A Fresh Approach to Immigration

Over the past 20 years, the control of immigration in the U.S. science and technology (S&T) sector has become a topic of perennial debate. Foreign-born scientists and students make up a significant and growing share of those holding or pursuing degrees in science and engineering. As the job market for new Ph.D.s becomes tighter and doctoral students’ future prospects appear less secure, the issue of whether and how to control immigration has once again come to the fore.

Devising a sensible set of policies to deal with the flow of foreign scientists and engineers to our shores is a daunting task even for the most creative policymakers. At present, immigration ceilings are established for a very broad range of skilled occupations, of which doctoral-level science and engineering employment is only a small component. Policymakers infrequently re-evaluate the level at which these ceilings are set. There is no explicit mechanism for evaluating the demand for scientists and engineers, and changing the aggregate ceilings will do little to balance demand with supply. Moreover, because science and engineering labor markets are highly dynamic, it is difficult to anticipate the long-term trends on which more narrowly defined ceilings could be based. Finally, it is unclear how those ceilings should be established. Many different stakeholders are involved, each with a different set of interests. How should we balance these competing objectives? To answer these questions, we need to determine whether the flow of immigration in science and engineering is, in fact, a problem-and if so, for whom.

Challenges to policy formulation

Scientists and engineers are a relatively small proportion of the immigrants for whom occupation is reported-about 2.5 percent in 1993 (the latest year for which data are available). Yet immigrants make up a large proportion of Ph.D.-level scientists and engineers. In 1993, for instance, immigrants comprised 23 percent of those holding doctorates in all science and engineering fields, and 40 percent of those holding doctorates in engineering. Even more important, foreign students account for nearly all of the increase in the number of doctorates awarded in these fields since that figure began to rise in the mid-1980s. Clearly, immigration is a critical element in formulating policy for the science and engineering labor market.

Perhaps the most significant obstacle in formulating immigration policy for doctoral scientists and engineers is the difficulty in assessing whether such immigration is, on the whole, good or bad for the country. A number of studies have sought to address this issue. In the mid-1980s, for example, the National Academy of Engineering (NAE) funded the study Foreign and Foreign-Born Engineers: Infusing Talent, Raising Issues by a committee of the National Research Council to describe and assess the advantages and disadvantages of relying on foreign engineers in our work force and graduate programs. The committee pointed out that immigration provides access to a large pool of highly qualified and motivated engineers, some of whom may make significant contributions to our S&T enterprise. Those coming from low-income countries such as China and India may be willing to work for low wages or stipends, making U.S. research activities cheaper. By keeping wages low and by attracting a broader pool of talent, immigration produces benefits for the universities, research institutes, and corporations that employ scientists and engineers. The benefits of this research are enjoyed by a number of stakeholders, from the industries and government agencies that fund the research to the individuals who consume the goods it generates.

However, the committee also noted that immigration brings disadvantages to U.S.-born students and science and engineering professionals. Competition with immigrants may bring down their wages and reduce their access to graduate training and job opportunities. Low wages and limited opportunities in turn may discourage future generations of domestic talent from pursuing science and engineering careers at the doctoral level. The committee also identified potential problems associated with language barriers and cultural orientation. Nonetheless, the committee concluded that during the 1980s, on balance, the advantages of having foreign and foreign-born engineers in our work force outweighed the disadvantages. A comparable study of scientists was undertaken by the National Science Foundation (NSF) several years later and reached similar conclusions.

The larger question, raised by both the NAE and NSF studies, is why more Americans are not going into these fields. A 1995 book by David North examines the role of immigrant scientists and engineers in the labor market and concludes that the relatively slow growth in the number of Americans choosing these careers is in part the result of the poor earnings potential as compared with that of other professions such as law, medicine, and business. A 1993 study by Derek Bok arrived at similar conclusions. Some part of the earnings disadvantage in these careers arises from the relatively low wages or stipends earned during the six to nine years it takes to acquire a doctoral degree and the additional three to six years that many new doctorates spend in postdoctoral appointments. But even after completion of these periods of study and apprenticeship, salaries in many scientific and technical fields are lower than those received in other professions requiring extensive postgraduate training.

In the past, the parsimonious stipend levels for new Ph.D.s presented few barriers to recruitment of young people, because they reflected an implicit bargain between faculty and students. Committed students were willing to make financial sacrifices for a few years in exchange for the promise of a meaningful post-training career in research. Unfortunately, the current tight market in academic employment for science and engineering Ph.D.’s means that, for significant numbers of young scientists and engineers, these implicit agreements are largely honored in the breach. We are now experiencing the costs of this failure in terms of frustrated expectations and thwarted careers, and they are substantial. Inability to honor these agreements may well discourage future generations of domestic talent from pursuing science and engineering careers at the doctorate level.

A large foreign presence in our graduate programs may be one factor causing native-born students to pursue careers in other fields. But the underlying problem-the long-term structure of the job market for science and engineering Ph.D.s and the career paths it offers-deserves careful attention as well.

Policy formulation is also hampered by gaps in immigration data. It is hard to know exactly how many doctorate holders enter the United States as immigrants, because the Immigration and Naturalization Service (INS) does not provide information on the degree level of immigrants. U.S. universities identify new doctorates by citizenship status and report whether they intend to stay in the United States. But this information covers only their plans for the following year. Thus we do not know how long immigrants who earn doctorates here actually stay, or whether those who leave eventually return. A 1995 study by Michael Finn reports that nearly half the foreign citizens who received doctorates from U.S. universities during the 1980s were still in the United States in 1992. We cannot tell, however, whether immigration is growing or shrinking in importance in the academic community.

It is also hard to determine whether immigration improves the quality of domestic R&D. Clearly, immigration would be more beneficial if immigrant scientists and engineers turn out to be more creative and productive than equivalent domestic scientists and engineers. A not-yet-completed study by Paula Stephan and Sharon Levin, using data from the early 1980s, suggests that immigrants may be more productive than nonimmigrants. Without more recent evidence, however, it is hard to know how much weight to give this conclusion.

Principles for policy formulation

In the absence of clear objectives and adequate information, the debate over immigration policy for science and engineering Ph.D.s has been guided primarily by supply considerations. Over the past few years, immigration ceilings have careened between open and more restrictive measures. More permissive policies are enacted to enhance the supply of talent when labor markets are strong, and less permissive policies are used to protect and enhance job opportunities for young U.S. scientists and engineers when labor markets are weak. In the late 1980s, NSF’s prediction of massive, looming shortfalls of scientists and engineers in the 1990s was one factor motivating large increases in employment-based ceilings for skilled workers-from 54,000 to 140,000 per year-embodied in the Immigration Act of 1990. When the forecasts of shortfalls proved dramatically wrong and the job market for doctoral scientists and engineers began to turn sour, concern shifted from future shortfall to current glut.

Although recent proposals for immigration reform have reflected greater attention to the impact of immigration on U.S. scientists and engineers, efforts to moderate the large increases adopted in 1990 were blocked during the 104th Congress. Instead, public attention focused on illegal immigration-an important issue with respect to total immigration, but relatively unimportant for science and engineering doctorates. No amendments were adopted to the regulation controlling immigration in science and engineering. This outcome reflects the built-in inertia of public policy as well as the political and financial superiority of those advocating no change, from organized ethnic and religious groups to organizations representing research universities and certain employers such as Microsoft and Intel.

What principles should govern policy formulation with respect to science and engineering immigration? First, we should set clear priorities among the competing objectives that such a policy can meet. Although the immediate goal may be to control supply, it is important to recognize that this is only a step toward the higher-order objective of ensuring the health and vitality of our nation’s R&D enterprise. Moreover, measures designed to restrict the supply of scientists and engineers may actually conflict with this larger goal by raising the cost of research. The dilemma is that in today’s financial climate of budgetary restrictions at all levels of government and retrenchment in corporate funding for research, policies that reduce the cost of research through a more permissive immigration policy for scientists and engineers might arguably serve some public interests but at the same time create disadvantages for some young U.S.-born scientists and engineers.

In this regard, it is striking to note the chorus of concern now being expressed by the American Medical Association, American Association of Medical Colleges, Association of Academic Health Centers, and other leading medical organizations about what they perceive to be the growing oversupply of physicians. They attribute this to the large number of foreign medical school graduates being trained as residents in U.S. hospitals and call for reduction of the generous funding for such training positions that currently flows from federal health care programs such as Medicare.

Second, efforts to regulate immigration should be based on long-term considerations of supply and demand, not short-term fluctuations. For example, we cannot be certain that the current weakness in the market for science and engineering doctorates represents a long-term oversupply. Although there is general agreement that we are producing more doctorates than required to meet the current demand for tenure-track faculty, particularly those engaged in academic research, this oversupply may turn out to be offset by demands in other segments of this market, such as secondary school instruction; industrial research; or professions such as law, consulting, or banking. On the other hand, the reluctance of academic employers to make the commitments implied by tenure-track appointments may represent a long-term structural shift rather than a response to short-term financial uncertainties. Certainly it is reasonable to expect that the budgetary constraints underlying the apparent weakness in demand will persist until at least 2002 and probably well beyond that unless steps are taken to balance the federal budget.

If, in the face of these market conditions, we wish to preserve young Americans’ commitment to research careers, we will need to address the problem through means other than controlling immigration. In particular, we should reexamine the use of low-paid graduate students to staff research projects. Current practices are rooted in Vannevar Bush’s vision of academic science as the joint production of research and research scholars. Because many universities can no longer afford to provide tenured positions for the research scholars produced by this arrangement, they should establish new staffing arrangements to produce the research; for instance, by contracting with employees on a project-by-project basis and providing appropriate financial compensation instead of the promise of a career that may never materialize.

Uncertainty about the future also raises questions about continuing to rely on immigration to meet our nation’s R&D needs. Although such a policy might be desirable from the narrow perspective of maintaining the volume of academic research, we must recognize that the future may not continue to provide a flow of foreign talent to our shores. We are already beginning to see declines in enrollment of foreign citizens in our graduate science and engineering programs, from 109,000 in 1992 to 105,000 in 1993. The size of the drop is not dramatic, but the direction is clear. We are also seeing declines in the number of foreign scholars at our major research universities. The Institute for International Education reports that the number of visiting scholars declined by 1,900 (about 3 percent) in 1995-and this was the second consecutive year of decline. These declines may be a temporary aberration, they may reflect the relatively poor job market in this country, or they may indicate improvements in the quality of graduate education and academic research overseas. Regardless of their origin, if these declines continue they will strongly suggest that our R&D and graduate education enterprises are beginning to contend with viable competition from abroad.

Some modest suggestions

Given that science and engineering immigration is but a small portion of total immigration, we believe that, at a minimum, policy formulation should address the needs of science and engineering more directly. Attempting to tailor overall immigration policy to the particular circumstances of scientists and engineers would be letting the tail wag the dog. Nonetheless, the issues uniquely associated with science and engineering immigration warrant special consideration.

Specifically, we recommend that a balanced panel of distinguished experts be created to propose separate immigration ceilings for scientists and engineers. The panel should operate under the aegis of the Office of Science and Technology Policy, with input from the Department of Labor, the INS, and federal science and technology agencies such as NSF, the National Institutes of Health, and the National Aeronautics and Space Administration. Its recommendations would be considered and administered by the Department of Labor and the INS as part of the larger numerical limits set by congressional legislation.

The goal of creating a balanced panel is to ensure objectivity in the assessment of current and future labor market conditions as well as wisdom in developing a set of recommendations that will resolve (or at least minimize) the various conflicting interests involved. The panel should base its recommendations on a comprehensive review of recent immigration ceilings; how they affect the health of our national R&D enterprise; and whether they are consistent with explicitly stated objectives, including the relative attractiveness of careers in science and engineering. Moreover, the panel should update its recommendations every three to four years to ensure that immigration ceilings are based on the most recent information available.

Ideally, immigration policy should emerge from careful consideration of proper social objectives and a serious assessment of the benefits and costs of attaining those objectives. In the absence of a mechanism specifically designed to regulate science and engineering immigration, there is no easy way to balance the objectives of enhancing the nation’s science and engineering enterprise and protecting the legitimate interests of U.S.-born scientists and engineers. A first step toward this important goal is to create a more rational framework for proposing recommendations that can refine and update the immigration policies established by previous legislation.