I have been invited to comment here today on the issue of high-tech worker shortages and U.S. Immigration Policy, particularly on the proposal to raise the annual quota of so-called H1-B visas, which enable skilled engineers to enter the United States to work for American high-tech companies.
Cypress Semiconductor is a $600 million international supplier of high-performance semiconductors for computer, networking, and telecommunications companies. We will be 15 years old on April 7 of this year. We employ 2,771 people. We design our chips in San Jose, CA; Minneapolis, MN; Seattle, WA; Portland, OR; Colorado Springs, CO; Austin, TX; Starkville, MS; London, England; Bangalore, India; and Cork, Ireland. We manufacture our chips in Minneapolis and Austin. Our retained earnings (lifetime cumulative profit) are $334 million, making us the 15th most profitable semiconductor company in the history of the American semiconductor industry. In short, Cypress Semiconductor is a typical Silicon Valley success story, and our experience with immigrants is also very typical.
The semiconductor business is one of the most competitive you will find, and we have strong feelings about the value provided by immigrant engineers—and about the factually hollow, emotion-driven claims of those who insist the U.S. semiconductor industry can retain its current global leadership without an adequate supply of high-quality engineers, including immigrants. Underlying the arguments of critics is the notion that immigration is a zero-sum gamethat each job for an immigrant means one less for a U.S. native. That idea is not only unquestionably false, but also the opposite of the true impact of immigrant engineers.
I will proceed by refuting a number of conventional arguments against legal immigration, using real data from Silicon Valley. The following points will become evident:
+ The need for skilled workers in the high-tech sector is growing exponentially, causing chronic shortages. The loss of one important source of skilled workers would cause America to lose a competitive advantage to other countries.
+ Foreign skilled workers do not take jobs from domestic workers, but, in fact, create additional jobs.
+ Congressional inaction on the current immigration cap of skilled workers would accomplish what our worst economic enemies hope to achieve through competitive strategies.
+ Members of Congress who would tolerate an increase in the skilled worker immigration cap only at the expense of family immigration attempt to pit the groups that support legal immigration against one another to effectively kill any chance of increasing the skilled worker cap. Such members do not have the interests of American high-technology business in mind, and Silicon Valley will view such a proposal as an attempt to divide the people, rather than to help our businesses.
+ The loss of a reliable supply of skilled workers would leave many high-tech companies with little choice but to locate operations in countries with an adequate supply of skilled workers.
Let's look at the arguments and counterarguments surrounding legal immigration. As we will see, the data supports only one conclusion: Legal immigration provides tangible economic benefit to the broad cross section of Americans.
Cypress recruits from a total of 26 major schools nationwide (see Appendix A), including Georgia Tech, Stanford, the University of California at Berkeley, UCLA, and the universities of Illinois, Minnesota, Washington, and Texas. This effort is supported by our Human Resources department, but the drivers of the recruiting program are our senior vice presidents. Many of them are alumni of the schools they are assigned to engage. The point is that college recruiting rates the very highest-level priority at Cypress as it does at most high-tech companies in Silicon Valley.
Why is there a shortage of engineers? High-technology companies pump tens of millions annually into bolstering science learning while sponsoring job fairs and recruiting intensely on U.S. campuses. Both my company and I personally have contributed directly and substantially to the Stanford Electrical Engineering program. Stanford is currently raising $200 million from Silicon Valley entrepreneurs specifically to endow 300 science and engineering graduate student fellowships to replace those lost to recent Federal R&D budget cuts. Despite these recruiting efforts and education funding, the domestic supply of engineers is too small, leaving the unemployment rate of electrical engineers at an all-time low of 0.4%, as reported this month in Electrical Engineering Times, our highest-circulation trade publication (see Appendix B). At the same time, most high-tech companies have dozens to thousands of positions open, in some cases for years.
After Cypress recruits engineers, we train them with dozens of in-house technical and nontechnical courses, adding up to thousands of hours of training each year. Larger companies like Intel have even more extensive in-house training courses.
Some of our courses are taught by experienced Cypress insiders with many years of experience in the semiconductor business, but we also go outside for best-in-class trainers, including professors from Stanford University, the University of Santa Clara, and U.C. Berkeley. And Cypress pays for outside education, as well. Engineers are offered the opportunity to return to school part-time or even full-time to get advanced degrees, partly underwritten by Cypress up to a maximum of $7,500 per year.
Despite our significant investment in recruiting and training, we have historically averaged between 75 and 250 openings at any given time. Currently we have 75 openings, even though we're right in the middle of one of the deepest pricing recessions the semiconductor industry has ever seen.
Fifty-nine of our current 75 openings are for engineers (see Appendix C)—we tend to cut non-engineering hiring during bad times, but we never stop recruiting engineers. Our current engineering shortfall translates into approximately 10 new-product development projects that we are unable to start. (Appendix D lists 17 projects that are ready to be launched and are waiting for engineering staffing.) The direct consequence of unlaunched new-product projects is slower growth, which, in turn, means we will fail to create all the new jobs we might be able to under full employment conditions. Even with our intensive recruiting efforts bolstered by the hiring of some immigrant engineers, we're still not filling all of our employment needs.
This is the weakest of all arguments against legal immigration, although it certainly has high emotional appeal to some. Consider the case at Cypress, which is run by 10 vice presidents and me, and which is overseen by four outside directors. Our management team supports 2,771 jobs, of which 2,011 are in the U.S.
Four of our 10 vice presidents are immigrants. John Torode, vice president of our Seattle Computer Products Division, came to America after World War II as a dependent immigrant with his father, a British sailor. In later life, John received a Ph.D. in computer science and became a professor at U.C. Berkeley. At Cypress, John's division makes the clocks used to synchronize 20 million personal computers a year. Recently, John's division invented a new chip that will allow us to participate in a rapidly growing market called "USB." That stands for the Universal Serial Bus, the new way in which personal computers are being hooked to all of their peripherals (printers, modems, etc.) beginning this year. Recently, we promoted John to the position of chief technical officer.
Lothar Maier, our vice president of manufacturing, came to America from Germany as a child. Lothar began his career at Intel and moved to Cypress for the opportunity to work at a start-up company and to get a promotion to manage a group of engineers. A decade later, he is now our vice president of manufacturing, entrusted with 1,067 jobs at six manufacturing sites.
Emmanuel Hernandez is our chief financial officer; he was relocated to the U.S. from the Philippines based on outstanding performance by his former employer, National Semiconductor, Silicon Valley's second largest chip company.
Tony Alvarez, our vice president of research and development, immigrated as a child in 1961 from Castro-controlled Cuba. He directs a force of 113 engineers in Austin, San Jose, and Minneapolis that develop our most advanced silicon technologies.
Jose Arreola, our chief scientist, works for Tony, and has the direct responsibility for our advanced semiconductor technologies. Jose's group just brought our largest chip into production—a memory chip containing 26.2 million transistors. Jose immigrated to the U.S. from Mexico in 1974 to attend graduate school, where he received a degree in transistor physics. He entered America on an employment visa, and later he became a citizen. Jose manages a group of 30 people, 80% of whom have postgraduate degrees—two-thirds of them legal immigrants. If you remember Pat Buchanan's last campaign, he derided immigrants by labeling them "Jose." I can state unequivocally that our Jose has made 2,011 Americans better off. Our top transistor physicist, Jeff Watt, is a member of Jose's team. He is a Stanford-educated Ph.D. who immigrated from Canada.
Cypress also has immigrants on its board of directors. These directors have helped us through the difficult environment in our world, and they have changed the course of business history.Pierre Lamond, our chairman, was born in France, received a technical degree there, and was recruited to the United States to work with Fairchild Semiconductor, the progenitor of virtually every important Silicon Valley company, including Intel, National Semiconductor, Cypress, and dozens of others. Pierre was a founder of National Semiconductor, the second-largest Silicon Valley chip company, which employs 13,000 workers worldwide.
Later as a venture capitalist and corporate director, Pierre's venture firm, Sequoia Partners, was instrumental in founding Apple Computer (8,000 employees), and biotechnology ground-breaker Genentech (3,200 employees), as well as Cypress and more than 200 other high-tech companies with a market capitalization of $175 billion and more than 150,000 total employees. In addition to his board position at Cypress, Pierre also serves as the chairman of Vitesse Semiconductor, the world leader in digital gallium arsenide integrated circuits, semiconductors five to 10 times faster than silicon, used to transmit data on the Internet at the astronomical rate of 250,000 typed pages of information per second.
Eric Benhamou is a Cypress director and is also president and chief executive officer of 3Com Corporation. Eric fled the Algerian civil war with his parents in 1960, ending up in France, before moving to the U.S. to attend graduate school at Stanford University. Eric went on to build 3Com, the world's leading supplier of data, voice, and video communications technology for the exploding Internet. 3Com has 100 million customers worldwide and employs 13,200 people. It is highly unlikely that any e-mail any of you receive does not pass through at least one 3Com product.
Without question, Cypress's 2,771 jobs and its success in Silicon Valley could not have happened without the immigrant engineers who help me run the company and the immigrants who serve as board members.
Perhaps the most-often cited example of an immigrant who brought value to this country is Hungarian refugee Andy Grove, the founder and CEO of Intel, the world's largest semiconductor company that produces the processor chips for 80% of the world's personal computers. Intel has more than 50,000 employees, 32,000 in the U.S.
But numbers tell only part of the story. If not for immigration policies that allowed them to come to America, Grove, Lamond, Benhamou, and the hundreds of other immigrant executives in Silicon Valley, would be applying their skills to the benefit of our competitors overseas. Consider the net effect to the American economy if these men had remained in Europe, designing products that helped to build a European Intel or a Taiwanese Apple or a Korean 3Com. Historical "what if" games offer none of the precision of the mathematical equations engineers are accustomed to dealing with. But it's interesting to ponder what high-technology advantage America would have, if Europe had been the birthplace of the microprocessor or the personal computer or the driving force behind the Internet.
My point is that it is absolutely preposterous to claim that any of these men took jobs away from native-born Americans. We would lose jobs without our immigrant talent. The logic of those who claim otherwise, including high-ranking members of the Clinton administration, borders on absolute folly.
Meanwhile, consider a more quantifiable measure of the economic value provided by immigrants, this one involving rank-and-file Cypress engineers. Cypress has 470 engineers and 2,771 worldwide employees. Roughly speaking, this means each engineer creates jobs for six additional people who make or administrate or sell the products developed by our engineers.
A disproportionate number of our R&D engineers—about 36.6% are staffed by immigrants, a typical situation at high-tech companies. It follows that if we had been prevented from hiring the 172 immigrant researchers we have hired, we would have failed to create about 1,000 other jobs—70% of which are held by native-born Americans. Immigrant engineers not only do not take away jobs, they actually create jobs for native-born Americans.
Meanwhile, it is important to recognize that only two of Cypress's six immigrant vice presidents and directors relocated here specifically to take a job. The other four job creators—including our vice president of research and development and our chief technical officer—came to America as dependent children. I would oppose strongly the linking of the passage of H1-B legislation with a reduction in family immigration, viewing it as a political maneuver made at Silicon Valley's expense.
If we believe that letting highly skilled people into America will make all of us better off, then we should not tell those who relocate here, "You can come to America, but you must leave your family behind." Simply put, I cannot attract top-flight engineers if I must tell them they will be prohibited by U.S. law from sponsoring close family members to join them in the U.S.
Our international competition could not wish for more than for Washington to reduce the number of skilled engineers available to American high-tech businesses. During the 1980s, Japan battered the U.S. semiconductor industry. Now Japan lags us again partly because of an acute shortage of engineers there due to Japanese population demographics and the fact that the Japanese have imposed a short-sighted limit to their influx of immigrant engineers. On a daily basis, our competitors in Tokyo scheme to stop the momentum of the American semiconductor and computer industries. Even if they tried, they could not come up with a better plan than to cut off our supply of critical engineering talent by halting immigration. Unfortunately, it appears they may have the United States government as their ally.
High-tech resembles football in that the team with the best players wins—and it appears that we are turning away the best players because we are unwilling to let a very small additional number of highly talented people into the United States, many of whom will be turned away after being educated in our colleges and universities!
Government meddling in immigration policy takes its toll here everyday. It has already hurt my company. Again, consider Cypress's involvement in the market for Universal Serial Bus microcontrollers.
Last year in our Austin, Texas plant, Cypress worked on a critical USB project. The project was disrupted when the U.S. hit its annual 65,000-person legal immigration cap in August, a month before the fiscal year's end (the cap was imposed arbitrarily in the 1990 Immigration Act). Immigration was summarily cut off, and one of our Canadian engineers was forced to return to his home country, delaying our project, which is still not complete. We delayed the sale of tens of millions of USB chips per year, and the employment of the hundreds of workers needed to make and sell them. This is a direct example of lost jobs due to bad government policy.
Attached to my testimony is a list of the projects that are currently backlogged at Cypress (see Appendix D), for lack of engineering resources. These projects include those for USB; computer memories, communications devices, such as cellular phones; and special memories used to launch information onto the Internet.
Cypress cannot find enough skilled workers to fulfill its maximum growth potential. Multiply our troubles by the thousands of semiconductor, system, software, computer, and networking companies in Silicon Valley, and the true proportions of this crisis emerge. The Information Technology Association of America estimates that there are 346,000 highly skilled positions going unfilled in U.S. companies! More than 70% of the companies surveyed by ITAA identified this as the leading barrier to their growth and competitiveness.
The Clinton-Gore administration has portrayed itself as a friend of Silicon Valley, yet on issue after issue, from securities litigation, encryption, the Internet tax, and now H1-B visas, it has turned its back on high-tech companies. Today, when Silicon Valley companies are in desperate need of increased access to skilled professionals, the message we have heard from the administration on the H1-B issue is simple: "Drop dead."
Yes, "Drop dead." I refer specifically to the intransigent position of Commerce Secretary Daley that it is "not feasible" to increase the cap on H1-Bs. If the Administration holds to this position, it should forget about portraying itself as anything but an enemy of Silicon Valley companies. Appealing to populist nativism by claiming these highly skilled individuals will be taking away American jobs may appear to have some short-term political appeal, but the implications to the high-technology industry are obvious.
If the administration holds to a position that forbids companies from hiring skilled foreign nationals beyond the current, arbitrarily imposed 65,000 cap, it will have achieved only one objective—to drive the resources of more and more companies offshore.
The committee should also know that most of our H1-B hires are individuals of either Asian-Pacific or Hispanic descent, just like many other immigrants. Neither these individuals nor anyone who comes through the family immigration or refugee system should be maligned unfairly for "taking away American jobs." H1-B immigrants work hard and honestly, and they create innovation and wealth in our society. I treat my employees with respect, and I hope others will, as well.
One more time, let's look at the data as it applies to Cypress. Take the 75 to 250 annual openings we've averaged since 1993, the vast majority of which were openings for engineers. No semiconductor company can find all the design talent it needs to grow in the tight Silicon Valley engineering market. In response, we started our first remote design center in 1987, in Starkville, Mississippi. After that, we added design centers by acquisition in Washington, Colorado, and Minnesota. But eventually we opted to expand outside the U.S.—first with an acquisition in London, England, and eventually through startups in Cork, Ireland, and Bangalore, India. Our offshore design centers communicate via satellite, and Cypress now has more than 60 engineers in offshore locations. We are being forced by government policy to fill the jobs offshore that we cannot fill here.
This common argument, usually stated as "High-technology industries require a large international pool of cheap workers to sustain growth," is also false.
San Jose's immigrant population is among the nation's highest. Yet competition for workers here is so intense that the average wage in 1995, the most recent year for which figures are available, was $42,000, No. 1 in the nation ahead of No. 2 New York, another highly concentrated immigrant center. According to a study by the Cato Institute's Stephen Moore, "Immigration and America's Cities: A Sign of Economic Health," the per-capita money income in cities with a high immigrant concentration outstrips that of low-immigrant cities, $14,748 to $12,851. If immigrants supposedly lower salaries, why does the data show exactly the opposite?
Higher immigration levels seem also to correlate with lower unemployment: According to the Manhattan Institute's Index of Leading Immigration Indicators, the 1995 unemployment rate in the states with the highest immigrant presence, including California, was a full 1.5% lower than the rate in low-immigration states. If immigrants take our jobs, why is unemployment lower where they concentrate?We enjoy a low unemployment rate in immigrant-rich Silicon Valley. In the broader U.S. electronics sector, the average unemployment rate of electrical engineers scraped bottom at 0.4% in the 1997 fourth quarter, according to Robert Rivers, editor of the Engineering Manpower newsletter, as quoted in Electronic Engineering Times.
This rock-bottom figure virtually assures all engineers of receiving full wages and top dollar. It is absurd to suggest, as do some anti-immigration zealots, that a two-tiered salary structure exists, with foreign-born engineers earning less than native-born Americans of comparable age and education levels. If a company tried to pay foreign nationals less, in addition to breaking the law, it would destroy itself as its engineers quit to go elsewhere. In Silicon Valley, almost any engineer in any company on any given day can quit and have another job by the end of the same day.
The competition for workers is so intense in Silicon Valley that Cypress's average San Jose employee, including vice presidents, line workers, and receptionists, earns $84,474 a year including 19% benefits. Our average secretary earns $54,408 with benefits. The Cypress immigrant managers and vice presidents I described earlier all earn six-figure incomes. Whose pay are they holding down? And with an unemployment rate of 0.4%, where are the out-of-work American engineers that the alleged oversupply of foreign talent is injuring?
Bad argument #6: Only government can strike a "correct" balance between the economic and social costs and benefits of immigration, safeguarding the interests of "real" Americans.
I always cringe when Washington tries to "manage" anything. The House Bank and White House Travel Office come immediately to mind. Consider some of the ridiculous immigration-control proposals that have circulated in Washington in recent years.
One proposal would have required companies to run identity verification tests on employees to get government clearance to hire. Such tests would have relied on personalized birth certificates with a fingerprint or other biometric data such as the retinal pattern in our eyes. The system would be enforced by yet another inefficient government bureaucracy with the potential for error and waste on a massive scale. Imagine the economic gridlock that would result from having to phone Washington for permission to make a job offer. And even if there were just a 1% error rate in the national employment database, 650,000 jobs a year would be denied by "mistake"—a government-caused human tragedy.
Another preposterous proposal was for the United States to send immigrants home for three years after they had completed their U.S. college educations. In other words, we were advised to burn millions of dollars educating immigrants (currently one-half of all new U.S. Ph.D. engineers are immigrants) and then send them back to their native countries, where they would have made our foreign competitors stronger!
This year, "trial balloons" have been sent up to see if industry would accept government controls on layoffs in return for relief on immigration. This bad idea sounds like the great French employment control system that has produced 11% unemployment by restricting the abilities of employers and employees alike to make free-market choices. My company actively avoids hiring in countries where we face unreasonable government restrictions.
In no other area of the national debate have I seen the wholesale abandonment of data, logic, and reason as I have in the debate on immigration.
Consider that today our legal immigrant population is only 8.5%, compared with more than 13% between the 1860s and the 1930s, and that immigrants add less than 0.4% to our population yearly. Our immigrant population is low—roughly half that of Canada, Australia, Switzerland, or New Zealand. Add to these statistics the documented value that immigrants provide to all Americans, and our response to the immigrant "problem" should be obvious.
The National Science Foundation reported recently that although immigrants represent less than 10% of the U.S. population, they make up almost 30% of scientists and engineers with doctoral degrees engaged in research and development. Cypress is a microcosm of that big picture. On a bulletin board in our research and development division, push pins on a world map identify our employees' home countries or regions, including India, China, Southeast Asia, and Western Europe. There's even a pin representing an engineer from Mongolia.
I have spoken in other Congressional forums about the American Dream—that every generation will enjoy a higher standard of living. But there's another part of that dream: that America's strength derives from a melting pot of diversity, and that diversity builds wealth rather than destroying it.
Immigration is well-controlled. Raising overall immigration quotas by only 3% specifically to bring in valuable new engineers and scientists would be good for all Americans. We should enjoy our diversity and work on the real problems of the country.