The workshop’s final panel session featured an open discussion moderated by planning committee member Edward Alden, the Bernard L. Schwartz Senior Fellow at the Council on Foreign Relations. The five panelists participating in this discussion were Felicia Escobar, special assistant to the (U.S.) President on Immigration at the White House; William Kamela, senior federal policy lead for Workforce Readiness and Immigration at Microsoft Corporation; Pia Orrenius, vice president and senior economist at the Federal Reserve Bank of Dallas; Madeleine Sumption, director of research at the Migration Policy Institute,1 and Michael Teitelbaum, senior research associate for the Labor and Worklife Program at Harvard University. The goal of this panel discussion was to identify policy lessons that can be drawn from the research that had been presented and discussed over the course of the workshop.
Before starting the discussion, Alden noted that efforts to pass comprehensive immigration reform had died in the U.S. House of Representatives after making it through the U.S. Senate. This Senate bill would have made some fairly significant changes to U.S. law on high-skilled immigration and it would have been the first time since 1990 that the nation made changes to its policy on skilled immigration. Commenting on the failed bill, Alden said, “It is a little hard to know where we stand now. If we had done this conference 6 months ago, we would all be debating those proposals. Now we are probably going to talk more about what is the power of the President? What can be done under executive order? Are there ways to re-think this going forward?” (Box 7-1).
Michael Teitelbaum noted that several of the presenters mentioned the tensions that are inherent in balancing family versus economic contributions
1At the time of this workshop, Madeleine Sumption was Director of the Migration Policy Institute. She has since moved on to become Director of the Migration Observatory.
On November 20, 2014, the President announced a series of executive actions on immigration. While most of those actions are not targeted towards high-skilled immigrants, the Department of Homeland Security will be issuing a series of rules and regulations geared toward improving access to visas for inventors, researchers, and founders of start-up businesses. Specifically, the Department of Homeland Security, U.S. Citizenship and Immigration Services was ordered to:
- Work with the State Department Work with the Department of State to develop a method to allocate immigrant visas to ensure that all immigrant visas authorized by Congress are issued to eligible individuals when there is sufficient demand for such visas;
- Work with the Department of State to modify the Visa Bulletin system to more simply and reliably make determinations of visa availability;
- Provide clarity on adjustment portability to remove unnecessary restrictions on natural career progression and general job mobility to provide relief to workers facing lengthy adjustment delays;
- Clarify the standard by which a national interest waiver may be granted to foreign inventors, researchers and founders of start-up enterprises to benefit the U.S economy;
- Authorize parole, on a case-by-case basis, to eligible inventors, researchers and founders of start-up enterprises who may not yet qualify for a national interest waiver, but who:
- Have been awarded substantial U S investor financing; or
- Otherwise hold the promise of innovation and job creation through the development of new technologies or the pursuit of cutting-edge research
- Finalize a rule to provide work authorization to the spouses of certain H-1B visa holders who are on the path to lawful permanent resident status. This rule was finalized on February 24, 2015. It has been in effect since May 26, 2015.
- Work with Immigration and Customs Enforcement (ICE) to develop regulations for notice and comment to expand and extend the use of optional practical training (OPT) for foreign students, consistent with existing law.
- Provide clear, consolidated guidance on the meaning of “specialized knowledge” to bring greater clarity and integrity to the L-1B program, improve consistency in adjudications, and enhance companies’ confidence in the program.
Source: U.S. Citizenship and Immigration Services, Executive Actions on Immigration, http://www.uscis.gov/immigrationaction#4.
when considering immigration policy. In particular, he recounted the observations that economic migrants do better, in general, economically, although family migrants may integrate better. Countries paid less attention than expected to the possible interactions between skilled migration and education
and employment decisions of native-born students. He was also surprised by how strong the findings were on the importance of language fluency for economic success in Canada.
In terms of policy comparisons, he noted that Canada and Australia have chosen markedly different policies with regard to allocation between economic and family visas than the United States. Based on the experiences of these countries, his conclusion was that U.S. system allocates too small a share of its total permanent immigration quota to meet economic and employment needs. While U.S. policy on temporary migration compensates in part for this imbalance, Teitelbaum noted that the U.S. system as a whole is not very well balanced. The result is that there are not enough visas in the skilled category to accommodate the large backlog of applicants who have temporary visas and are trying to become permanent residents.
The other lesson Teitelbaum gleaned from the presentations on Canada and Australia is that their systems are far more flexible and nimble than the U.S. system, which he acknowledged arises from the structure of Canada and Australia’s Parliamentary system of government and something that is not likely to be duplicated in the U.S. system with its separation of powers. That flexibility enables countries to try various policies that might have positive impacts and then adjust those policies easily based on experience. Teitelbaum also reiterated Cissna’s point that any policies enacted in the United States should not be too binding because of the difficulty in “tinkering at the margins” in the way that parliamentary systems can.
His last comment on Canada and Australia was that he was struck with how many of the scientists and engineers in those countries were immigrants. “What does that tell us about the trends of domestic science and engineering education and career preparation, particularly within Australia? Are such fields being defined gradually as immigrant-dominated fields that Australians don’t want to go into?” He thought those were important research questions for the future.
The most important lesson that Teitelbaum learned from the United Kingdom’s experiences is how effective the MAC is at helping formulate policy based on data and facts. He credited the MAC with helping the British government evolve from one that he thought was rather naïve about immigration to one that is now seeking the best independent assessments before it makes changes. The MAC has established its credibility as a nonpartisan advisor on immigration policy, so much so that the new coalition government of the United Kingdom not only kept the MAC in place, but all of its members, too. An additional benefit of the MAC’s credibility is that it has lifted the quality of press coverage of immigration issues in the British media, which may improve the public’s understanding of the value of immigration. “Could the MAC model be adopted in the United States?” asked Teitelbaum. “I think it’s worth discussing.”
Madeleine Sumption agreed with Michael Teitelbaum’s takeaways from the workshop and added that something she found striking is how much more strategic and dynamic other countries are in setting their immigration policies. “I love the phrase that the Australians used about ‘adjusting policy settings’ as if the immigration system is this perfectly calibrated machine that responds with sensitivity to their every desire,” said Sumption. She noted that these countries adjust their policies when those policies are not producing optimal results. One example of this was when Australia discovered the flaw in its student selection process. In addition, she echoed Teitelbaum’s and Cissna’s concerns about not making policies too rigid in the United States and about the flexibility that comes with having a parliamentary system of government versus the U.S. system with its separation of powers.
Sumption also asserted that policies in the United States can often be explained only by the history of how they were conceived and not by any substantive reason based on evidence. One example of this is the permanent labor notification process under which an employer does not have to advertise a job to domestic workers before hiring a person via the H-1B visa process. Instead, employers are only required to advertise that job within 6 years of hiring, long after they have employed and invested in an immigrant. Lessons learned from other countries’ experiences often become distorted when they are translated into the U.S. situation. For example, the MAC was distorted into a commission-like entity in the failed Senate immigration reform bill. In the U.S. version, the commission would have had decision-making abilities, but the final decision would be determined through an algorithm combined with performance metrics. In addition, Sumption pointed out the lack of good data in the United States to show how well or poorly the system is working compared to other countries discussed at the workshop.
What the workshop had made abundantly clear, said Pia Orrenius, is that the U.S. system compares poorly with those of Canada and Australia and that immigration policy cannot be the only means by which the United States reaches its goals in terms of meeting its STEM-trained labor force needs. Immigration policy is just one tool of many that can result in a better, more qualified, nimble and innovative workforce. Luckily for the United States, the nation does well in other areas—the quality of our institutions of higher education, the salaries that U.S. employers pay, the flexible labor markets with many job opportunities, and the relative ease with which foreign workers integrate in the U.S. workforce, among others—that enable the country to be competitive in the international market for high-skilled workers. In that respect, Orrenius said that Canada and Australia could learn about other reforms they can make to their labor markets and other regulations that might block immigrants’ access to economic opportunity.
Orrenius was also skeptical of using immigration policy for regional development. “I think if regions are otherwise failing to generate growth for natives, it is folly to think that you can bring in immigrants and somehow generate growth,” said Orrenius. “Again, I think it is not only immigration policy that matters. There are all of these other policies that go into where businesses are created and where firms hire workers and immigration policy is not going to solve those types of problems.” She said the same holds true for the investor programs that Jean-Christophe Dumont discussed. If a country or a region cannot attract investment on its own merit, then investor visas are not going to be a solution either.
According to William Kamela, the real challenge is to integrate proposals and recommendations from academia into the reality of the U.S. political system, particularly given the difficulty in making any major changes in our current immigration policy. De-politicizing immigration policy in the United States by creating a commission is not feasible. Instead, there must be a way to bring the right people to the table before even having the next level of discussion on immigration policy.
Companies like Microsoft will move jobs to Vancouver, Canada if they cannot access more H-1B visas. How do policy makers determine the right policy levers to bring the right employees to the United States to meet the needs of companies that have high-skilled positions to fill. He agreed with Orrenius that the U.S. system will only implement incremental changes to existing policies, despite the need for a completely new immigration policy built from the ground up, the reality is that he does not think that will be possible in the next 50 years. If that is the case, then we need to think of how to navigate the current system so that we have the correct balance between the family-based humanitarian side and the employment-based side, Kamela added as a final comment.
Felicia Escobar acknowledged how nice it was to be in a room full of people who were trying to study high-skilled immigration from a rational framework, and sounded a note of optimism in saying that the immigration debate has benefitted from the research that has been conducted over the last several years. “Research in the last several years has really generated a rich base of knowledge for those who are trying to advocate in Congress for immigration reform,” said Escobar. She voiced her belief about the importance of helping people understand how much this issue really affects the U.S. economy, local communities, and the standing of the United States in the global economy. “It affects the race for talent and the people that we invest in at our institutions and whether they stay here or go elsewhere,” she added.
Turning the discussion to the subject of data, Alden noted that for those who work on trade policy, the International Trade Commission is a rich source of data. Given that there is no analogous source of data for immigration questions, Alden asked the panel to comment on where the biggest gaps in data are and whether it would be useful to have a richer source of information on immigration outcomes in the United States. For Sumption, the main problem is that there are no good data on who is migrating to the United States. There are some basic data on who is receiving H-1B visas, but the data are not publicly available for other types of visas.
Sumption noted, in contrast, that Australia conducts longitudinal surveys that follow immigrants over time to monitor their progress. Canada mines data from its tax system that enables it to track people who came in at different periods and via different avenues and make adjustments based on the data they receive. For example, Canada closed its investor visa program because they found out from tax data that many of these people who were supposedly high net worth individuals and who were coming to contribute to the economy were actually doing much worse than immigrants in almost any other category. The United Kingdom is reviewing its applications and attempting to track people and learn how immigrants are navigating their way through the British system. Sumption said that she believes that the United States could do many of those things if it were able to create systems within the Customs and Immigration Service to track where applications are coming from and where people go once they are in the country.
Teitelbaum noted that when the United States lacked good economic data for policy makers, the results were massive economic booms and busts, panics and bank failures, and so on. There was no Federal Reserve System, no Congressional Budget Office, and no Council of Economic Advisers. In fact, it took about 100 years to develop a Federal Reserve that has people such as Orrenius who can develop data and analyze data in an objective manner, he said. The same is true for the Congressional Budget Office, a more recent innovation. While he agreed with the earlier comment that it was unlikely that the United States would create a body akin to the MAC, he thought that the Council of Economic Advisers could do more to develop and analyze immigration data. Kamela and Escobar agreed. Escobar also agreed that additional data that is shared more publicly with researchers would be helpful, but she also cautioned that immigration policy is not just about creating jobs and increasing economic growth. The United States has a long and rich tradition regarding family reunification and accepting migrants for humanitarian reasons.
In response to a question from Stephen Merrill of the National Academies of Sciences, Engineering, and Medicine about data access, both Sumption and Orrenius noted that there are data being collected that could be used to answer important questions, but that too often they are in a format that is unusable or are not made publicly available because of political reasons. “How else can you interpret the fact that data that existed earlier has been taken away, such as the micro data on green card recipients?” asked Orrenius. Escobar noted that everyone can agree that the Department of Homeland Security needs to update its antiquated systems so that they can generate data in useable formats and in a timely manner, but that funding is the key challenge. She added that the U.S. Digital Service, a new entity within the Executive Branch, was created to help agencies deal with data questions.
Competition for Talent
The panel then considered the question of whether global competition the right way to think about high-skilled immigration. Kamela said that the competition is not just global, it is local, too. Today, he said, there are 27,000 companies participating in the H-1B visa program, and the real challenge is for the nation’s startups. Microsoft, he said, will file 1,000 H-1B visa applications this year and it will get about half of those in the annual lottery. In contrast, a startup that might file one or two applications may not get any of these visas. “What happens to that company? They are competing against us and Google and the challenge is to make sure those companies get a bite of the apple,” said Kamela.
Orrenius said that there absolutely is a global competition for talent, and what is needed is for there to be the same kind of liberalization in immigration policies as there has been in trade policies. She suggested that just as countries and workers have benefitted from free trade agreements in terms of creating more competitive industries and improving the well-being of workers, the same could hold true if countries were to develop free migration agreements. “There are huge untapped gains to had from migration globally,” said Orrenius. The problem, she said, is that most of the gains in the short-term are likely to accrue to the migrant and not the natives.
Teitelbaum noted that the term high-skilled is itself ambiguous because it refers to anyone with a tertiary education, from an associate’s degree to a PhD. What is high-skilled from an immigrant’s country, he said, may be low-skilled or medium-skilled in Sweden or the United States.
Sumption said while competition between companies of the sort that Kamela noted exists, countries have made so little progress on selecting talented
migrants that even a discussion about competition between countries may be premature. People do not move because of immigration systems—they move because there are good economic opportunities. The immigration system then either allows them to pursue those opportunities or it does not. She also noted that the talent pool from which to draw is actually quite large, which makes the supposed competition really about selecting better from those pools. To consider high-skilled migration a zero-sum game at this point is not appropriate in her opinion.
Escobar pointed out that the White House uses the term “best and brightest” and does not frame the discussion in terms of a global competition. She also stressed that when this Administration thinks about a race for talent, it is thinking about the nation’s investments in its domestic workforce which would limit the need to pull in high-skilled individuals from outside the United States.
Noting that many countries are now moving beyond quotas as a means of controlling immigration, Alden asked Orrenius to comment on why the United States seems so wedded to quotas. “Is there any real possibility in the United States of moving beyond a quota system?” asked Alden. Orrenius replied that she found the Australian emphasis on an “expression of interest” system to be an interesting idea, particularly when coupled with the concept of clearing the queue every year. She said that the United States could create a system for employers or firms, rather than potential immigrants, to register their expression of interest to hire a foreign worker. Instead of using a points system, as is the case in Australia, allocation would be one using price—letting employers bid for these permits to hire foreign workers. This type of system, would do a much better job at identifying the best applicants from the pool while also generating revenue.
Sumption agreed that numerical limits are not an effective way to run an immigration system. Quotas have a political purpose and they are useful for governments to demonstrate to the public that immigration is under control, she said, but that are not an effective means for selecting people because you end up with lotteries and a huge backlog. A better approach, based on the experience of other countries, is to establish skill standards and language proficiency requirements. Instead of creating a first-come, first-served system, the United States could implement a more targeted approach. Teitelbaum expressed strong agreement with this idea.
Immigration and wage growth
Turning to the possible impact of immigration on wage growth, Alden asked Kamela to respond to the data presented at the workshop that suggests that STEM occupations have not experienced the kind of wage growth that might be expected. Kamela replied that part of the challenge in answering this question is that STEM covers a wide range of occupations and that wage growth is very occupation specific. He noted that the Senate immigration bill that failed to pass the House had many provisions that would have mitigated challenges around wages and did so in a way that he characterized as thoughtful and supportive.
Kamela added that there are already provisions in existing regulations that mandate certain wage minimums for different kinds of visas, but that enforcement is a challenge. “We are more than willing to say that we need more enforcement,” he said, “We are more than willing to say that there are bad actors in this system that need to be cleaned up.” He also supported the idea of allowing visa portability as a means of addressing wage challenges.
Orrenius then commented that the evidence on wage growth is mixed, and there is no evidence of an adverse effect from immigration on high-skilled natives. She also noted work using H-1B data showing that H-1B visa holders are not underpaid relative to natives and in fact are paid up to 10 percent more than natives.2
What can the President do?
Given that Congress is unlikely to act on immigration reform in the near future, Alden asked Escobar if she could comment on what kind of flexibility the President has to address high-skilled immigration. Escobar replied that the White House is in the process of determining what flexibility exists in the legal immigration space, and that those who are dealing with immigration issues are learning that “there are real limitations to what we can do, which is why we all need to continue to think about changing the actual law to really make the kinds of reforms that people are thinking about.” She explained that the White House is looking in particular at green card and temporary worker reform, as well as at the Optional Practical Training Program, which allows graduates to stay here for a period of time after graduation depending on the type of degree that they have and if they have an employer that will give them additional
training. Other possible options include looking at how the Department of Labor can get involved in the immigration system.
Escobar commented that the workshop’s presentations impressed her by showing how much innovation and change is going on with immigration policy outside of the United States. Kamela remarked that the White House’s legal team is working on this issue which he believes will be well thought out and will withstand legal challenges. He agreed with Escobar that in the end, there is only so much the President can do and the country needs to change its entire immigration system. “These are going to be stop-gap measures because the next administration could have a difference view of these measures and could move to rescind some of these actions,” said Kamela. “While we are incredibly appreciative of the President undertaking this effort, we know that in the long-run we have got to get Congress to act responsibly for comprehensive reform.”
Alden added that there is substantial agreement between the Administration and both parties in Congress on the high-skilled immigration piece of immigration reform. “Substantively, we really went a long way in the process between the Senate and the House, so it was frustrating to not see something come out of it,” said Alden. “Maybe there will be a chance to reengage.”
Jennifer Hunt started the open discussion by asking Kamela and Orrenius to comment on a claim that she hears often in connection with the software industry. According to this claim, older workers in this industry earn more than younger workers but younger workers, because of their more recent training in the tools of the modern software industry, are more qualified. It is further claimed that since the immigration systems bring in and expand the pool of younger, cheaper, and actually more productive people, it leads to older people losing their jobs in this sector. Kamela replied that the “whole younger worker versus older worker thing is a little sensitive, and the reality is that we look for talent, period.” He explained that Microsoft’s older workers tend to be the company’s team leaders with younger workers being part of the team, and that a blend of older and younger workers and of natives and immigrants produces the best outcomes.
Orrenius said that while she has not looked at this issue in any detail, the H-1B data show that those workers tend to be young. She did note, however, that the typical career progression in technology fields has younger workers handling the technical details at the beginning of their careers and evolving into management or the business side of the industry as they get older. Teitelbaum wondered how all of these younger workers can be so technically advanced when they are being taught by older professors. “Have the faculty at these institutions somehow stayed right up to date with all of the new technologies,
but the people in the companies haven’t? That is possible, I suppose,” said Teitelbaum. If that is indeed the case, it would argue for increasing the amount of internal training for corporate technical staff, he added.
The discussion then turned to the subject of skill sets and how immigration policy should reflect the fact that companies put more emphasis on skills rather than degrees when it looks for workers, whether natives or immigrants. Meredith Singer, from IBM, noted that a more flexible and mobility-based immigration system would enable companies to better meet their needs for specific skills. She challenged the economists at the workshop to figure out how to make economic determinations based on skills rather than degrees. Alden mentioned that the workshop planning committee discussed the subject of mobility, and that questions were raised as to whether policy makers need to think in terms of people moving around instead of moving in and settling in one place.
Sumption remarked that Canada’s radical overhaul of its old system, which selected immigrants based on qualifications, was replaced with a system that emphasizes skills and gives companies the ability to select immigrants with the right skills. She added that even with this change, Canada’s system is still based on the idea of permanent immigration and is too slow to allow for the kind of mobility that Singer mentioned. She also said that if the U.S. system actually worked and did not have a massive backlog of visa applications, it would be a system that would be well-adapted to this kind of mobility. “The U.S. has the right model, it is just falling apart,” said Sumption. “We need to fix that model rather than think of something totally new.” Orrenius agreed with Sumption that skills are the key attribute and that employers are the best at gauging whether an applicant has the right skills. Giving that power to companies would also take the selection burden off of the government’s plate. “You can rely on employers to screen these workers and bring them in. You don’t have to have to have a complicated point system, but you have to somehow adopt and change over time,” said Orrenius.
According to Jean-Christophe Dumont, one lesson that OECD countries can learn from the U.S. experience is that they should build an immigration system that adapts and changes over time instead of creating an immigration policy for the next 20 years. “Flexibility and dynamic management are key elements in an efficient system,” said Dumont. He then asked if it would be possible to add some flexibility to the U.S. system so that the country would not have to go through the same debate that it is engaging in now five or ten years down the road. Alden questioned whether a congressionally-mandated system would ever be flexible given Congress’s tendency to micromanage.
Another attendee asked if there have been any studies on an auction system for immigration and what the market-clearing price would be. He noted that migrants from Mexico and South and Central America are paying mules to
bring family members to the United States, and that students are graduating from college with average indebtedness of around $30,000. He wondered if immigrants would be willing to go into debt to secure a U.S. work visa. Orrenius said that she and Madeline Zavodny have written about auction systems from a theoretical position, but that she did not know of an auction system in the real world in which companies could bid for permits to hire foreign workers. She noted that her work looked at a system where 3- to 5-year work permits would run between $10,000 and $15,000 per person, which she qualified by saying that was a back-of-the-envelope calculation. She did not envision a system where visas were auctioned to individuals (including family and humanitarian applicants) but rather one that was employment-based. She was looking at a system that would clear the market yearly or quarterly, eliminate queuing and backlogs, and would raise revenue for the government. This system she and Zavodny envisioned would provide subsidies to some institutions, such as small businesses and universities, so that they would not be priced out of the market.
Kamela remarked that he was involved in an initiative from the Chamber of Commerce and the National Association of Manufacturers which proposed that businesses should pay more to process visas more quickly with the money put into a fund to support STEM education. This fund would have generated close to $500 million dollars and would have been the largest STEM program in the United States. He then said that he and his counterparts in the technology world believe that the long-term solution to meeting the demand for high-skilled workers is to train them here in the United States. The challenge, he said, is that of 42,000 high schools in the United States, only 2,100 offer Advanced Placement computer science courses. He also noted that computer science today is like physics was in the 1950s—a foundational subject that will be the basis of many occupations, not just in the information technology industry. Teitelbaum agreed wholeheartedly with Kamela’s assessment.
The final point of discussion had to do with immigration caps and the current labor certification process. According to Kamela, the labor certification process costs the Department of Labor $40 million annually but does little to protect American workers from being displaced by immigrants with similar skills that are willing to work for lower salaries. He also questioned the need for both a cap system and a labor certification system. Escobar commented that she does not see Congress ever eliminating caps for the green card program or for H-1B, H-2B, and other immigration programs. She also noted that legislation has been proposed that included provisions that would raise or lower caps based on several factors such as employer demand and the economic cycle, but this legislation has never made it through Congress. She did say, however, that she supported the idea of creating a commission or other advisory body to provide support for a flexible cap system, though again, she thinks it will be a difficult sell to convince Congress to provide that kind of latitude.