In the recent presidential campaign, neither Barack Obama nor Mitt Romney said anything meaningful about how to fix the U.S. innovation system. Both ran as status quo candidates who acted as though there was nothing to fix. Virtually no one who follows the issue agrees with this, but for the moment the political system remains oblivious.
About ten years ago, I began to hear murmurs from high-tech industry people that American innovation wasn't working as advertized. Some of this was disappointment tied to the dot-com bust, but most of it consisted of people looking ahead in their particular technology-based industry and not seeing much. I had a number of off-guard conversations at wedding receptions or post-conference airport cabs in which someone would say, "we don't really have a next big thing in the pipeline anymore." I heard this from a pharma scientist in 2002, then from a corporate computer engineer, and something similar from IT executives at different firms over the ensuing months. Official reports started coming out, such as "Rising Above the Gathering Storm" (2005), which warned that the U.S. was losing its innovation lead. At the time, however, most people assumed, like Romney and Obama this year, that the innovation system just needed a few tweaks and bigger inputs --- more funding for research, more science and engineering (STEM) graduates.
Phase 2 arrived a few years ago, and the gathering storm became a heavy rain. We now have a steady patter of worried, critical books, sometimes written by the U.S. innovation system's biggest boosters. There was Fareed Zakaria's Newsweek cover story in November 2009, bluntly titled, "Can America Still Innovate?" This was followed by his book, The Post-American World (2011). New York Times columnist Thomas Friedman captured the same problem in his title, That Used to Be Us: How America Fell Behind in the World It Invented and How We Can Come Back.
Many systematic academic studies have been moving in the same direction. In 2007, Richard Florida published The Flight of the Creative Class: The New Global Competition for Talent, whose implications for the U.S. innovation system is clear in the subtitle. More recently, the economist Robert J. Gordon caused a stir with his NBER paper offering quantitative evidence that the rate of increase of multi-factor productivity was higher in 1928-1950 than in the vaunted Internet era. Scholars are now questioning the effectiveness of one after the other of our main innovation pillars. In a recent book by two principals at the Information Technology and Innovation Foundation, Robert D. Atkinson and Stephen J. Ezell, note that over the past 30 years the U.S. has lost its position in one major industry after another -- consumer electronics, capital goods industries, machine tools, automobiles. The same then happened in the high-tech sector, when U.S. market share slid in industries from printed circuit boards and
desktop and notebook PCs to liquid crystal displays (LCDs), advanced batteries, and compact fluorescent lightbulbs (CFLs). Moreover, U.S. leadership in the industries that will define the future -- including high-performance computing, artificial intelligence, biotechnology, nanotechnology, robotics, energy storage, and clean energy production -- is by no means assured (33).
Why do these declines keep happening? Every element of the country's past success is coming under scrutiny -- the patent process, the higher education system, and the condition of the physical and digital infrastructures, among many others. One perennial target is the bottleneck in U.S. access to the world's technical workforce. This is the subject of Vivek Wadhwa's new book, The Immigrant Exodus, in which he claims that U.S. innovation is gravely threatened by a greatly inadequate number of visas for immigrant entrepreneurs.
Building on statistical research by AnnaLee Saxenian, Wadhwa shows first, that Silicon Valley companies have had a high dependence on immigrant founders, and second, that the ranks of these founders are no longer growing as expected. From 1995 to 2005, he writes, "more than 25 percent of all the technology and engineering companies in the country had one or more immigrant founders. The proportion of technology and engineering companies in Silicon Valley with at least one immigrant founder was 52 percent." A large share of these entrepreneurs came from two countries, India and China (which together have 35 percent of the world's population). India has been a particularly large donor to the skilled California workforce: during the 1990s, "the population of Indian scientists and engineers (S&E) in Silicon Valley grew by 646 percent" (six times the growth rate of the S&E workforce overall). As a result, "within a decade the proportion of Indian-led startups had increased from 7 percent to 13.4 percent."
The problem, as Wadhwa defines it, is that this immigrant flow "has slowed and reversed to an Immigrant Exodus." The term is riveting -- "exodus" suggests a sudden mass outflow of biblical proportions. The book does not actually present evidence of this kind of a reversal. When Wadhwa conducts a 2012 followup to a previous nationwide survey, he finds that "the proportion of immigrant-founded companies had slid from 25.3 percent to 24.3 percent." Wadhwa's research thus indicates not an exodus but a stagnation of immigrant entrepreneurship. He believes that this stagnation indicates a peak and a possible turning point in the proportion of U.S start-ups that are lead by immigrants.
Although the title, Immigrant Exodus, overstates the problem, there is little doubt that the U.S. visa system is clogged. Wadhwa nicely describes the discouraging effects of the long waits and uncertain outcomes on thousands and thousands of highly educated immigrants. His solution consists of seven visa reforms. One is to break open the quota system by eliminating the 7-percent-per-country limit for employment-based and related visas (only 7 percent of visas each year can at present be granted to nationals from a given country). Another is to increase by "3 or 4 times the number of employment-based green cards issued per annum." He also advocates letting spouses of H-1B visa holders work, and proposes a new "start-up visa."
Wadhwa claims that his visa reforms would get the growth in the number of immigrant entrepreneurs back on its earlier, New Economy pathway. This would in turn, he says, give a jolt to the overall American economy: "these measures could conservatively add $100 billion to $150 billion in revenues to the U.S. economy and generate between 1 million and 1.5 million jobs." Since STEM employment is about 6 percent of U.S. employment, or roughly 8 million jobs, Wadhwa is suggesting that ending current visa restrictions could increase U.S. STEM employment by 20 percent in the coming years.
Wadhwa discusses another explanation for the stagnation of STEM immigration figures, one explored in books like Florida's mentioned above, and that is the growth of high-tech opportunities in traditional donor countries like China and India. His current surveys find reduced percentages planning on staying in the U.S. long term. "72 percent of Indian and 81 percent of Chinese returnees [to their countries of origin] said that the opportunities to start their own businesses were 'better' or 'much better' in their home countries." Wadhwa suggests that the "American dream ... has lost its luster," but what we might more rigorously find in his data is the reemergence of the Indian dream and the Chinese dream, towards which Indian and Chinese STEM workers might logically gravitate.
Wadhwa is right that the U.S. employment-based visa system can and should be reformed. But is the stagnation in the growth of immigrant company founders a serious problem for technology innovation? This connection is much less clear, and may be the reverse of what Wadhwa suggests.
First, the numbers involved are too small to move the innovation jobs needle. Only a fraction of immigrant STEM workers become company founders, the average time to founding, Wadhwa indicates, is 13 years, and startups take years to hire significant workers if they ever do. Accepting for the moment Wadhwa's estimate of a 20 percent increase in startups through visa reform, this increase would be spread over many years. If it took only a decade for a visa to produce many new jobs -- an accelerated time frame, the increase in STEM employment would be about 2 percent a year from this source.
Second, Wadhwa does not clearly identify why the U.S. should want immigrant entrepreneurs as opposed to their homegrown variety. His main answer seems to be their superior technical training, although Wadhwa doesn't delve into the complicated question of qualifications. He is most convincing on the topic of the immediate practical difficulties faced by both large and small technology companies when they look for highly qualified STEM employees. But this suggests that STEM immigrant workers are most essential as working engineers rather than as deal-making company founders. The U.S. has no shortage of M.B.As awaiting their chance to be entrepreneurial as Wadhwa defines it. In fact, the U.S. may have too many STEM entrepreneurs and not enough STEM workers. We may have a shortage of people who stay away from finance, fund raising, and business plan pitching, in order to solve hundreds of everyday problems in their technical jobs. Bringing foreign entrepreneurs to the Valley has the air of bringing coal to Newcastle. They should not be excluded harsh visa restrictions, but that doesn't mean more of them will add new unique innovation value.
I have already broached a third issue, which is the impact of a focus on increasing immigration on the domestic innovation system. The core problem Wadhwa identifies is a shortage of great STEM workers, which follows from the U.S. failure to produce enough STEM graduates domestically. As a result, U.S. companies have spent decades hiring advanced workers on the open global market. Each individual decision was reasonable, but the aggregate policy has been cheap and lazy: it is less expensive and easier to let someone else's country educate a creative workforce than it is to fund and teach your own. In California, the epicenter of the New Economy, the Valley has done little to restore the public education system to the post-war glory that had provided the workforce for the initial high-tech boom. This has been particularly unfair to California's large Latino population. Wadhwa's book would be more compelling, and more useful as policy, if it encouraged the Valley to spend as much time and money renewing public education as it has spent lobbying for visa expansion.
There is also the issue of innovation's impact on global justice. Innovation theories can be lumped into two broad categories--those that focus on the ultra-cultivation of a small elite, and those that focus on what I think of as mass creativity, the maximum expansion of high-order human capabilities throughout the global population. From this latter, democratic, global perspective, Wadhwa's immigration stagnation is good rather than bad news. Silicon Valley loves graduates of India's world-beating system of technology institutes, which are funded by the Indian people. But why should taxpayers in a poor country create engineers for a rich one? If brilliant Indian engineers make jobs in India instead of California and Massachusetts, this reduces an unjust Indian subsidy to a rich Western nation, while more effectively reducing poverty rates and the absurd waste of human potential that poverty creates. From these perspectives, America's loss of Chinese and Indian STEM workers is China's and India's gain, and also the world's.
Finally, mass creativity is more important to innovation success than is the totemic American figure of the entrepreneur, the Schumpeterian hero who creates wealth through personal greatness. This can be seen by reading a biography of Steve Jobs, which beneath the surface is actually a collective biography of hundreds or thousands of people who made Apple Computer possible. Many of them were Indian and Chinese immigrants, who played a major role in creating Silicon Valley wealth in their role as superb technical workers rather than as founders and entrepreneurs. And not only as technical workers. STEM employment, as I mentioned, is 6 percent of total employment, while "creative class" employment is closer to 30 percent of the total. Legions of non-STEM "creatives" -- from all over the world -- built an innovation ecosystem, and we are finally beginning to understand in a deep and sustained way that innovation depends less on entrepreneurs than on the cognitive and social dynamics of the "creative class" -- really classes -- that form a multitude or ensemble that drives all sustainable development. The Valley's obsession with the entrepreneur is a cultural blind spot, and in Wadhwa's case it misstates as a need for start-up founders what is actually a need for a whole range of creative employees -- including artists, designers, narrative specialists, cross-disciplinary integrators among many others.
We are learning to question the orthodox elements of the current innovation system, including an overreliance on venture capital and strong, uniform intellectual property rights. We need to do the same with the role of immigration of technological workers. Vivek Wadhwa's specific reforms should be discussed and in most cases implemented. But as the presidential campaign has taught us, we need a rebuilt innovation system rather than an extension of the old one. This will depend on putting STEM immigration issues back into the framework of innovation as a function of social development.