Getting Schooled: The Role of Universities in Attracting Immigrant Entrepreneurs

Natee Amornsiripanitch is a PhD candidate in Financial Economics at Yale School of Management; Paul A. Gompers is Eugene Holman Professor of Business Administration at Harvard Business School; and Kaushik Vasudevan is a PhD candidate in Financial Economics at Yale School of Management. This post is based on recent paper authored by Mr. Amornsiripanitch, Mr. Gompers, Mr. Vasudevan, and George Hu.

Immigrants play a vital role in innovation activities and entrepreneurship. Given the substantial contribution of immigrants in these areas, a set of natural questions arise: what are the pathways that high-skilled immigrants take to arrive in the United States and how has the importance of these pathways changed over time? What are important institutions that serve as gatekeepers for high-skilled immigrants and does it affect the types of immigrant founders that come to the United States? Do certain parts of the United States benefit disproportionately from high-skilled immigration, and if so, what are some factors that contribute to these benefits? The answers to these questions have important implications for designing immigration policy and regulation which have become increasingly acrimonious topics in public discourse. They also have important implications for firms and universities which recruit talent from abroad and the communities that hope to promote vibrant entrepreneurial ecosystems.

In this paper, we seek to address the questions that we raise by leveraging a combination of unique datasets that allows us to identify immigrant entrepreneurs and to more closely study their backgrounds. Particularly, we combine a dataset from Infutor, which enables us to proxy for the immigration status of individuals in the United States, with VentureSource, a workhorse dataset for the study of VC ecosystem, which contains detailed information on the near universe of venture capital-backed startups in the United States. We supplement these data with hand-collected information on the education and fields of study and prior work experience for entrepreneurs in our sample. Combined, these data provide a detailed source of information that we take advantage of to understand the pathways that high-skilled immigrants take towards entrepreneurship.

We begin our empirical analysis by documenting several facts about immigrant entrepreneurs in our sample. We estimate that approximately 20% of the founders in our sample are immigrants, broadly in-line with estimates in other work of the immigrant share in other entrepreneurial and innovation activities. Consistent with overall immigration trends, we find that the proportion of East Asian and South Asian founders has increased substantially over our sample. We also find that immigrant founders are just as likely as native founders to start firms that have initial public offerings (IPOs) or that are acquired at valuations greater than 100 million dollars, markers of success for VC-backed firms. Moreover, relative to natives, immigrants disproportionately tend to start Information Technology (IT) firms. On a relative basis, venture-backed firms founded by immigrants also tend to be founded in coastal states, most notably California, Massachusetts, and New York, as opposed to smaller states in the southern and middle parts of the country. Thus, even relative to their overall greater share of venture capital-backed companies, these coastal states have higher proportions of founders who are immigrants. This last fact indicates that the benefits associated with high-growth potential, immigrant-founded firms have been especially geographically concentrated in certain parts of the country.

Having established these broad facts about immigrant entrepreneurs and their firms, we move to leveraging our unique data on the educational and work history of immigrant entrepreneurs, to better understand immigrant pathways towards entrepreneurship. We categorize each immigrant entrepreneur in our sample into one of three categories: those that came to the US first for college, those that came to the US first for post-graduate education, and those that came to the US after receiving their education elsewhere. This classification allows us to more closely explore how the different pathways of high-skilled immigration to the United States contribute to the pool of entrepreneurial talent. This classification reveals a striking fact; more than 75% immigrant entrepreneurs that we have education information for received some form of education in the United States. Of this 75%, more than half of the entrepreneurs received an undergraduate degree in the United States. The substantial portion of immigrant entrepreneurs educated in the US is one of our key novel findings and suggests the extent to which universities play an important role in importing global entrepreneurial talent to the United States.

More closely focusing on entrepreneurs across the different groups, we find that immigrant entrepreneurs that come to the US for college tend to have similar educational backgrounds to native entrepreneurs. In contrast, entrepreneurs who come to the US for graduate educations or for work are substantially more likely to have studied a science, technology, engineering, and mathematics (STEM) discipline; consistent with this fact, these groups of entrepreneurs are also responsible for the disproportionate representation of immigrants among founders of IT firms in our sample. Interestingly, while the share of immigrant entrepreneurs has stayed relatively constant over our sample, we find that the relative share of US college-educated immigrant entrepreneurs has been rising, while the share of immigrant entrepreneurs that came to the US for graduate school and work has been falling.

Lastly, we study the location in which immigrant founders across the different groups start their companies in order to unpack some of the drivers behind which geographic areas tend to benefit from immigrant entrepreneurs. We find that more than 40% of founders in our sample found firms in the same state that they were educated. This fact is not simply driven by Berkeley- and Stanford-educated entrepreneurs founding companies in the Bay Area, and Harvard- and MIT-educated entrepreneurs founding companies in Massachusetts. Instead, it captures a more general phenomenon that applies to non-venture capital hubs as well. The evidence suggests the presence of top universities has likely been an important determinant of which areas have benefited most from the immigration of high-skilled immigrants that go on to start firms. More generally, this result provides additional evidence for the role that universities play in local agglomeration economies. Universities are known to contribute to local economies in a variety of ways, for example by training a skilled labor force, or by knowledge diffusion from innovation activities. Our results suggest that this agglomeration benefit extends to attracting skilled immigrants, some of whom end up starting high-growth potential firms.

From a policy perspective, our results emphasize the importance of immigrant entrepreneurs as founders of high-growth potential startups. While a substantial focus in the current public discourse revolves around work visas, such as the H-1B visa, our evidence suggests that student visas may deserve almost as much attention. Policy proposals in the past few years have sought to place restrictions on student visas for foreigners. Given the substantial contribution of immigrant entrepreneurs educated in the United States, our results suggest that such policies likely carry significant costs for the country, by restricting the supply of talented potential entrepreneurs. Our results also suggest that there is a substantial lag between when immigrant entrepreneurs enter the country, and when they found their firms. Accordingly, the effects of policies that increase or decrease the flow of immigrants may have very persistent effects on immigrant entrepreneurship which only show up decades after the implementation of the policies.

The complete paper is available for download here.

Both comments and trackbacks are currently closed.