January 28, 2017. Across the United States, thousands were crowding major international airports. Heavily bundled and scarved protestors shuffled through arrival terminals in Chicago; demonstraters prayed in the baggage claim area at Dallas Fort Worth International Airport. They were responding to the closure of national borders for the next three months to individuals from seven predominantly Muslim countries.
Lawyers flocked to file petitions for recent arrivals being detained at Kennedy International Airport. Prominent politicians trumpeted dissent to the masses, such as Senator Elizabeth Warren saying at Boston Logon Airport, “We will not turn away anyone because of their religion. We are a better people than that!” Hordes bearing bright signs spilled outside San Francisco International Airport, chanting, “Muslim ban has got to go, hey, hey, ho, ho!”
Among these throngs was a middle-aged Russian immigrant, whose family had fled the Soviet Union to escape Jewish persecution. When asked by a reporter why he was present, he commented simply, “I’m here because I’m a refugee.”
To most of the world, however, he is more widely known as a computer scientist, entrepreneur, and—oh, yes—the cofounder of Google: Sergey Brin.
In the following days, many tech leaders voiced their opposition to Executive Order 13769, more commonly known as the president’s travel ban. Some objected on humanitarian grounds, such as Brad Smith in an internal email to Microsoft employees stressing the company’s belief in “the importance of protecting legitimate and law-abiding refugees whose very lives may be at stake in immigration proceedings.” Others, like Netflix’s Reed Hastings, lamented the company-wide impact of Trump’s actions as “hurting Netflix employees” and “so un-American it pains us all.”
Many cited their own reliance on international talent. Apple CEO Tim Cook pointed out that Steve Jobs was himself the son of a Syrian immigrant: “Apple would not exist without immigration, let alone thrive and innovate the way we do.”
Economists are increasingly interested in the question of how immigrants shape the U.S. economy—and the economics of innovation in particular.
But politicians and business leaders aren’t the only ones who have been turning their heads. Economists are also increasingly interested in the question of how immigrants shape the U.S. economy—and the economics of innovation in particular. While it is still early to understand the full implications of the Trump administration’s immigration policy, there are plenty of historical precedents and data that can be mined for a broader picture of how immigrants influence national innovation and economic growth.
Long before Sergey Brin set foot in San Franscisco International Airport, another middle-aged Jewish refugee sailed to American shores aboard the SS Westernland in 1933. He was one of thousands who would flee Nazi persecution over the next decade, arriving to the United States against a backdrop of nativist hostility and suspicion.
Throughout the 1930s, there were political fears about Jews bringing “dangerous ideology” and security risks, not unlike contemporary concerns around Syrian refugees. But there were also live economic anxieties: the United States was surfacing from the Great Depression, and many worried that immigrants would compete for scarce jobs and submerge the economy again. In 1930, President Herbert Hoover directed consular officials to exclude all immigrants “likely to become a public charge”—even those capable of work but poor at the time of admission, a stricter reinterpretation of earlier policy.
This particular immigrant, however, managed to settle in a quiet township in New Jersey. He lived there for his remaining two decades of life, where he described feeling “more at home than ever before.” His name was Albert Einstein, and his scientific discoveries contributed to the Allied victory in the Second World War and the rising prestige of U.S. higher education and military dominance afterwards.
Einstein has become synonymous with “genius.” To be an Einstein is to be the exception, set apart from the rest of the human herd: he can hardly be said to represent the overall demographic of immigrants, much less their economic impact in any direction. Individual accounts like those of Einstein, or the more contemporary Sergey Brin, are case studies, not the systematic reviews economists value as more complete answers to the question of how immigration affects innovation.
But New York University economist Petra Moser and her colleagues have helped to rectify this evidence gap. Intrigued by historical accounts about the German Jewish scientists who fled Nazi Germany, the economists wanted to see if it was possible to quantify their impact on U.S. innovation. After all, Einstein was one of over 133,000 German Jewish immigrants to find safe harbor in the U.S. by 1944.
“Historical records tell us that these immigrants revolutionized U.S. science,” she recounted over the phone. “What the historical accounts could not provide—what was missing—was an estimate of the effect of these arrivals on native invention.”
The researchers focused on German Jewish chemists whose innovative output could be tracked using U.S. patent data from 1920 to 1970. What they found was that inventions jumped by over 30 percent in the fields that these immigrants entered after 1933. In some cases, patent production nearly doubled. But that was just the beginning: their native-born collaborators also became more inventive. Even the coinventors of the collaborators of these immigrants were energized, continuing to be significantly more productive into the 1950s and 1960s. Like stones dropped into a pond, the immigrant scientists created ripples that widened as they spread.
* * *
Economists have long known that innovation is a key driver—perhaps the most important driver—of economic growth. It’s easy to see why. Entrepreneurs and inventors fueled the Industrial Revolution and the tech boom. The creation of new firms and technologies provides jobs, markets, and profit.
What’s less popularly known is just how much immigrants are driving this innovation. As we learn more about the quirks of our modern knowledge economy, a fascinating data set is emerging on the impact of immigrants on innovation, from scientific breakthroughs to new inventions to business creation.
“I’m always trying to understand what affects innovation,” Shai Bernstein, an economist at Stanford University, told me over the phone. He has previously studied the effects of public equity markets and the role of venture capital on innovation. “And then, clearly, immigration became a very big topic over the last few years.”
He had been hearing grumbles from others in the Bay Area about the increasing difficulty of hiring high-skilled workers. Rumors of H-1B visa delays, increases in denials.
Then, in August 2018, a letter that 59 chief executives of the Business Roundtable sent to the Trump administration caught Bernstein’s attention. It warned that current immigration policies inflicted “substantial harm on U.S. competitiveness,” discouraging talented and skilled workers from coming to the United States.
That got Bernstein asking questions. “We became really interested in trying to understand how to quantify innovation,” he said. “Before we even talk about policy, we don’t even really know the underlying number … what’s the contribution of immigrants to innovation? What’s the number?”
A fascinating data set is emerging on the impact of immigrants on innovation, from scientific breakthroughs to new inventions to business creation.
It was one thing to cite extraordinary individuals like Einstein, or even the interesting historical data on German Jewish immigrants that Moser and her colleagues analyzed. But a great deal had changed in the landscape of immigration since the mid-1900s. For the innovation economist, the political arguments and rhetoric were baseless without empirical studies.
Using data from the U.S. Patent and Trademark Office, Shai Bernstein and his Stanford colleagues Rebecca Diamond, Timothy McQuade, and Beatriz Pousada homed in on the roughly one million inventors patenting in the United States between 1976 and 2012. Patent citations serve as a useful way of tracking the innovative output of each inventor. Every invention includes a patent history, a kind of bibliography citing prior technologies that contributed to its development. Hence, the more patent citations a given technology has, the greater its innovative impact on developing new technologies. This meant the economists could measure not only the quantity but also the quality of inventions and identify the top 10 percent of patents. Their method also allowed them to compare the output of immigrant to native-born inventors.
“The results are very consistent,” Bernstein reported. “Immigrants contribute disproportionately to innovation. They’re [responsible for] 16 percent of the inventor population, but 22 percent of the total patents, and roughly the same number if we’re looking at citations.”
That translated into a roughly 40 percent increase in the relative contribution of immigrant over native-born inventors to innovation. This heightened productivity was visible across all their measures: patents, citations, and top-quality patents.
Using a new measure based on stock market reactions, they were also able to calculate the economic value of patents over the 1976–2012 period.
“What’s interesting is looking at the market value [of patents], because you can actually attribute the dollar sign to this,” he added. “The stock market movement not only captures the price value of the firm, but it also captures the economic value of the innovation. Almost 25 percent of the economic value created through innovation among public companies is created by immigrants.” Relative to their share in the population, this reflected a 50 percent increase in their contribution in economic value.
* * *
On the other side of the country, Wellesley and Harvard economists Sari Pekkala Kerr and William Kerr have been looking into another domain of innovation: entrepreneurship. Unlike Bernstein’s sample of inventors, their data set includes both low- and high-tech sectors, from small “mom and pop” shops to start-ups in Silicon Valley.
In their 2019 working paper on immigrant entrepreneurs, they found startling figures: first-generation immigrants create about 25 percent of all new businesses in the United States; this percentage jumps to 40 percent in some heavily immigrant-populated states like California and New York. (In the overall U.S. population, first-generation immigrants composed roughly 14 percent.) These immigrant-owned firms performed similarly to the native-owned firms: they exhibited comparable pay levels, employment, and survival and growth, though they also tended to offer fewer benefits and were more involved in international activities (e.g., import/export, overseas facilities).
So what’s going on? Why are immigrants disproportionately starting new businesses?
“First of all, the share of immigrants in our working-age population has steadily increased,” Pekkala Kerr explained. “The second part of it is that even given the share of immigrants in the population, it looks like each immigrant on average is more entrepreneurial than the typical native.”
Tracing causal mechanisms is hardly clean-cut, but Pekkala Kerr offered several explanations for these differences. Immigrants are, by definition, individuals who have taken a significant risk in leaving their home country to seek better opportunities abroad. This kind of act is itself an entrepreneurial endeavor, and as a group they appear to be more inclined to calculated risk-taking.
In addition, entrepreneurs primarily sit at the upper and lower ends of the education and skill distribution: the “stars” and “misfits.” Immigrants exhibit a similar pattern, populating the lower and higher ends of the education distribution in the United States: as a group, immigrants are 20 percent less likely than native-born Americans to hold a high school diploma, but also 40 percent more likely to hold a doctorate.
As a group, immigrants are 20 percent less likely than native-born Americans to hold a high school diploma, but also 40 percent more likely to hold a doctorate.
They may also be pushed toward starting their own business because they’re less able to find a job that matches their skill set. Because U.S. employers often don’t recognize experience and education gained abroad at the same level, some immigrants embrace entrepreneurship.
“Part of the reason why immigrants are more likely to become self-employed is that they just don’t get the same employment opportunities in regular wage jobs,” Pekkala Kerr said. “Their skills aren’t fully appreciated in the regular labor market, so they may be better off starting their own firm.”
This can be especially attractive if they have strong relationships with tight-knit communities, which enable them to share knowledge and specialize in specific industries like nail salons, dry cleaners, and convenience stores. Various studies have found both of these dynamics at work.
But perhaps one of the biggest reasons is related to a dynamic specific to the economics of innovation. The way that the best new ideas are generated in the first place: talent clusters, or geographic “hotspots” of innovation that concentrate global talent.
“If we’ve all grown up under the same education system, the same perspective on life, the same TV shows, then it’s harder for us to find something new about the way we’re approaching things,” Harvard economist William Kerr observed. “And we’re finding some evidence about how that plays out.”
In addition to coauthoring the working paper on immigrant entrepreneurship, Kerr has written extensively about the economics of talent clusters in his book, The Gift of Global Talent. Historically, popular anxieties around immigration have been tied to a simple model of supply and demand. Native-born workers envision a set number of jobs (fixed demand) and fear that the arrival of foreign workers (increased supply) will push them out of those jobs or drive down wages.
“If we’ve all grown up under the same education system, the same perspective on life, the same TV shows, then it’s harder for us to find something new about the way we’re approaching things.”
He cites the example of local dentistry. This model holds true assuming there’s a set number of townspeople who need their cavities filled. If more dentists move into a town, they’ll have to charge lower prices to compete for those cavity-filling services; some dentists will end up losing out, and others might decide to move out. This is a model in which there are winners and losers, in which more workers means less to go around, i.e., 2 + 2 = 3.
However, in today’s knowledge economy, talented workers are one of the most important drivers of growth. The economics of talent clusters operate on a very different model than those of local dentistry. Instead of a simple equation of supply and demand, one of the most important inputs for innovation among creatives is the presence of other talented creatives.
One of the most prominent examples of this dynamic is Silicon Valley, home to a proliferation of start-ups. The concentration of talent from around the world allows new ideas and networks to come together in unexpected ways. It also pools specialized workers in a single place, making it easier to form matches among entrepreneurs, programmers, and investors.
Unlike the local cavity fillings, the demand for technology stretches far beyond the geographical limits of Silicon Valley. This means the influx of new talent doesn’t necessarily drive down wages, and several studies actually suggest that skilled immigration bumps up the fortunes of local native-born workers.
Kerr argues that this is because adding more talented workers in these cases actually makes the whole greater than the sum of its parts: 2 + 2 = 5. Many immigrants concentrate in these kinds of knowledge economies, making up twice the number of workers per capita in STEM and over half of all doctorate holders in the United States.
Across low- and high-tech sectors, the economists’ research suggests a dynamic that is not competitive, but complementary and collaborative.
“We show some evidence of how immigrant- and native-owned firms are differentially located in different areas and sectors of the economy. For example, the service sector in particular is very immigrant heavy, but manufacturing maybe not so much,” Pekkala Kerr said. “But I don’t see any evidence that there would be a competitive relationship, in the sense that if I start a firm, then that prevents you from starting a firm. We think that firms sort of generate their own market…. It’s much more everyone adds onto the market.”
* * *
Even as economists are gathering evidence about how much immigrants drive innovation, they’re also seeing encouraging spillover effects. Just as the German Jewish refugees of the 1930s and 1940s had ripple effects on their collaborators in chemistry, there’s emerging evidence that modern-day immigrants enhance innovation in their native-born peers.
In their study of immigrant inventors, Bernstein and his colleagues exploited data on the premature deaths of 9,405 inventors in their sample to investigate the impact of immigrants on their native-born colleagues.
For example, say Bob and Ali are a native-born and immigrant inventor team. They’ve coauthored a few patents in their 30s, but then Ali unexpectedly passes away at age 40.
“It’s a little bit morbid, but useful from an identification standpoint,” Bernstein explained. “The idea is that if I didn’t contribute to your productivity at all, then my removal wouldn’t affect your productivity…. However, if I did, then the fact that you’re no longer able to bounce ideas with me, we’re not able to work together, that may hurt your productivity.”
Bernstein and his team matched these deceased inventors with “control” inventors who were similar on various dimensions to them—including immigrant status, age, patent applications, and number of collaborators—but who happily didn’t die before age 60. They could then compare Bob’s productivity before and after Ali’s tragedy to his matched inventor counterpart Rob, whose coauthor Ari didn’t pass away. So what did they find?
Instead of immigrants competing for a slice of American pie, more immigrants grew the whole pie.
“There’s a decline across all the metrics that we’re exploring,” Bernstein said. “A decline in the number of patents, but also the number of top patents, so really high-quality patents are declining.”
Unsurprisingly, the death of a coauthor—whether immigrant or native-born—was bad and sad news for all inventors. But there was a steeper drop when immigrant inventors were removed from the equation: a roughly 26 percent drop in native-born inventor productivity when an immigrant coauthor died, compared to an approximately 10 percent dip when a native-born coauthor passed away.
“I would call it suggestive evidence that the death of an immigrant has the biggest negative effect on productivity of native inventors,” he said. They also found that native-born collaborators improved immigrant inventors’ output more than additional immigrant collaborators, suggesting a synergistic relationship. “Our interpretation is that this supports the idea that the combination of different pools of knowledge is actually really important for the creation of innovation.”
Bernstein’s team was also able to find support for the importance of global talent clusters. When they looked closer at the data, they found that immigrants were actually seeking larger sets of collaborators. Immigrants were also more likely to cite and be cited by foreign inventors, suggesting a more international knowledge network. Finally, immigrants were more likely to concentrate in gateway cities, like Boston and New York.
But there’s still an element of mystery to the mechanics, Bernstein admitted. Even after trying to control for geographical location, technological field, and the year in which inventors were innovating, the economists were only able to explain about a third of the innovation gap between immigrant and native-born inventors. The major contribution of their study was its ability to quantify the disproportionate influence of immigrant inventors: when incorporating both direct and spillover effects, they accounted for roughly 30 percent of total U.S. invention in the past several decades.
“It’s not terribly surprising for most people who have looked at any of the data that immigrants create a lot of firms and the innovation that we’re so heavily relying on,” Pekkala Kerr said. “In terms of our STEM population, in terms of our graduate student population—all of these highly educated populations are heavily immigrant-focused in the U.S. That’s going to then manifest in the innovation as well.”
Even for immigrant entrepreneurs in low-tech sectors, the direction of the results was the same. They were disproportionately creating new businesses, and with them, more opportunities for immigrants and native-born Americans alike. Instead of immigrants competing for a slice of American pie, more immigrants grew the whole pie.
“It’s almost one out of every four jobs among new firms that is created by an immigrant entrepreneur,” Pekkala Kerr observed. “There’s a clear subset of immigrants who are disproportionately creating jobs in the U.S.”
That doesn’t mean immigrants are responsible for a quarter of the entire job market. Preexisting large firms like Walmart and Amazon also contribute to job creation. But given that the vast majority of net job growth comes from new firms, it’s safe to say that immigrants are driving a significant chunk of expanding employment opportunities in the United States.
In other words, these immigrants aren’t taking jobs. They’re making them.
* * *
Even as a growing body of evidence supports the outsized contribution of immigrants to innovation, there is also emerging evidence about the dangers of turning them away. New York University economists Petra Moser and Shmuel San are at work on a new paper exploring a darker side of U.S. history: the nationality-based immigration quotas of the 1920s.
In the years leading up to World War II, nativist sentiments surged as sources of mass migration shifted from Britain and Germany to Southern and Eastern Europe. Many white Americans feared that the United States was losing its predominant ethnic character. In 1921, Calvin Coolidge requested an “ethnic law” to restrict immigration lest the country become a “dumping ground.”
The result was a series of immigration quotas intended to exclude low-skilled immigrants from Southern and Eastern Europe and Asia. From 1923 to 1925, immigration plunged from around 360,000 to 165,000. As it turns out, though, the quotas also led to a dramatic decline in scientists and inventors from the barred regions.
“We show that roughly each year the size of a major [university] department was kept out of the United States—around 30 people each year—because of their country of birth,” Moser said. “You could have two scientists who both worked in Munich, but one who was born in Hungary and the other in Bavaria. U.S. policy at the time would let in the German, but not the Hungarian.”
Using data on over 80,000 U.S. scientists, Moser and San tracked patenting changes before and after the quotas were implemented. They compared the productivity of U.S. scientists working in fields that Southern and Eastern Europeans previously populated with that of their peers working in other fields. Far from protecting native-born scientists, the result was a 60 percent drop in their patent production. Overall, there was a 30 percent decline in U.S. invention—a decline that persisted into the 1960s.
It’s safe to say that immigrants are driving a significant chunk of expanding employment opportunities in the United States. In other words, these immigrants aren’t taking jobs. They’re making them.
“What we find in that paper is that the U.S. never recovered from that,” Moser noted soberly. But she also highlighted another, darker aspect of the exclusionary laws: “The immigration quotas happened in the 1920s. But, importantly, they were active in the 1930s, when you had Jews fleeing from Nazi Europe. Now, when the Jews were German born, they were allowed to come in. But if they were from Eastern Europe, they were kept out.”
Even more devastating than the lasting damage to U.S. innovation was the irrevocable human cost of turning these people away. In May of 1939—just six years after the SS Westernland bore the German-born Einstein to the United States—the SS St. Louis arrived at the shores of Florida. There were over 900 Jewish refugees on board. They were returned to Europe, “despite direct appeals to President Franklin Roosevelt.” Over 250 were later killed by Nazis.
Amid the contemporary immigration debates and refugee crises, the historical echoes around nationality-based restrictions pose an unsettling challenge to our national conscience, raising the question about consequences not just for us, but also for those whom we reject from our harbors.
* * *
The United States has historically been a dream destination for entrepreneurs and inventors. It disproportionately attracts global talent, capturing a whopping third of skilled immigrants from around the world. This extraordinary advantage enables top universities and companies to create vibrant talent clusters, which in turn fuel the nation’s privileged position as a worldwide leader of higher education, technology, and the sciences.
The disproportionate contributions of immigrants are reflected repeatedly across various sectors. Nearly half of all Fortune 500 companies are founded by immigrants or their children. Slightly over half of all Nobel laureate migrants in Chemistry, Medicine, Physics, or Economics lived in the United States at the time of their award-winning research.
But the trends are shifting.
Recent macro-level evidence suggests the United States is attracting less global talent at the university level. From 2015 to 2018, the number of student visas issued dropped by 42.5 percent, from 677,928 to 389,579. At the same time, other countries are drawing more international students. Canada took in over 572,000 international students in 2018, a jump of over 16 percent from 2017. The United Kingdom admitted a record 40,720 international students in 2019, which represents a 6 percent increase (including a 30 percent growth in Chinese students, one of the largest demographics for innovation that may have been deterred from the United States by trade tensions).
Similar patterns are reflected in the broader legal immigration landscape. Recent analysis from the Brookings Institution showed immigration net growth dropped by more than 70 percent in 2018 to 200,000 immigrants: the lowest growth in a decade. This was accompanied by a rate of visa rejections 37 percent higher than it was in 2016 under the Obama administration, which translates to about 155,000 more denials.
The result is an overall immigration slump, and a tapering off in the flow of global talent to the United States.
The result is a tapering off in the flow of global talent to the United States.
The fretful Bay Area murmurs that prompted Shai Bernstein to look into the data around immigrants have also been echoed in William Kerr’s workplace at the Harvard Business School, where applications have been falling steadily. (Across the country, applications to U.S. business schools are down 9.1 percent from last year at 135,096 students, even as Canada’s experienced an uptick of 9.4 percent at 7,703 students. International students, it seems, are moving to Canada.)
“At least for MBA admissions, there’s discussion about access to visas and a worry that [they’re] not going to be able to enter the U.S. labor market afterwards,” Kerr reported. “There’s also frequently a discussion about the kind of animosity towards immigrants, the Trump administration and its rhetoric.”
Of course, the coastal elites and the skilled immigrants they attract are hardly representative of the overall picture of either immigrants or the economy. The tech bubble of Silicon Valley and hallowed halls of Harvard Business School may seem like remote realities for most Americans. But the fruits of innovation—and the immigrants driving it—benefit everyone’s daily life.
At the individual level, Scottish-born Alexander Graham Bell helped to create the telephone; German-born chemist Herman Frasch worked on mineral extraction techniques similar to modern-day fracking. Many of our medical technologies, like surgical sutures, have been developed by immigrants. On a macro level, innovation drives long-term economic growth, higher living standards, and greater job opportunities. And for better or worse, the shifts in immigration policy and rhetoric will affect the entire country.
Innovation drives long-term economic growth, higher living standards, and greater job opportunities. For better or worse, the shifts in immigration policy and rhetoric will affect the entire country.
“If you think about the Muslim travel ban, and some of the restrictions towards the Middle East, these are the places where we’ve been drawing a lot of students in recent years… and that’s becoming more tapered,” William Kerr said. “The overall context of being in America and what that means matters to all immigrants, and if you have an administration that is all up in arms against a caravan at the southern border, or is threatening to revoke birthright citizenship to people that are born in the United States, but have parents that have come from abroad, that creates a lot of uncertainty that spills across the categories.”
One of the things economists know is that greater uncertainty leads to lower investment: the kind of investment required for innovation.
“If the stock market is going up and down, your income is volatile and so forth, you’re less likely to buy a home; a chemical manufacturer is less likely to build a new plant,” Kerr continued. “The same thing happens in a relative sense if there’s a lot of uncertainty about the path America may take and your ability to set up a long-term home here.”
Immigrants, after all, are individuals who are willing to take a risk. To make a calculated investment. And just as good entrepreneurs know that an uncertain landscape is less likely to yield strong returns, many immigrants realize a hostile country is less likely to make a good home.