TLS: The Causal Impact of Foreign and Domestic Doctoral Students on Knowledge Creation and Innovation in US Universities: Evidence from Enrollment Shocks

由于各种原因，包括2001年9月11日之后对外国学生发放美国签证的更严格限制，外国研究生申请美国大学的人数有所下降。这项基于微观数据集的实证研究评估了这一政策的一个潜在的重要经济后果：美国学术机构中高质量的外国研究生的减少。 这可能会对科学和工程部门产生的知识和创新产生不利影响，从而影响美国在科学和工程以及科学创新和知识创造方面的长期领导地位。因此，必须就这种可能性提出认真的统计证据。使用所有博士的机密微观数据。根据1960年至2004年美国大学的获奖者，该项目创建了博士课程入学人数的衡量标准，并按学生的国籍、学习领域和日期以及大学和学术部门的入学情况进行了细分。然后将这些指标与特定学术部门的知识和创新产出数据（包括专利、专利引文、研究出版物和引文）相结合，研究科学和工程领域的国内外研究生是否对知识创造和商业创新做出贡献。实证识别策略利用了来源国经济特征和政策的变化，这些变化导致了与美国创新无关的外国研究生潜在供应的外生变化。 特别令人感兴趣的是，国内和国际学生在知识和创新生产中的替代程度。以适当的基本参数模拟了国内外研究生与大学和研究领域匹配的动态模型，然后受到各种冲击（如学生工资机会的变化）。这些模拟被用来解释计量经济学的结果，并评估有关学生签证限制的替代政策的可能影响。对外国申请减少的一个主要担忧是美国大学研究能力的潜在下降以及由此导致的更广泛经济中创新的损失。这一分析应使政策制定者了解这种担忧是否有道理，以及随着时间的推移创新减少的程度。它也有可能告知联邦关于签证和研究补贴的政策优先事项，包括赠款计划。该项目评估了其他签证限制政策的创新影响，包括有利于高质量外国学生的政策，以及对申请人财务状况有一定影响的现行政策。这项研究还可以为加速永久居留权的潜在好处提供指导。美国研究型大学越来越依赖于联邦赠款和许可收入来支持他们的研究计划。这一分析提供了强有力的证据，证明外国和国内研究生在创造支持这种收入的知识方面的生产力。因此，调查结果应能提供有关对大学进行科学和工程学赠款和研究生培训的投资所产生的回报的信息。此外，这项研究将产生一个重要的新数据库，将入学微观数据与大学知识创造和创新的衡量标准以及对更广泛商业部门的溢出效应联系起来。该数据库应立即有价值的（a）未来的研究知识创造的具体领域，（B）管理研究的研究组织在大学设置，（c）研究赠款评估。

For various reasons, including tighter restrictions on the issuance of U.S. visas to foreign students after September 11, 2001, applications by foreign graduate students to U.S. universities have fallen. This empirical study, based on micro datasets, evaluates one potentially important economic consequence of this policy: the decrease in high-quality foreign graduate students in U.S. academic institutions. This may adversely affect knowledge and innovation produced in science and engineering departments, thereby affecting long-run U.S. leadership in science and engineering and in scientific innovation and knowledge creation. Thus, it is important to develop careful statistical evidence on this possibility. Using confidential micro-data on all Ph.D. recipients at U.S. universities from 1960-2004, this project creates measures of enrollments in doctoral programs, broken down by national origins of students, field and dates of study, and university and academic department of enrollment. These measures are then combined with data on knowledge and innovation outputs attributable to specific academic departments, including patents, patent citations, research publications and citations, to study whether foreign and domestic graduate students in science and engineering fields contribute to knowledge creation and commercial innovation. The empirical identification strategy exploits variation in economic characteristics and policy in source countries that led to exogenous changes in the potential supply of foreign graduate students for reasons unrelated to U.S. innovation. Of particular interest is the extent to which domestic and international students are substitutes in the production of knowledge and innovation. A dynamic model of matching between foreign and domestic graduate students with universities and fields of study are simulated with appropriate basic parameters, and then subjected to various shocks (such as changes in wage opportunities for students). These simulations are used to interpret the econometric results and also to evaluate the likely impacts of alternative policies regarding student-visa restrictions. A major concern about declining foreign applications is the potential reduction in the research capacities of American universities and the consequent loss of innovation in the broader economy. This analysis should inform policymakers about whether this concern is warranted and the extent to which reductions in innovation over time are anticipated. It also has the potential to inform federal policy priorities regarding visas and subsidies for research, including grant programs. The project evaluates the innovation impacts of alternative visa restriction policies including a policy that favors higher-quality foreign students versus the current policy that puts some weight on the financial status of applicants. The research can also offer guidance on the potential benefits of accelerated tracks toward permanent residency. American research universities are increasingly reliant on federal grants and licensing income to support their research programs. This analysis offers strong evidence of the productivity of foreign and domestic graduate students in generating knowledge that supports such revenues. Thus, the results should be informative about the returns generated by investments in universities for grants and graduate training in science and engineering. Further, this research will generate a significant new database linking micro-data on enrollments with measures of knowledge creation and innovation by universities and spillovers into the broader commercial sector. The database should be immediately valuable for (a) future research on knowledge creation in specific fields, (b) management studies of research organization in university settings, and (c) research on grant evaluation.