” If the purpose of the fee is to cover the cost of the application, that’s reasonable. If the purpose is to force compliant employers to cover the costs of DOL actions against noncompliant employers, that’s unfair. This authority should be at least clarified.
15. Bans B‑1 temporary business visas for anyone who would “normally” be classifiable as an H‑1B:
    • Comments: This provision is unambiguously negative. Foreigners visiting the United States “temporarily for business” can receive a B‑1 visa if they are “participating in business activities of a commercial or professional nature in the United States.” The State Department has stated since the 1960s that in cases where a person who could qualify as an H‑1B is only coming for a short period, they are “more appropriately” classified as a B‑1 so long as they are paid from sources abroad. In a proposed rule, the comment period of which ends in December 21, the State Department is proposing its own reversal of this policy. Presumably this legislative provision is also intended to stop this practice, though the word “normally” adds some ambiguity. The “B‑1 in lieu of H‑1B” option is important because there is no other option specifically for skilled professionals on short-term assignments, especially those that come up suddenly and need to be completed quickly.
      
      
      U.S. businesses contracting with foreign companies, foreign multinationals investing in the United States, or foreign companies without a physical presence in the United States use this option, but it’s unclear how widespread it is because the State Department doesn’t separately record B‑1s-in-lieu-of-H-1Bs from the total 38,000 B‑1s. In 2010, however, the State Department stated that the consulates in India (the largest source of H‑1Bs) made “fewer than 1,000” such grants against nearly 60 times as many H‑1Bs. Nonetheless, this rule directly restricts legal trade, travel, and employment to no benefit to the United States.
16. Requires employers that retaliate against people who “reasonably believe” are disclosing evidence of an H‑1B violation to pay backpay.
    • Comments: This is an extension of current law prohibiting employers from retaliation by explicitly requiring them to pay backpay.
17. Requires DOL to review H‑1B LCAs for “fraud or misrepresentation” rather than only for “completeness and obvious inaccuracies”.
    • Comments: This undermines the type of expedited review that LCAs receive. As soon as adjudicators must undertake a more substantive review than completeness and obvious inaccuracies (such as, an internal inconsistency), the review will add significantly more time and expense to an already expensive and time-consuming process.
18. Requires employers pay at least the actual wages paid to similar U.S. workers in the local area.
    • Comments: This provision extends a current provision of the law to state the “actual wages” must be based only on wages of workers in the area of intended employment. There is also a slight tightening in how it defines similar U.S. workers from “similar experience and qualifications” to “substantially the same duties and responsibilities.” In some cases, using only U.S. workers in one area might raise the wage in some cases, while it might lower the wage in other cases. Again, narrowing the workers to those with “substantially similar duties and responsibilities” would have the same ambiguous effect.
19. Vastly expands DOL audit and investigation authority:
    • Allows DOL to conduct compliance surveys or annual audits of any H‑1B employer.
    • Requires audits of anyone with 100 H‑1B workers if more than 15% of their employees.
    • Allows investigations based on anonymous sources not in the form of a complaint from workers or other harmed parties.
    • Allows DOL to audit or investigate based on information in an LCA.
    • Eliminates the requirement that the DOL secretary personally certify that reasonable cause exists for an H‑1B investigation.
    • Removes the 60-day time limit on investigations.
      • Comments: Currently, H‑1B audits are based only on complaints or other verified, nonanonymous sources that come to the DOL from people DOL knows would have knowledge of an H‑1B violation. According to DOL, this latter authority had never been used as of this year, so H‑1B audits have exclusively been based on complaints. Congress imposed these restrictions because it wanted to limit the authority of DOL to conduct meritless investigations. These provisions would allow DOL to target employers for audits without any reason to believe a violation has occurred. This is yet another burden in an already burdensome and expensive process.
20. Eliminates the protection from penalties for employers that made a good faith effort to follow the rules or that underpaid employees based on use of a prevailing wage methodology based on industry standards.
    • Comments: These harmful provisions are replaced with a benignly labeled “information sharing” provision on page 19.
21. Triples the fines for H‑1B LCA violations. Fines increase from $1,000 to $3,000 for non-willful violations, from $5,000 to $15,000 for willful violations, from $35,000 to $100,000 for displacement of U.S. workers.
    • Comments: Adjustments for inflation since 1998 would not quite double the fine amounts, so these increases are clearly intended as more than an update to outdated statutory figures.

One of the most frequent and less serious criticisms that comes across my desk is that immigration is bad because non-citizens vote illegally in such large numbers that sway elections. A new report by James D. Agresti, pushed by some news outlets, argues that the number of noncitizens who illegally voted in 2020 substantially increased Biden’s vote share but did not affect the outcome of the election. It has been illegal for non-citizens to vote for federal elected officials since 1996, so these noncitizen voters would all be breaking federal law. Is the Agresti paper reliable? Are large numbers of noncitizens voting in federal elections to such an extent that several states voted for Biden as a result?

No, but to understand why you have to follow how the Agresti paper arrived at its conclusion. The Agresti report relies on a peer-reviewed academic paper by political scientists Jesse T. Richman, Gulshan A. Chattha, and David C. Earnest that was published in 2014 that estimates the rate at which noncitizens voted for president in 2008. Their paper relies upon responses to the Cooperative Congressional Election Study (CCES) for the 2008 election that found a substantial proportion of noncitizens voted in that year. The Agresti paper combined two figures from the Richman, Chattha, and Earnest paper to get their primary estimate that 15.8 percent of non-citizens voted in 2008. Agresti then apples that 15.8 percent rate to the non-citizen population in swing states in 2020 to reach their conclusion.

The big problem, as explained in two succinct pieces, is that non-citizens voting illegally is a small subset of a small population of Americans measured in the CCES survey. In the CCES survey, as in any survey, a certain number of respondents click the wrong box. Thus, some respondents will incorrectly click that they are non-citizens by accident and that they voted. Or they will make any number of other errors. This general problem is called measurement error and it afflicts every survey. These errors are common in surveys, but if it surveys enough people and there isn’t a tragic flaw in design that causes large numbers of people to make the same error, then it doesn’t matter much for the final result.

The problem is that the authors focused on a small number of non-citizens in a very large survey that likely accidentally said they were noncitizens who voted when they were really citizens who voted. The CCES survey asked about 20,000 people how they voted and about 19,500 of them said that they were U.S. citizens. Since the CCES is about federal elections, it oversamples citizens who can vote and under sample non-citizens who can’t vote. In fact, the number of reported non-citizens in the CCES survey who said they voted in a federal election is just about exactly the number who should have misidentified themselves as non-citizens in such a large survey:

This problem arises because the survey was not designed to sample non-citizens, and the non-citizen category in the citizenship question is included for completeness and to identify those respondents who might be non-citizens. We expect that most of that group are in fact non-citizens (85 of 105), but the very low level of misclassification of citizens, who comprise 97.4 percent of the sample, means that we expect that 19 “non-citizen” respondents (16.5 percent of all reported non-citizens) are citizens who are misclassified. And, those misclassified people can readily account for the observed vote among those who reported that they are non-citizens [emphasis added].

Survey misuse, misdesign, and misinterpretation is a serious problem that we all witnessed right after the 2020 election. This strain of research appears to be another instance of that. There are likely many problems with America’s voting system and there is no doubt that a non-zero number of non-citizens illegally voted, but there is no good evidence that noncitizens voted illegally in large enough numbers to actually shift the outcome of elections or even change the number of electoral votes.

Employment-based immigrants from India and China put up with a lot of bad governmental policies. Despite having wage offers far higher—on average—than immigrants from the rest of the world, these highly skilled immigrants face wait times for green cards (i.e., permanent residence) of a decade for Chinese to 8 decades for Indians, while nationals of other countries hardly wait at all. This is a consequence of the country caps that limit immigrants of any single birthplace to no more than 7 percent of the green card cap unless they would otherwise go unused.

This means that because nearly all of them are already here working on an H‑1B visa, skilled workers from India and China are constantly fearing the prospect of being forced to leave the country if they lose their jobs. They are less likely to get promotions and have trouble changing jobs as a result of the stilted immigration rules for temporary workers. They cannot start businesses, and they must pay thousands of dollars to attorneys and the government to obtain repeated renewals of status. Each administration changes the rules for them, causing additional uncertainty.

But the worst indignity brought upon these talented future Americans is how the system treats their children. Every “minor” child of a temporary H‑1B worker is entitled to H‑4 temporary status. They are also entitled to a green card if their employer sponsors the worker for one. But the moment that the child turns 21, the government cancels their H‑4 status and boots them from the green card queue. They must choose either leaving the country or finding another temporary status to jump to, such as a student visa.

These young people must fight to remain in the country that they have grown up in, graduated from high school in, and have built their lives in. Even if they obtain a student visa, they must then try to win an H‑1B visa through the lottery system, where staying with their family and their adopted country is up to random chance. Of course, even if they get the H‑1B visa, they are thrown to the back of a massive eight-decade long wait for green cards, even though they had already waited in line for a decade or longer with their parents.

High-skilled immigrants will suffer many indignities to stay in the United States—including nationality-based discrimination, closed job opportunities, unnecessary fees, and much else. But many will not tolerate this forced separation. I’ve spoken to hundreds of H‑1B families and every single one has spoken of the stress and anxiety of the thought of having to choose between America and their children. EB‑5 investors who are waiting for visas abroad face the choice of whether to come to the United States at all without their children.

About one quarter of the employment-based immigrants who receive green cards are children. Based on this fact, Figure 1 estimates the composition of the green card backlog in April 2020, showing that about 253,293 children were waiting for green cards with their parents in 2020. That means a quarter of a million children are fearing the prospect of aging out of green card eligibility and status in the United States.

Of the children in the green card backlog, there are 157,064 from India, another 49,835 from China, and 46,394 from other countries. Figure 2 shows that Indian children represent 62 percent of the child green card backlog, Chinese 20 percent, and others 18 percent. Again, the fact that Chinese and Indians dominate the backlog is the result of the country caps where green cards are not issued proportionally to the number of pending applicants in each country but rather limited arbitrarily at 7 percent per nation of birth.

Figure 3 breaks down the child backlog by country and category as of April 2020. Almost 87 percent of the child backlog from India is in the EB‑2 and EB‑3 green card categories for employer-sponsored immigrants with master’s and bachelor’s degrees, while nearly 57 percent of the Chinese backlog are in the EB‑5 investor category. Because nearly all EB‑5 investors come from abroad and are not already working in the United States like those in the EB‑2 and EB‑3 categories, aging out for them would mean never coming to the United States at all. These investor families would have to decide whether to keep the family together or leave their children and immigrate.

Table 1 shows how many immigrant children will age out in each category by country of origin at current rates of green card issuances (accounting for the 2021 increase from spillover of unused green cards from family-based categories). About 104,000 children will age out of eligibility over the next two decades. This is about 40 percent of the entire green card child backlog. More than four in five of the aging out children will come from India—a higher proportion than even their current share of the backlog (62 percent).

Congress should entirely repeal the employment-based green card limits, both the country caps based on birthplace and the category limits overall. But even if Congress does nothing else, it should stop the monumental injustice of aging out. The vast majority of these children grew up in America. This is their home. They should get to stay—just like the Dreamers in the DACA program should get to stay. This injustice harms America as well. More and more talented immigrants are going to Canada, Australia, and other countries because they treat their immigrants well: a reasonable and timely path to citizenship for them and their families. It’s the least they should expect.

Congress has no reason to continue to exclude legal dreamers from its paths to citizenship in bills like the Dream Act that provide for immigrants brought to the United States as children—essentially requiring a violation of the law to obtain citizenship. It’s senseless and unjust. Congress can and should do better.