Participants’ Roundtable—Where Do We Go from Here? Policy Issues?
William J. Raduchel
Dr. Raduchel opened the roundtable by voicing his hope that the attendees would, above all else, take away from the conference an understanding of the uniqueness and complexity of software, as well as of the ecosystem that builds, maintains, and manages it. It is because one does not “really know what’s happening through that whole ecosystem that brings down the software,” he said, that an issue like offshoring “gets so complicated so quickly.” The day’s second most important point was that the way software was created, maintained, and distributed had not changed in 40 to 50 years. It should come as no surprise, therefore, that other countries had learned the prevailing techniques and become competitors of the United States. Unless there were investment that fundamentally changed the ways in which software was created, the industry would have to be considered mature even though what it produced was at the cutting edge of advanced technology.
Dr. Raduchel then asked each member of the panel to take two minutes to identify the one issue relating to software that he would put at the top of a public-policy agenda for the United States. He called upon Wayne Rosing of Google to go first.
Dr. Rosing related that he had posed the following question to four or five friends of around his age who had lost jobs in the dot-com bust: “When was the last time you taught yourself something new—that you really took the trouble to learn a new field?” The initial response, he said, was most frequently “a blank stare,” followed moments later by an answer of 10, 12, or 15 years ago. That many individuals did not stay current was one of the biggest single problems the United States faced, because in a global economy those who do not keep their skills at the cutting edge are inevitably sidelined. What the country needed was a system that created extraordinary incentives for people to take charge of their own careers and their own marketability. Much change would be necessary because there was a lot that did not work well; the colleges, for example, were designed for young, full-time students. Awareness would have to be built, but the situation was not without opportunity.
University of Texas at Austin
Noting that the day had begun with “lofty questions about ‘What is software?’ ” and concluded with “a somewhat tense discussion” of offshoring, Dr. Flamm proposed to offer four points tracing the route from one to the other. The first harked back to the history of the semiconductor industry, which became one of the United States’ first to go offshore when, in the mid-1960s, virtually every manufacturer moved the unskilled, labor-intensive assembly part of the production process to Hong Kong. A decade later the only assembly that remained in the United States involved relatively short runs of very specialized product that had to be fabricated close to the market. Meanwhile, the commodity production had stayed in Hong Kong only until wages for unskilled labor were bid up there, when it moved to Korea. The same pattern was repeated several times: Production moved on to Taiwan, then to the Philippines, then to Malaysia, each time staying only until wages had gone up. Even in an area in which one could expect a relatively undifferentiated labor input going into the production process, the supply of unskilled labor was not infinitely elastic.
Second, Dr. Flamm maintained that the offshoring of software services in many respects greatly resembled the semiconductor story of the 1960s, the exception being that very skilled labor rather than unskilled labor was now involved. Central to the movement of semiconductor manufacturing abroad had been the growth of the air-traffic and telecommunications infrastructures. Transport costs are very low for the semiconductor because of its small physical size and light weight; and, by the mid-1960s, telecom had developed to the point
where it was possible to coordinate production abroad and to do so in a cost-effective way. “Similarly,” he observed, “if ever you are going to find a service that can be offshored economically, it’s going to be software, which is not really a physical product but just bits flowing through wires.”
WHY THE ABRUPT EMERGENCE OF THE OFFSHORING PHENOMENON?
Third, and with this in mind, Dr. Flamm asked why offshoring seemed to have suddenly swept the software industry in the previous 2 or 3 years. Why hadn’t this happened 5 years before? He speculated as to two potential answers. The first was that major investment in communications infrastructure outside the United States might have equalized the playing field, helping make this possible for the first time. The second was that the skill of the U.S. labor force had declined relative to that of its competition. He offered his impression as a member of a university faculty that the nation had not truly been investing in its educational infrastructure, and he asserted that the California public school system of the 2000s was a shadow of that “glorious and wonderful” system that had “invested all kinds of resources” in him when he went through it in the 1960s. He suggested that the diminution of resources available to public education in California had been replicated across the country, and that it had had an impact on the relative attractiveness of hiring U.S. workers and their competitors in other countries.
Fourth, the semiconductor industry had become subject to another round of competition in the 1980s, this time from other advanced industrial economies in a process very similar to that affecting the contemporary software industry. In this second round relatively skilled jobs rather than unskilled jobs had been at stake. The U.S. semiconductor industry had ended up coming back, and while part of the reason was that it had invested in resources, the bottom line was that it never returned to the commodity products it had gotten out of. It had instead gone into design-intensive products and stayed ahead.
Dr. Flamm drew the following policy prescriptions from this story:
“Don’t expect to get back into what you’ve been pushed out of.” The rest of the world has competence, and those nostalgic for the old monopoly should remember that the rest of the world had caught up and refrain from looking back.
“Invest in the things that our new competitors invest in,” which make economic development more accessible for all to participate in. While he was skeptical of the widespread notion that an infinitely elastic supply of top-notch university graduates was coming out of the schools of China and India, he argued that if the United States really wanted to compete, it had to invest in education. Likewise, to make less-industrialized areas of the country more attractive and to keep workers from the departing industrial sector employed, domestic investment in communications infrastructure would be necessary.
Ernst R. Berndt
MIT Sloan School of Management
Seconding Dr. Varian’s earlier point that not all advanced technological sectors were behaving alike, Dr. Berndt observed that while enrollment in MIT’s electrical engineering and computer science departments was indeed down, that was not true of fields related to biotechnology, such as biology and chemistry. At 30 years old the PC industry was, after all, no longer sexy but mature, and the action seemed to be elsewhere. Novartis was moving 900 scientists from Switzerland to Cambridge, Massachusetts, which he saw as part of “a Schumpeterian process that we just have to recognize and try to exploit.” One of the best investments in this regard would be in getting better data.
Senate Committee on Banking
Congressional staffers like himself, Mr. Socas stated, had been hearing every day from people who were losing their jobs and who could not afford the luxury of retraining themselves, as they were working two jobs or caring for small children or a sick parent. Such people reported seeing U.S. companies send offshore training and capital, resources that “used to stand behind U.S. workers”; as a result, workers abroad were being armed with all the skills that formerly had allowed U.S. workers to command the world’s highest wages. “U.S. workers were paid more not out of the kindness of the heart of companies,” he declared, “but because they produced more”—an ability now being transferred to their counterparts in other countries.
The “relentless race for profits” might be forcing U.S. companies into this, Mr. Socas conceded. But he recalled the phrase “what’s good for General Motors is good for the country,” which he interpreted as a “bold—and at the time arrogant but valid—statement” of GM’s intention to invest capital and technology in the community of Detroit and its workers “to help them be the best that they could.” In return, the U.S. government gave General Motors tax benefits for R&D and other advantages, such as the benefits of a strong public education system, and as a consequence both General Motors and the United States prospered. The substance of the complaint that the American workers had been bringing to Capitol Hill was: “It seems like this whole deal has been broken.” Since they understand how the world works, being familiar with similar stories from agriculture and manufacturing, they were not so much objecting to the loss of a specific job as to what they saw as evidence that “the game has changed.”
FREE TRADE AND PROTECTIONISM IN U.S. HISTORY
Such economists as Gregory Mankiw, chairman of the President’s Council of Economic Advisers, had reached the conclusion that this change was a good thing, Mr. Socas said. Americans have been raised to believe that free trade is good for the economy in that it will cure displacement and many other ills. But because the country had endorsed free-trade policies over the previous half-century, an important fact was being left out of account: that its history is in fact one of protectionism, that McKinley ran on a protectionist platform, that the Civil War was fought over protection and tariffs.
Free trade is a wonderful policy nevertheless. Offshoring, however, is not free trade, and the current landscape was very different from the one that David Ricardo envisioned when he put forward his Theory of Comparative Advantage. In a nutshell, this theory states that each country should use its internal cost ratios and direct its productive resources to what it does most efficiently. This does not mean that a country needs to produce something at the lowest cost around the world; simply, a country must do something well and direct its resources thereto. The consequence will be that, through trade, world output will increase and, with it, the economic pie.
SEEKING COMPARATIVE ADVANTAGE OR ABSOLUTE ADVANTAGE?
But there is an assumption in Ricardo that, Mr. Socas claimed, nobody ever talks about: that the factors of production remain immobile. If instead they can leave a country, then they will not go from South Carolina to California but from South Carolina to Guangzhou Province or from South Carolina to Bangalore. They will no longer give the United States comparative advantage but will chase what Ricardo called “absolute advantage.” In his famous example of wine and wool trading between Portugal and England, Ricardo lays this out very clearly. But Ricardo notes that the “factors of production” would never leave England, for no English capitalist would ever invest so substantially in another country out of loyalty to his mother country and out of fear for the security of his investments overseas.
Expressing skepticism that anyone in attendance believed this economic patriotism still applies today, Mr. Socas declared: “It certainly didn’t apply in my life when I was in the world of investment banking, it certainly doesn’t seem to be applying with those who talk about corporate strategies today.” Even though that fundamental assumption was no longer valid, advocates of free trade continued to fall back on Ricardo to defend their position. This raised two questions: Was offshoring something old, meaning the latest chapter in the history of free trade, or something brand new, the first manifestation of a globally integrated economy? And, if the latter were true, were the nation’s policies—reflected in everything
from the way the Bureau of Labor Statistics collected data to the way corporate accounting was handled—pinned to a vision of trade oriented toward a national system, when in fact companies were oriented towards a global economy? He called the latter “the central political question of our time.”
WHAT DOES “PROTECTIONISM” AIM TO PROTECT?
Mr. Socas reiterated that he had observed a high level of grass-roots concern about these issues, adding that “average people” believed politicians to be exploiting them. The negative epithet “protectionist” obscured the question of what was being protected—the answer to which, he asserted, was the American system and the American community. He rejected a comparison between current reservations about free trade and the motivations for the “Buy American” campaigns of two decades before, characterizing the latter as efforts to shield Ford from its failure to invest in just-in-time manufacturing or to reward “workers that [had] lost their way.” Now at issue was giving U.S. workers, who had suddenly been exposed to many new competitors worldwide, a shot at trying to compete. They needed the skills that would allow them to continue to command the high wages and opportunities “that, frankly, the whole structure of our country is founded on.”
Rochester Institute of Technology
Declaring himself free of the orthodoxy shaping debates surrounding the issue of “free trade vs. protectionism,” Dr. Hira observed that both terms were somewhat vague and that they tended to be “situationally implemented.” As to the specific issue of software, he recalled the 1980s question of whether manufacturing matters, asking whether software matters. “If it does,” he stated, “we need to think about how to go about making software a viable profession and career for people in America,” since the field did not appear sufficiently attractive at present. He expressed his impatience with what he termed “the old stories of ‘We just have a bad K-12 math and science education’ ” as the reaction to a call for policy alternatives. While educational improvement was in order, it was also important to think through and debate all possible options rather than tarring some with a “protectionist” or other unacceptable label and “squelching them before they come up for discussion.”