My Stream of Consciousness

Thank you for the opportunity to appear before the Subcommittee on Technology, Environment, and Aviation. I am here at the invitation of Representative Walker and the minority. Two years ago, I appeared before the Subcommittee on Technology and Competitiveness at the invitation of Representative Valentine and the majority. I appreciate the bipartisan interest in free-market technology development.

In my right hand I have a data-communications chip made by Cypress Semiconductor. We call it HotLink. It is capable of transporting information over a wire, or through an optical fiber, at the rate of 330 million bits per second. In my left hand I have a 4,196-bit static random-access memory (SRAM) chip made by Cypress. It is capable of storing and retrieving data in three nanoseconds--about the time it takes light to travel one yard. It is the fastest SRAM of its type available from any company in the world. Our HotLink chip would undoubtedly be part of any data communications network created in the United States--and in high volumes. Our super-fast SRAM is currently being used in conventional supercomputers.

I also have here a Cypress module containing two powerful SPARC processors of the kind that are used--500 modules at a time--in a massively parallel computer made by a company called Thinking Machines, in Cambridge, Massachusetts. Thinking Machines is a competitor of MasPar, one of the companies whose CEO, Jeff Kalb, is testifying here today.

In other words, Cypress makes data-communications chips used in electronic superhighways, memory chips for supercomputers, and microprocessor modules for massively parallel computers. We would benefit greatly if billions of taxpayer dollars were showered on the various technology projects favored by the Clinton administration. It would be easy for me to support these projects. I could spend one minute talking about our products, a few more discussing the wonders of the basic technologies, a few more minutes on the serious peril we face from other countries, especially the government-financed Japanese and Europeans, and finally, I could ask for a dole--to save American high technology.

But I am here to say that such subsidies will hurt my company and our industry. Why? Because they represent tax-and-spend economics--a brand of economics that is a known failure. I do not want handouts. The men and women of our company do not want handouts. And if Congress wants to help American high technology, handouts are the wrong way to go--especially if they are funded with huge tax increases on individuals and corporations.

The subject of today's hearing is the High Performance Computing Act of 1991. But it is impossible to separate high-performance computing from the broader Clinton technology program. And it is impossible to separate the Clinton technology program from the administration's broader economic program. Thus, I will begin by presenting my views on the administration's general approach to economic and technology policy. I will then address the particular issues surrounding high-performance computing. I will conclude with some concrete suggestions for enhancing America's technology leadership.

CLINTON PROGRAM MISCONCEPTION ONE: UNANIMOUS SILICON VALLEY SUPPORT

Let me begin by correcting a serious (albeit politically useful) misconception. The administration would have us believe that the business leaders of Silicon Valley stand unanimously behind its program. The image of John Sculley, CEO of Apple Computer, sitting beside the First Lady and applauding the State of the Union address has been beamed far and wide by White House political operatives. John Sculley and I are friends and neighbors. We live in the same small town in the hills above Stanford University and Silicon Valley. But on this issue we are thousands of miles apart.

Indeed, I am here today in strong opposition to the administration's economic program in general and the big-spending aspects of its technology agenda in particular. I am not alone. Over the last week, in preparation for this testimony, I corresponded with directors, founders, and CEOs of ten high-technology companies. Not one agreed with the proposition that the right way to enhance America's technology leadership is to increase individual and corporate taxes to finance government mega-programs--even if those mega-programs would benefit their own companies. What follows are ten opinions expressed to me over the past week:

Scott McNealy, CEO, Sun Microsystems, America's largest manufacturer of workstations: "In the current economic climate, the proposed increase in the corporate tax rate does not encourage job growth, business investment, or global competitiveness. Rather, it penalizes profits and will result in further loss of jobs."
Joe Zemke, CEO, Amdahl Corporation, America's largest manufacturer of IBM plug-compatible mainframes: "Whether it is sugar subsidies or ‘investments' in high-performance computing, the Clinton program represents the same logic: siphoning dollars from individuals and corporations and allocating them through a process that is terribly inefficient--a process that is responsive not to market requirements, but to bureaucratic empires and political payoffs. This use of tax money is disturbing at any time, but to increase expenditures as we face a historic deficit is unconscionable.

"As a high-technology executive who faces the rigors of the market every day, I view both the data highway and any subsidy of high-performance computers as the most recent examples of industries lining up to feed at the public trough. There may be a few select winners, but the majority, and the taxpayers, lose."

Don Valentine, venture capitalist, founding venture capitalist and director, Apple Computer, currently a director of five companies, including $600-million Cisco Systems: "Don't assume that the Pepsi-Cola kid [John Sculley] speaks for Silicon Valley. We do not need pretenders who speak for us. We have visionaries who are rare, important, and doers.

"To Washington I say, please do not help us. The world of technology is complex, fast changing, unstructured, and thrives best when individuals are left alone to be different, creative, and disobedient. Go help the Russians. They are a Third-World technology state. Go help all the people who know how ‘pork' works, and who want to be taken care of. But please do not help us: Anyone who thinks corporate taxes promote employment does not understand the problem."

Wilf Corrigan, CEO, LSI Logic, America's largest gate array manufacturer: "I am a strong supporter of industrial policy, but lowering taxes would be the best form of industrial policy we could have. We should balance the budget by cutting spending, and if that means we cannot put money into the high technology infrastructure, that is okay. If wealthy individuals get taxed more, they will spend more time figuring out how to minimize taxes and less time creating wealth."
Finis Conner, founder and CEO, Conner Peripherals, Silicon Valley's leading computer disk-drive maker: "President's Clinton's proposed tax program will raise the U.S. corporate tax rate by 2 percent. We believe this strategy will accelerate the move of U.S. jobs offshore. There are plenty of countries that will welcome these jobs with open arms--and will offer tax incentives rather than tax increases. The development of all technologies and products involves risks and rewards. The government should not be in the business of speculating with taxpayers' money on which of those risks will be winners and which will be losers."
Gil Amelio, CEO, National Semiconductor, $1.6 billion chip company: "Our current tax code encourages and instigates class warfare. Today, the top 5% of all wage earners pay 44% of all income tax--and if Mr. Clinton has his way this will increase further because he has campaigned on the basis of ‘the politics of envy.' He wants to punish with high taxes Americans who have been successful--that is, the people we need to revitalize our economy!"
Pierre Lamond, venture capitalist, founder, National Semiconductor: "Every dollar that is taxed away from individual investment or corporate R&D will weaken America's high-technology companies."
L.J. Sevin, venture capitalist, former chairman, Convex, a supercomputer company, chairman, Cyrix, chip supplier: "The companies that the administration claims got a ‘free ride' generated all the jobs and foreign exchange. And the so-called free ride probably earned the government a factor-of-ten return on the investment. Somehow, the administration's attitude seems to be that any money the government does not take in taxes is a gift to corporations."
John Adler, CEO, Adaptec, a $300-million supplier of components and software to the personal-computer industry: "I was delighted with President Clinton's initial two-for-one deficit-reduction target. I am now deeply concerned about the trend of moving away from significant deficit reduction to significant increases in government spending. I am not in favor of increased government spending--even if it is called investment, and even if it is directed to high technology."
Roger Emerick, CEO, Lam Research, a leading supplier of semiconductor-manufacturing equipment: "Large investments in R&D and in building a world-class manufacturing capability have allowed Lam Research to gain global market share and create 350 new jobs over the past 12 months. Raising Lam's corporate taxes without strong additional R&D and investment incentives will reduce our ability to create jobs in the future."

These comments represent only a selection of recent communications with senior executives in Silicon Valley. So as we evaluate the President's program, let's be clear: Silicon Valley does not stand as a monolithic group in support of these tax-and-spend economic policies. If anything, the opposite is true.

CLINTON PROGRAM MISCONCEPTION TWO: SOAKING THE "EXCESSES OF THE '80s" IS GOOD POLICY

There is a second dangerous misconception about the President's economic program. It too serves a useful political purpose. Indeed, it is a game as old as politics itself: divide-and-conquer, or, as Gil Amelio says, preaching "the politics of envy." Yes, the White House tells the American people, we plan to increase spending by hundreds of billions of dollars. But we plan to spend that on "you." Even better, "they"--the bad guys--pay for it.
The bad guys, of course, are successful individuals and profitable corporations. Throughout the campaign last fall, and through the first two months of the administration, we have heard endless talk about companies which, in the words of Labor Secretary Robert Reich, got a "free ride" in the 1980s. Or, as the President himself argued in the report that followed his State of the Union, the burden of the economic program will fall on "those who profited most from the uneven prosperity of the last decade."

Somehow, we never get to meet these bad guys. They are never in the room when tax increases are being discussed. Instead, political operatives offer caricatures--images of Michael Milken and Ivan Boesky defrauding investors of billions. Who would not want to tax such "excesses of the 1980s?"

So please allow me to reintroduce myself: I am an excess of the 1980s. Based on my ownership stake in Cypress, I am one of the people who, in the President's words, "profited most from the uneven prosperity of the last decade." I became a paper millionaire in the 1980s--eight times over, in fact.

How did I profit? I started a company in Silicon Valley. I obtained stock in that company when it had one employee (me) and one used computer. I worked with that company for a decade--sixteen hours a day, six days a week--to help get it where it is today.

And where is it? Over its ten-year history, Cypress has generated over $1 billion in cumulative revenue, made over $160 billion in profits on which we paid $60 million in taxes, created 1,500 jobs which paid cumulative salaries of nearly $500 million, on which our employees paid further taxes of $150 million. We have shipped cumulative exports worth $300 million. We have generated a market value of $500 million for our shareholders and employees--all of whom own stock in the company.

If that is an "excess of the 1980s," let's have more! As an entrepreneur, I should not have to apologize for my success and that of my company. I am offended by the administration's divisive rhetoric. As we debate the virtues of raising taxes on individuals and corporations, let's not debate abstractions. Let's debate the realities of who pays taxes and the impact of raising taxes on those people and companies.

I don't want sympathy. But I do want to expose the shaky foundations of the logic behind the administration's program. I am a person of simple tastes; therefore, I still have most of the wealth associated with my Cypress shares. What have I done with that wealth? I invested it. In fact, I invested it in precisely the kinds of companies on which the administration wants to shower taxpayer subsidies--the world's most advanced competitors in fields such as semiconductors, biotechnology, software, networking, environmental sciences, and health care.

Attached to my testimony is a list of 99 companies in which I hold investments through my participation in three venture-capital funds. Nineteen of those companies are innovators and leaders in high-speed data communications--real companies that are making real components of today's existing data superhighway. Other companies are innovators in the field of high-performance computing--including MasPar.

Every incremental dollar that Washington takes from me comes directly out of my investments in these companies. I cannot sell my house or car or cut my food bills. But I am going to be forced to invest less. After all, the cash to pay my higher taxes has to come from somewhere. In essence, the administration is arguing that by taking my money in the form of higher taxes and "investing" it in subsidies, it can make better investments--more jobs and wealth--than the venture-capital firms with which I invest--firms that are the envy of Japan and Europe. That logic defies common sense. Does anyone believe that Washington invests more effectively in high technology than the free market?

Last month, President Clinton and Vice President Gore visited Silicon Graphics, one of the great new-generation computer companies in Silicon Valley. By all accounts, they were amazed by what they saw. They declared their eagerness to help produce "more successes like Silicon Graphics."

I own shares in Silicon Graphics. It exists because hundreds of institutions and individuals like me--excesses of the 1980s--put their money into the company through venture capital. Washington cannot create more companies like Silicon Graphics. The way to create more Silicon Graphics is to allow knowledgeable investors, steering their money through world-class venture capitalists, to try to fund just the right companies with just the right technologies at just the right time. Even these venture experts are not right all the time. But surely they are right more often than Washington.

The Clinton plan also raises taxes on Cypress as a corporation. Suppose, as a result of the plan, that Cypress's corporate taxes increase by $1 million next year. As CEO, my only choice is to take that money directly out of R&D--the lifeblood of the organization. Again, let's be clear about the logic: A tax increase of $1 million means that Cypress will employ ten fewer Ph.D. technologists than it would otherwise--technologists that would be working on high-performance chips for data superhighways and supercomputers.

CLINTON PROGRAM MISCONCEPTION THREE: GOVERNMENT SUBSIDIES PRODUCE TECHNOLOGIES EFFICIENTLY

The third misconception about the Clinton plan brings us directly to high technology, high-performance computing, and the politically named "electronic data superhighway." This misconception is ideological rather than political--but it is no less dangerous. It is the proposition that the best way for good ideas to become realities in the market is for government to subsidize them. There may well be, under certain limited circumstances, a legitimate place for government as a customer of last resort for high technology. But the administration wants to make government a customer of first resort--to the tune of tens of billions of dollars.

We have been down this road before. In the 1960s, we had a government that threw money at social problems. It didn't work. In the late 1980s, under President Bush, we had a government that threw money in many of the same directions that the Clinton program is now proposing. It didn't work; it brought us a meager one percent growth rate for four years. In the 1990s, we have a government that wants to throw even more money at such opportunities. It won't work.

From a purely economic point of view, it is a mistake to discuss the "Reagan-Bush" era. As the non-partisan Cato Institute points out, an independent "bean counter" would conclude that the Bush administration deviated from the conservatism of the Reagan administration and that the Clinton administration is going even farther in the direction Bush started. "Reagan-Reagan" and "Bush-Clinton" are better approximations of economic reality.

ONCE MORE: THE CASE FOR FREE MARKETS

What does work? The ragtag, unmanaged, sometimes-painful melee of the free market. It's not pretty, it's not neat, but it is what has made the United States the world's technology powerhouse.

Consider my personal investments. Via the venture-capital firms with which I deal, I am invested in Jeff Kalb's company, MasPar, a massively parallel computer manufacturer. On the other hand, venture capitalist John Doerr of Kleiner Perkins Caufield & Byers, another firm with which I invest, believes that the conventional (non-parallel) supercomputing field is heading for a fall. His view: "The supercomputer industry is collapsing on itself. It is not a competitive way to solve problems. Why should we invest in dying markets?"

I hope Jeff Kalb succeeds with MasPar. But there is certainly plenty of risk in the investment equation. The best place to sort out that risk is the free market, with dynamic real-time decision-making--not with government programs that often take as long to implement as it takes for major technologies to run their life cycle.

The administration is not alone in putting its faith in Washington over the free market. A few weeks ago, Massachusetts Representative Edward Markey, who chairs the Subcommittee on Telecommunications and Finance, addressed an important computer-industry conference. First he complained that the industry did not lobby enough. I guess those computer executives were just spending too much time back home starting companies, creating products, entering new markets, creating wealth and opportunity. (Representative Markey's comment prompted me to dust off and modify a Vietnam-era phrase: What if they gave a subsidy and nobody came?)

Representative Markey went on to urge the executives to devote special attention to the data superhighway. His reasoning: "This is too important to be left to the invisible hand of the marketplace." I was amazed. That socio-economic experiment was tried once--from 1917 to 1989--and it failed. Which government-sponsored technology advances would Representative Markey like to compare to the embarrassment of riches generated by the "invisible hand"?

Today, in industry after industry--semiconductor chips, computers, biotechnology--U.S. companies lead the world or are mounting remarkable comebacks against Japan and Europe. Why are we moving forward against our foreign rivals? Because we relied on Darwinian competition--the invisible hand--while Japan and Europe relied on government targeting and subsidies.

Think about Japan. Just a few years ago, America was in panic about the Japanese government's massive research program in high-definition television, or HDTV. Today, everyone agrees that Japan's effort was a multi-billion-dollar flop and that America has prevailed--thanks to the messy, uncoordinated innovations of many private companies. Five years ago, we lived in fear of the Japanese Fifth Generation Computer Project. "Tron" was going to walk, talk, and eat our lunch. Today, everyone agrees that it too was a flop and that U.S. computer manufacturers are continuing to extend their global lead.

Or think about Europe. Amazingly, we still have "experts" who want us to emulate Europe's alphabet soup of technology consortiums such as JESSI, their equivalent of the U.S. chip consortium Sematech. JESSI showered billions on the European semiconductor industry. It also "rationalized" the industry by allocating certain market segments to various companies. Siemens became the DRAM company for Europe--and has since gone out of the business. Philips became the SRAM company for Europe--and has since gone out of that business.

After inadvertently weakening its chip industry, Europe then established 14% import duties on foreign chips--the next logical step of desperate government policy. The import duty had precisely the effect we might expect: It raised the price of components to the European computer industry and virtually wiped it out as well. Today, there is no European chip industry or computer industry to speak of--thanks to the role of government programs like JESSI. European taxpayers gave up part of their income to wipe out two critical industries! We can't afford to emulate such failed experiments.

Industrial-policy advocates like to describe a few high-profile cases as success stories. Let's consider two of them. The first is Airbus. It is true that European governments have unfairly subsidized their aircraft manufacturers and gained a large share of the world market. But at what cost? European governments have spent $26 billion on Airbus--to directly create 40,000 jobs. That works out to $650,000 per job. The National Venture Capital Association estimates that venture-funded startups generate employment at the rate of $45,800 per job. If Airbus represents industrial policy at its best, is it any wonder that Europe has experienced such chronic unemployment?

SEMATECH: A BAD EXAMPLE

The second alleged "success story" is here at home. It has become fashionable to describe Sematech, the chip-industry manufacturing consortium, as a triumph of industrial policy. It is not. I have been called to Washington three times in the past few years to testify about the consortium and other proposals to prop up the U.S. chip establishment. Each time, I have argued that most of the U.S. semiconductor industry is innovative and healthy--that Washington should not equate the struggles of a few giant companies in the mid-1980s with the fortunes of the entire industry.

What has happened? America again leads the world in semiconductors. In each of the last three years, the U.S. has won back market share from Japan. In 1992, our worldwide share actually exceeded Japan's share for the first time in a decade.

Sematech has waged a public-relations campaign to claim credit for the comeback. It is a preposterous claim. It is true that Washington spent or will spend $1 billion of taxpayer money on the Sematech boondoggle. It is true that the U.S. semiconductor industry has experienced a resurgence. It is not true that one had much to do with the other.

The U.S. semiconductor industry resumed its leadership position because a new generation of chip companies forged in the 1980s--a generation of which Cypress is just one example--created innovative new categories of chips, new business models for how to compete, and new levels of efficiency. In short, we out-innovated the Japanese. America also resumed its leadership in great part because the giant companies of our industry--companies like Intel and Motorola--realized that they had to adapt or become also-rans. They adapted--and now lead the world.

Consider Intel. In the mid-1980s, Intel was losing money and in crisis. In 1992, Intel had become the largest and most profitable chip company in the world. I am afraid to imagine what would happen if I cornered Intel's CEO, Andy Grove, at a cocktail party and said, "I am glad Sematech saved your company and turned Intel from a loser into a winner. I guess you ought to thank Uncle Sam." Of course, it was Intel's remarkable innovations in microprocessor architectures, along with top management's refusal to accept also-ran status and lots of hard work by thousands of men and women, that allowed it to capture the lead.

Sematech has made little contribution to America's comeback in chips--certainly no contribution worth $500 million of taxpayer money. Indeed, sometimes this money was downright counterproductive. In the 1980s and early 1990s, Sematech spent taxpayer dollars to develop equipment for the exclusive use of its members. Thus, taxpayer money, earmarked to help the U.S. semiconductor industry, was being used to hurt the 90% of American chip companies that were not members of Sematech--to hurt the very industry it was meant to help.

William Spencer, the new president of Sematech, has cleaned up the worst abuses. But Sematech's return on the taxpayer dollar is still very low. Five years ago, when it was created, Sematech's membership included only 14 giants out of the hundreds of America semiconductor companies. In the last year, two of the member companies (the two most entrepreneurial companies, LSI Logic and Micron Technology) left the consortium. Two more companies are reported to be considering departures. The simple lesson: Relentless competition and fast-paced innovation saved the U.S. chip industry--not taxpayer subsidies.

Surely this is no surprise. Life in Silicon Valley is a daily sprint; government moves at a crawl. Do any of us really believe that Washington can play a decisively positive role in fields as complex as semiconductors, high-performance computers, or electronic data superhighways?

Think for a moment about the realities of life at Cypress and then extrapolate it to the chip industry and Silicon Valley as a whole. Our company has 150 product designers. We have over 70 technologists. We sell more than 1500 products. We are working right now on 50 different new products--from high-speed computer memories to data communications chips. With my technical training and my managerial background, it takes me 16 hours a day to stay on top of this organization.

Cypress is but one $250-million company in a $50-billion semiconductor industry. Thus, if you take the details I just described and multiply it by 200, you have a sense of the complexity of the chip industry. If you take that level of complexity and multiply it by another factor of ten or more, you have the complexity of Silicon Valley. How can the government possibly hope to cope with the details of Silicon Valley? How could the government even know who the players were in any week, let alone pick winners and losers?

Now think for a moment about something less complex: the tobacco leaf. Today, the U.S. government spends tens of millions of dollars through the Office of the Surgeon General to warn Americans about the dangers of smoking. At the same time, through loan guarantees and occasional direct grants from the Department of Agriculture, it has spent tens of millions of dollars to subsidize tobacco farmers.

If government cannot figure out whether to discourage smoking or to subsidize it--and if it spends taxpayer money to do both--how can it architect a coherent policy about technology choices in fields as complex as optical fibers, wireless data communications, or high-performance computers? If, after several decades, it cannot make a rational decision regarding the tobacco leaf, how can it make winner vs. loser decisions in Silicon Valley, where the game changes weekly?

TAX-AND-SPEND--BUT WITH A HIGH-TECH TWIST

How, then, do I evaluate the administration's multi-billion-dollar technology plan? Put simply, it is classic failed tax-and-spend economics with a new coat of paint and new jargon. Bridge-and tunnel pork-barrel programs may have been replaced by high-technology pork-barrel programs--but it is pork-barrel just the same.

I want to be fair. The administration is beginning to move the government toward a high-technology vision, a development that I view as favorable. I certainly share its enthusiasm for an America in which computers and communications carry data and video into companies and schools, and eventually into the home. But why does the administration want to spend tens of billions of dollars of taxpayer money to fund technology programs that the free market would pay for without one cent of expense to the taxpayer?

For example, the administration proposes big spending increases on a range of projects to speed creation of the "data superhighway." The technology plan outlined by the administration offers few substantive details about these projects. But the details it does offer make one point clear: Everything in the program is already funded by venture capitalists and in development by innovative private companies, many of which I personally support.

Multiple, competing data highways are being built day-by-day, company-by-company, across the United States. Entrepreneurs are racing to develop new networking systems, new software interfaces, new value-added services. Gil Amelio, whose comments I cited earlier, was a senior executive at Rockwell before he joined National Semiconductor. Gil reports that his division at Rockwell created the hardware to lay 23 million miles of fiber-optic cable!

MCI, AT&T and Sprint already have three independent, coast-to-coast, fiber-based long-haul networks. The real issue is extending those networks into the home. There is a role for government in this, but it is not to spend billions of taxpayer dollars on a field that the private sector is willing to fund. The role for government is to untangle the morass of bureaucracy and regulations that prevents private companies from hooking up the "last mile" of fiber to the home.

For example:

The regional Bell operating companies would gladly hook fiber optics from the long-haul network to the home. But they are prevented from doing so by regulations that make the huge capital investments uneconomical.
Cable operators are already hooked into 60% of American homes. They too could make the connection with existing long-haul data superhighways, but they are prevented by regulations that declare them a "natural monopoly" and restrict them to the television and movie business.
Finally, the long-haul superhighway could be hooked to the home through wireless circuits. But the frequencies required are currently being held up by the Federal Communications Commission.

Washington does not need to "help" by spending billions of dollars on data superhighways. These highways will reach the home for free if government becomes less unhelpful--if it gets its regulatory house in order and then gets out of the way. Even the Berkeley Roundtable on the International Economy (BRIE), the liberal think tank that contributed Laura Tyson to head the President's Council of Economic Advisors, agrees that one centralized data superhighway--in its words, "the all-singing, all-dancing, all-integrated broadband network"--is not how computers and communications will come to the home. What we need from Washington are common-sense rules and a sense of the limits of government activism.

The same logic applies to high-performance computing. The prevailing image of supercomputing in Washington remains tied to the Cold War: giant machines, funded by the Defense Establishment, deployed to design nuclear warheads or track thousands of incoming missiles. But the Cold War is over. And we may be at the beginning of the end of the era of high-performance computing, at least as it has been conventionally defined. We have seen the limits of gigantic "number crunchers" of the kind developed by Seymour Cray. We are seeing the arrival of smaller machines, massively parallel machines, machines with new architectures designed for ease of software development rather than raw number-crunching capability.

In short, the high-performance computing industry is entering a period of profound change. This is precisely the time when venture capitalists are most effective at setting new directions and funding new players. Heavy-handed intervention from Washington is guaranteed to retard change rather than to encourage it. Venture capitalist L.J. Sevin, the former chairman of supercomputer maker Convex, puts it this way: "The supercomputer market is changing dramatically because the nature of the technology is changing. The only thing government can do is get in the way of that change--with disastrous results."

A PROPOSAL FROM THE PAST: A RISING TIDE LIFTS ALL BOATS

So much for what not to do. Let me offer several constructive alternatives. Please forgive me for not having a convenient, five-point program to enhance America's economic leadership. In fact, I don't believe in such programs, because I understand first-hand the messiness, unpredictability, and rapid changes in high-technology markets. As illustrated here, well-intentioned industrial policies usually create damage, even though their architects want to help.

Washington should stay away from the intricacies of high-tech competition--whether the issue is the data superhighway, high-performance computing, or advanced manufacturing. It should focus instead on the infrastructure of competition--those factors of production that help all companies equally. President Kennedy said it best: "A rising tide lifts all boats."

Members of the Clinton administration love to give speeches about infrastructure: deteriorating roads, falling bridges, aging railroads, even data superhighways. But the real threat to our country's economic future is the condition of our financial infrastructure--the scarce supply of reasonably priced capital that successful companies need to build their manufacturing muscle and to create new jobs. If Washington adopted policies designed to lower (and keep low) the cost of capital and extend the time horizons of investors, it would make a genuine contribution to America's entrepreneurial advances in the 1990s.

Washington can take two steps to restore our financial infrastructure. After years of the most reckless fiscal behavior in American history, federal government must get serious about reducing wasteful spending. And it must get serious in a hurry.

The Clinton administration has boasted about its proposed spending cuts and its reduction in the White House staff. But it has not begun to approach the dramatic cuts we need. You can pick up a newspaper any day of the week and read about the painful efforts of giant companies--General Motors, IBM, Sears--to slash their costs, streamline their overhead, reduce their payrolls. Boeing recently announced plans to dismiss 20 percent of its workers--and Boeing is a world-class company by any standard. My company, in a decision of great personal pain to me, recently dismissed 20% of our workers to become more efficient--and we started out as one of the leanest and most efficient chip companies in the United States.

How then, in light of these sacrifices, can the administration trumpet its plans to eliminate 100,000 federal jobs over the next four years--less than 5 percent of a total workforce of over two million? Does the administration expect us to believe that if a company as great as Boeing can figure out how to survive with 20% fewer people, the Department of Agriculture cannot live with an immediate reduction in headcount of 25%? Let's lay off the agricultural station head in Fairfield, Connecticut, who mentioned in the Wall Street Journal that she was looking for one corn subsidy customer!

The opportunities for cuts go beyond headcount. We could spend hours listing wasteful and unnecessary programs--programs that may have made sense 30 or 40 years ago, when they were created, but that make no sense today. We could begin with two of my personal favorites: the federal wool-uniform subsidy, created during World War II and later expanded to include mohair sheep (for no reason other than pork-barrel); and the Interior Department's strategic helium reserve (created during World War I), which is keeping America well stocked for the next blimp war at $120 million per year. Why are we asking for more money from America's hard-pressed people and companies when follies like this exist?

I believe there is only one way to impose real spending discipline: Congress must pass a balanced-budget amendment and approve the line-item veto. Let me restate the point: No industrial policy or national investment strategy or trade negotiation, no matter how brilliant, would have the same energizing effect on American entrepreneurship as a return to fiscal sanity through dramatic spending reductions. It is a prerequisite to sound economic leadership.

Washington should take a second important step. It should restructure the capital gains tax to penalize short-term speculation and encourage long-term investment. The Clinton administration has proposed a modest reduction in the capital-gains tax targeted to tiny companies. The proposal is good political symbolism, but it does not address the real problems of real entrepreneurs.

We already have plenty of venture-capital funding for start-up companies. No good engineer in Silicon Valley will be refused his or her first $10 million to prove the feasibility of a new technology or product. The problems start at the next stage, when a company needs its second or third $10 million to build a plant and acquire real manufacturing muscle. This is the stage when young companies must turn to the public capital markets--or to cash-rich foreign operations eager to acquire valuable American technology at bargain-basement prices. It is here that they face the most serious barriers. The best way to reduce the barriers is to enact a sweeping restructuring of the capital-gains tax.

Unfortunately, a majority of Americans have been convinced that any reduction in capital gains is a "tax break for the rich." That's why we should increase taxes on assets held for less than six months by imposing a surcharge over the nominal capital gains rate. This surcharge would be a stiff disincentive to unproductive speculation. It would allow us to argue, rightly, that we plan to "soak the rich"--or at least the rich who play hunches, trade on rumors, and churn investments to the detriment of the American economy. In return for this penalty on speculation, capital-gains taxes on assets held for more than three years, assets that help build America, would be totally eliminated. Assets held between six months and three years would be taxed at the current income tax rate.

So there you have it. A program for American renewal that revolves around immediate and dramatic cuts in government spending, a balanced-budget amendment, and a restructuring of the capital gains tax. Hardly the stuff of brass bands and whistle-stop tours, especially with all the energetic talk these days of critical technologies, high-speed data communications, and high-performance supercomputing.

I wish I could propose a razzle-dazzle plan, complete with spectacular computer graphics, to capture the imagination of Washington and put America back on track. But those of us who are out competing every day understand that there are real limits to Washington's potential contributions to our success--and countless opportunities for mischief and missteps.

America has plenty of work to do on the economic problems we have created for ourselves--problems that trace their roots to fiscal recklessness. Ultimately, though, the economic battles of the 1990s will be won in America's factories, labs, and offices--not in the halls of Congress or the corridors of the Commerce Department. That's good news. America's entrepreneurial companies have the guts, brains, and drive to beat the best the world has to offer. All we need from Washington is the confidence to let us fight it out.

The following is a list of 99 investments I made in high technology companies directly or through venture capital firms. The asterisk identifies the 19 companies specializing in data communications or high speed computers.

COMPANY
3DO
Alantec
*Amylin
Amylin Pharmaceuticals
Applied Micro Circuits
Appsoft
Aptix
Arbor Development
Arris Pharmaceuticals
Ascend Communications
*Ask Computer
Aspect Development
Auspex
*Avid Technology
Axion Pharmaceuticals
Banyon
Biosite
BIOSYS
Business Insurance
Cadence
Camino Labs
Cardiac Pathways
Cellular Data
*Central Point Software
Chemtrak
Citrix Systems
*Coactive Computing
*Cocensys
Communications Ventures
Conductus
Covalent Systems
Crescendo Communications
*CV Therapeutics
C-Cube Microsystems (V)
*Datalogix
Delineate
Domestic Automation
*D.S.I. - David Systems
*Endosonics
Endovascular Technologies
EO Computer
Evernet Systems
*Farallon
*Genetrix
GenPharm International
Global Village
*GMS
GO Corporation
Great Lakes Environ. (GLES)
Harmonic Lightwaves
*InSite Vision
Intuit
IST
Kera Vision
Labintelligence
Ligand Pharmaceuticals
Logic Modeling
Macromedia
Magellan
Maspar Computer
Microchip (SCIV)
Microtec
Ministor Peripherals
Miris Medical
Molecular Dynamics
nCHIP
Networking Computing
Notable Technologies
*ON Technology
Ontogen
Oxford Glycosystems
O.B. Tech
Paramed Technology
Pharmacylics
Plus Logic (SCIV)
Power Integrations
Progress Software
Quicklogic
Quickturn
Radius
Rambus
Raster Graphics
Redwood Design Automation
Relevant Technologies
Ribogene
Rose Communications
S3
Sangstat Medical
Shaman Pharmaceuticals
Shiva
*Shographics
Sierra Semiconductor
Slate
Southwest Network Services
*Storm Technology
*Tivoli Systems
Total Pharmaceutical (SCIV)
Unisurge
*Vitesse (SCIV)

My Stream of Consciousness

Thursday, July 8, 2010

Infotainment

Connection

On Structure & Function

Simpson Paradox

TJ: Cypress

Women Needed in Finance

That Feeling!