part0006

Industrial development and diversification are about taking risk. Encouraging the growth of industry in new and unfamiliar sectors is always risky. It invariably involves costs, and the rewards are uncertain. Yet it is important to consider different kinds and levels of risk in industrial upgrading. Some endeavors are relatively less risky than others. Some bets are safer than others. Potential payoffs differ as well, and the expectation of any payoffs at all is uncertain. Generally speaking, higher-risk propositions yield higher rewards, though the likelihood of success is much lower. Science-based industries and the prospects of science-based industrialization are the riskiest enterprises of all, extraordinarily high-stakes bets. Cutting-edge innovation—that is, creating something new—requires massive amounts of resources committed to a wide range of actors and R&D activities with the faint and distant hope that such efforts are technically and economically productive. Life sciences innovation and biotech commercialization are not just risky; they are inherently uncertain enterprises.

This chapter distinguishes between risk and “primary uncertainty.” Typically, risk and uncertainty are thought to be synonymous, as both are defined by the absence of certainty. They both describe circumstances in which there are unknowns. However, I seek to make a more fundamental conceptual distinction between risk and uncertainty, to suggest that they are in fact distinct and thus bear different implications with respect to the organization of political economies. Risk scenarios are those in which decision makers have some prior knowledge about potential outcomes. For instance, decision makers may have some sense of the odds of a particular outcome or the range of viable choices available. Decisions are risky, but they are not uninformed. Situations of primary uncertainty, in contrast, are those in which actors possess so little (or no) information or knowledge about a decision and its consequences that they must essentially make what are uninformed choices. Put another way, whereas risk taking involves calculated if uncertain decisions about alternative choices, the notion of primary uncertainty suggests a more ambient context in which decision-making processes are obscured by the lack of alternative choices, and even before that, the information and knowledge with which to generate clear and viable alternatives.1

This chapter sharpens the conceptual distinction between risk and uncertainty by examining the processes of industrial upgrading in East Asia during the postwar period; it then develops an analytical framework with which to make sense of the current processes of growing biotech industries in Korea, Taiwan, and Singapore. The core argument is that the postwar political economy in Asia was one in which the state strategically intervened in the economy for the purposes of mitigating the risks of industrial upgrading. The developmental state could not guarantee industrial success, but it demonstrated the capacity to reduce the risks of upgrading and technological “catch-up,” compelling firms and industry to move into sectors they otherwise would have eschewed if left on their own. The strategy of mitigating risk is less effective, however, when it comes to cutting-edge biotech. In the current state of the biotech sector, there remains such tremendous technological, economic, and long-term uncertainty that decision makers can, at best, only strive to manage and cope with such uncertainty. The distinction I draw between risk and primary uncertainty, therefore, is not one of semantics; it is intended to illuminate how decision makers make political economic bets and how they rationalize them in the face of two different modes of industrial upgrading, each with, I assert, its own set of challenges .

The developmental state is credited with being the main engine for postwar Asia’s rapid and dynamic economic growth. It set about creating comparative and competitive advantages to gain credibility in industrial sectors that had been dominated by firms in advanced nations. The model was about using state aid to develop private sector firms, transforming them into national champions and global players. Inspired by Friedrich List’s theories of nationalist economic growth, the developmental state in Asia cultivated, often through authoritarian means, nationalist myths—in postcolonial Korea, self-reliance; in Taiwan, the return to China; in postcolonial Singapore, paranoia about national survival—to deepen aspirations for social and economic modernization. Asia’s developmental states were motivated by lofty “transformative” goals.2

The bureaucracy was at the center of the developmental state model. Meritocratic recruitment ensured that the state apparatus in Korea, Taiwan, and Singapore drew from among the nation’s “best and brightest.” Bureaucrats were not only talented; they were well positioned in hierarchical organizations structured along clear lines of authority. The verticalization of power was afforded in part by technocrats’ dependence on the authority and patronage of the ruling political elite. But the organizational structure of bureaucratic institutions contributed as well to hierarchical forms of top-down state management. Economic pilot agencies such as Korea’s Economic Planning Board (EPB), the Council for Economic Planning and Development (CEPD) in Taiwan, and Singapore’s Economic Development Board (EDB) coordinated the strategies and activities of government agencies responsible for industrial growth. Because few actors were involved in economic policymaking, power and authority were not only verticalized but centralized and concentrated. A small number of powerful actors allowed technocrats to resolve coordination problems within the state.3 The institutional organization of the developmental state ensured that it remained autonomous and insulated from narrow societal interests below. It also allowed state-level decision makers to behave decisively and authoritatively.

Peter Evans reminds us, however, that the coordinative capacity of the state was not solely a function of its political structure and autonomy but also a consequence of its strategic “embeddedness” within productive sectors, most notably industry.4 As the literature on the developmental state goes to great lengths to emphasize, market forces were the basis of industrial growth in Asia. State and industry cooperated in the implementation of industrial policies. Information feedback loops between them were institutionalized, formally and informally, allowing the state to strategically allocate vast resources in order to coordinate industrial activity more generally. The Korean state thus “husbanded” the chaebols. The Kuomintang (KMT) party-state in Taiwan ensured that government ownership of firms in key infrastructural sectors facilitated rather than crowded out private sector industry growth. The state in Singapore was instrumental in growing government-linked corporations in tandem with the presence of multinational firms. The state’s embeddedness within society did not mean the devolution of state power, however. Rather, Evans’s point is that the externalities gained from the state’s strategic alliance with industry actually reinforced its market-regarding role in coordinating industrial development in postwar Asia.

Theories about East Asia’s postwar miracle generated over the past three decades have more or less specified the structural conditions that allowed the developmental state to effectively implement interventionist economic policies.5 This chapter looks at the Asian developmental state from a different vantage point. In the following sections, I revisit East Asia’s postwar experience in industrial upgrading from the perspective of technological and economic risk, and specifically the ways in which the developmental state in Korea, Taiwan, and Singapore was instrumental in mitigating the risks of early industrialization and later industrial upgrading and diversification.

Industrial upgrading in Korea was led by large, diversified, and export-oriented conglomerate business groups, or chaebols.6 They benefited from state patronage. For instance, the state received, brokered, and distributed foreign technology licenses and managed the transfer and diffusion of commercial know-how.7 The chaebols gained domestic first-mover advantages in technology development because competition was “oligopolistic,” limited to a select handful of national industrial champions.8 Just as important, the state leveraged its financial capacity to allocate investment credit to favored firms.9 The banking sector was nationalized in the early 1960s, and the distribution of domestic and foreign capital allowed the state to discipline and reward firms according to national economic plans and targets.10 Essentially bankrolled by the state, Korea’s chaebols enjoyed economies of both scale and scope. Scale ensured competitive prices in global markets. The chaebols’ “brand” gave Korea’s export-oriented companies global commercial exposure that far exceeded the market recognition of any Taiwan-based or Singaporean firm. Meanwhile, economies of scope were achieved with diversification inside firms, facilitating the internal transfer of capital, technology, and management best practices.

The development of Korea’s semiconductor industry illustrates the productivity of this state-chaebol strategy for industrial upgrading. Korea’s entry into VLSI (very large scale integration) chip fabrication in 1978 was prompted when a technology transfer agreement was brokered between Silicon Valley’s VLSI Technology and the publicly funded Korean Institute of Electronic Technology. The big push into the industry came during the 1980s when the government invested US$400 million in semiconductor R&D. The parastatal Electronics and Telecommunications Research Institute (ETRI) coordinated a DRAM (memory chip) R&D consortium, which was concentrated among a few chaebols.11 In 1982, Samsung announced a then unprecedented US$130 million (100 billion won) investment in VLSI development. This sparked similarly significant investments from chaebol competitors Goldstar and Hyundai. From U.S.-based Micron and Japan’s Sharp, Samsung acquired DRAM design and processing technologies, which enabled it to develop the more advanced 64K DRAM in 1983.12 It is important to point out that supply-side push in the informatics sector was matched with demand-side pull. Domestic demand for such technology was high, as was foreign demand.13 The U.S.-Japan Semiconductor Trade Agreement of 1986 and the revaluation of the yen after the Plaza Accord a year earlier weakened Japan’s export dominance in the sector, increasing global market share for Korean producers. There was little uncertainty about demand. Indeed, only five years after Samsung made its historic announcement during the early 1980s, Korea’s DRAM technologies neared the cutting edge and were competitive in the global marketplace.14

Whereas the Korean approach to technological and industrial upgrading centered primarily on national champions “husbanded” by the state, Taiwan’s developmental state favored the growth of small and medium-sized enterprises into eventual global competitors. Politically motivated to prevent a concentration of industrial activity within a few firms, industrial policy planners in Taiwan eschewed the economies-of-scale logic pursued in Korea, opting for an industry-wide economies-of-scope approach.15 Firms were nimble and flexible, and they tended to focus on specific niches.16 Nurturing small firms, as opposed to large ones, meant that investments into specific initiatives were less concentrated. It also meant that industrial firms required less direct investment from public coffers. Unlike in Korea, the KMT party-state in Taiwan was without the financial resources to invest in large-scale diversified firms; Taiwan’s largest firms were tiny compared with Korea’s chaebols. Taiwan instead relied on fiscal-based incentives and the state’s coordinative leadership to nudge firms into higher-tech sectors and to offset the costs and risks of industrial upgrading.17 Firms adapted by devising strategies to bootstrap and take advantage of the informal financial sector.

Technological and industrial upgrading flowed out of the commercialization efforts of publicly funded labs. Technology acquisition and development, and the risks associated with these midstream processes, were concentrated within parastatal labs and then transferred to the private sector. The Industrial Technology Research Institute (ITRI) was critical in developing the state’s midstream capabilities. The ITRI adhered to a “license-in, license-out” model of technology acquisition, development, and private sector commercialization. The ITRI’s role in midstream technology development thus effectively delivered near-market technologies to otherwise risk-averse firms. During the mid-1970s, the Taiwan government brokered a licensing agreement with U.S.-based RCA to acquire what was then its already obsolete integrated circuit (IC) technology. This was assimilated into the ITRI and specifically its Electronics Research Service Organization (ERSO). The technology was reverse-engineered, and a pilot plant for manufacturing was established soon after. Because private sector firms were hesitant to invest in IC technology, initial funding for the RCA deal came largely from public sources. The state assumed the role of what John Matthews and Dong-Sung Cho refer to as “collective entrepreneur”18 when it spun out United Microelectronics Corporation (UMC) during the late 1970s.19

The ITRI and the ERSO continued to play a similar leadership role when the government decided during the early 1980s to further upgrade its electronics sector by developing Taiwan’s VLSI manufacturing capabilities. Between 1983 and 1988, more than US$72 million of government funds was allocated for microelectronics R&D, the bulk of which was channeled to the ERSO.20 UMC initially sought to lead the VLSI project, though the state resisted the overconcentration of resources and market share in a single firm.21 Consequently, the ERSO formed an R&D partnership with U.S.-based Vitelic to develop DRAM chip technology. The ERSO intended to spin out a private firm in 1986, though again, local entrepreneurs were hesitant to lead the investment. The Taiwan Semiconductor Manufacturing Corporation (TSMC) was eventually formed, with 48% of the firm’s initial investment coming from public sources, about the same proportion of government funds for the UMC deal a few years earlier.22 TSMC’s pioneering pure-play foundry model allowed it to fabricate chips economically and to specific design requirements, and the firm quickly captured significant market share in a key segment of the IT value chain. As experts such as Douglas Fuller note, IT’s value chain is “granular” and “decomposed into distinct functional parts,” which allowed TSMC to target and eventually to dominate a particular value-added niche in IT manufacturing.23

Singapore’s industrial development followed a different pathway. Its economic vulnerability after independence fomented a “discourse of survivalism.”24 Human capital was Singapore’s primary economic resource, and economic development was driven by a full-employment industrial strategy. Job creation was the main priority, not technological upgrading per se. During the 1970s, state planners looked to strengthen Singapore’s technology base, specifically in “skill-intensive, higher value-added export industries.” However, a mid-1970s recession forced the government to revert to its earlier strategy of wage restraints and the continued growth of labor-intensive sectors.25 The overwhelming importance of job creation and labor market absorption was also reflected in the growth of government-linked corporations (GLCs), de facto state-owned enterprises. GLCs fostered infrastructural development in Singapore. They were also “in competition and partnership with foreign and local private enterprise,” which meant that even into the late 1970s most of Singapore’s industrial base was tied to the government.26 It is estimated that during the early 1980s, there were almost five hundred wholly or partially government-owned firms in Singapore.27 As a result, growing local private firms was not a priority for economic planners.28 In fact, Linda Low contends that the Singaporean state “maintained a certain distance if not outright antipathy toward the local private sector.”29

The other pillar of Singapore’s postindependence industrialization strategy was the inflow of foreign direct investment (FDI), especially the location of multinational corporations (MNCs) in Singapore. During the 1980s, Singapore received more FDI than any other economy in the developing world.30 The MNC strategy was viewed by policymakers as less risky, or a way to off-load risk. Multinational companies, by virtue of their scale and credibility, ensured access to global markets. And most important, they created jobs. In 1998, MNCs and GLCs together accounted for over half of Singapore’s national product.31 MNCs alone captured the largest share of the city-state’s exports. In 1994, foreign enterprises contributed to about 75% of Singapore’s manufacturing output, 70% of value-added production, and roughly 85% of exports.32 During the 1990s, foreign sources accounted for over 27% of Singapore’s fixed capital, a share that dwarfed fixed capital stocks in Korea and Taiwan, at 0.8% and 2.6%, respectively.3 3

The Economic Development Board (EDB) of the Ministry of Trade and Industry was instrumental in attracting foreign firms to Singapore, highlighting the city-state’s “locational advantages,” including low-wage skilled labor, investment in manufacturing infrastructure, transparent policies, and strong regulatory regimes.34 The EDB strategy paid off in electronics components and IT, which became two of Singapore’s largest export sectors. U.S.-based hard disk drive firms, initially led by Silicon Valley’s Seagate, located manufacturing plants in Singapore during the early 1980s. Other MNCs, notably Texas Instruments, set up chip assembly plants as early as the 1970s. These early efforts paved the way for a joint venture between the GLC Singapore Technology Group and American firms National Semiconductor and Sierra Semiconductor. The presence of foreign electronics and IT firms stimulated the growth of local SMEs, which primarily serviced the MNC sector through supply chain linkages. The large presence of MNCs was also intended to be a “conduit for transfer of advanced technology” into Singapore. Results were mixed, however, on that front. Some upward R&D spillover was realized (i.e., the development of R&D capacities),35 though the state was unable to compel or stipulate that foreign firms be required to promote technology transfer for the purposes of local commercialization.36 Singapore was suited for skilled manufacturing.37

The postwar developmental state was effective at mitigating the risks of industrial upgrading. As Evans puts it, the developmental state nurtured “entrepreneurial perspectives among private elites by increasing incentives to engage in transformative investments and lowering the risks involved in such investments.”38 East Asian political economist Yun-Han Chu goes so far as to say that in Korea the “chaebol were able to expand rapidly and aggressively because they faced a basically risk-free environment.”39 But what exactly is meant by risk? Economists understand risk to be a quantifiable calculation of the likelihood of a certain outcome. Risk is conceived as a probability distribution. Risk calculations therefore presuppose that one has a most preferred outcome and that one attaches values to a range of potential outcomes, even if the likelihood of such outcomes is far from certain. It is because of this that economists see risk as an intrinsic characteristic of the market economy. Indeed, imperfect information and knowledge among actors (both buyers and sellers) limits certainty. Contingencies have to be accounted for. Risk is so ubiquitous in the market that firms build risk into their business models as a fixed cost. Mitigating risk, therefore, is about knowing, and if possible increasing, the probability of success.

What makes a proposition risky is the unknown, to be sure. Yet economists remind us that in all risk scenarios there are many things one does in fact know, which helps decision makers calculate the probability or odds of certain outcomes. First, risk entails what Frank Knight referred to as “known unknowns,” that is to say, the decision maker or risk taker is at least aware of what she does not know or about what he has imperfect information. Such awareness, for obvious reasons, is critical. Second, decision makers know what are the preferred outcomes of any risk scenario, knowledge that is derived from a prior valuation of the range of potential outcomes. And third, because risk is conceptualized as a probability distribution, decision makers can use existing information to estimate the likelihood of certain outcomes, to have some sense of the odds—good or bad—of preferred outcomes.40 Data and technical knowledge, for instance, are important for evaluating risk propositions because they help risk takers better understand the causal relationships they are attempting to forecast.41 Knowing the science involved in a technological breakthrough, building on existing knowledge, and devising a theoretical model for a particular application of a technology are forms of relevant information that help one calculate the probability of, for example, a successful commercial outcome. While in real-world situations risk takers can only roughly estimate the probability of certain outcomes, theirs are nonetheless educated guesses about the future informed by the past and relevant knowledge. Risk propositions are those for which decisions are not complete shots in the dark.

Thinking about the Asian postwar experience from the perspective of risk, then, one can argue that the growth of the electronics and information technologies sectors in Korea, Taiwan, and Singapore was facilitated by what I have described as the postwar developmental state’s efforts to mitigate risk. Despite key variations among the three cases, the strategic approach each took was very similar: the state was central in mitigating the risks of entry for firms and technologists in new sectors. The absence of complete certainty notwithstanding, strategic decision makers in Korea, Taiwan, and Singapore enjoyed a significant amount of knowledge about their prospects for successfully entering the electronics and IT sectors. Alice Amsden and Wan-Wen Chu offer an important observation about Taiwan—an observation that can be extended to other economies in the region—when they suggest that there was a fair degree of technological certainty in Taiwan’s efforts to industrially upgrade during the postwar period. By technological certainty, Amsden and Chu mean that Taiwan by and large imported proven technologies from abroad, which were then copied, reverse-engineered, and in time improved on for export.42 Taiwan and other late developers in the region, such as Korea and Singapore, benefited from second-mover advantages; they exploited “mature technologies.”43 Moreover, technology acquisition also centered on sectors for which there was a preexisting “pathway of learning” among engineers and entrepreneurs.44 The adoption of advanced informatics, for instance, was built on already accumulated expertise in the electronics sector and more rudimentary experience in the IT sector. High-tech industrialization in Korea, Taiwan, and Singapore was afforded by the accumulation of existing knowledge rather than the assimilation of completely new knowledge bases; in short, they copied.45

The risks of industrial upgrading in East Asia were due to what Amsden and Chu call the economic unknown. What was unknown to decision makers was whether investments in specific sectors or technologies or even firms would create timely economic returns. There was, in other words, economic risk. However, several factors mitigated this risk. First, market demand was relatively clear. The auto sector, electronics components, and semiconductors were proven markets. As Poh-Kam Wong reasons, the fact that technology upgrading in East Asia was based on a strategy of reverse engineering existing technologies and products meant that markets were effectively guaranteed.46 Dan Breznitz, in his study on innovation and the state, argues that late movers enjoyed the “double advantages of knowing the market, and accordingly being able to predict needs fairly accurately.”47 Second, industrial upgrading focused on niche areas on the value chain that were relatively close to market, such as in skilled manufacturing or the production side of high-tech industry. The producer’s distance to market was narrowed because development centered on engineering problems rather than commercializing novel technologies. Near-market technology development strategies also ensured that economic returns would generally be quick. Not only were investments recouped in the nearer term, but failed upgrading initiatives could be identified relatively quickly and adjustments made. Stakeholders could exit without having sunk overwhelming resources into a losing sector, technology, or firm.

Third, and perhaps most important, the centralized postwar developmental state coordinated the allocation of public resources to lower firms’ costs of entry into technology-based industries. In doing so, it absorbed, manipulated, and mitigated the risks of industrial upgrading. The state, for instance, subsidized fiscal and credit-based incentives to lower firms’ up-front investments and other loss-leader transaction costs. Industrial technology R&D was by and large funded from public sources. The state was the primary investor in industrial infrastructure with direct investments or indirectly through state-owned enterprises and government-linked corporations. It leveraged its credibility and resources to acquire technologies from abroad. The state also structured domestic competition in order to manufacture winners. It coordinated the development of supply chain networks among firms, in effect creating markets at home and globally. As Amsden famously remarked, the state essentially got the prices “wrong,” both to correct for anticipated market failures and to construct markets for potentially enterprising firms.48 In other words, the state picked and made winners, not by eliminating risk but by mitigating it, and by increasing the probability of success for chosen sectors, technologies, and even firms.

Thus, when Samsung boldly announced its intention to enter the competitive semiconductor industry during the early 1980s, it was a move that of course entailed risk. But stakeholders there had some sense of the likelihood of Korea’s eventual success in the sector. They knew they had a technology that worked. They were relatively certain that there was demand for their goods, especially after the yen was revalued in 1985, pricing Japanese exports out of the market. They also knew that wage restraints and the economies of scale in chaebol firms allowed them to manufacture chips at competitive prices. And they were assured by the state that it would “husband” Samsung’s entry into the sector. Though the move was risky, stakeholders in Korea had enough information—relative certainties, as it were—to reasonably forecast the probability of success and increase that probability through strategic state interventions.

The key point is not that the process of industrial upgrading into high-tech sectors was easy, nor was it a foregone conclusion by any stretch in postwar Korea, Taiwan, and Singapore. If it had been, then late-developing countries in other regions would have similarly climbed the industrial value chain. Rather, while the prospect of successful industrial upgrading was always a high-risk bet, the developmental state was both willing and able to strategically intervene in ways that increased the chances of success, such that the odds were stacked more favorably.49 The allocation of state resources lowered the entry barriers for industry by offsetting costs and absorbing the potential costs of risk taking. In short, picking winners to make meant that the state created relatively surer bets.

Betting on biotech is different. The cutting-edge frontier of the knowledge economy—whether in biotechnology, nanotechnology, advanced information and communication technologies, or other science-based businesses—and the ability to forecast innovative outcomes in these industries are characterized less by risk and more by primary uncertainty.50 On the supply side, for instance, there is tremendous technological uncertainty in first-order innovation. Upstream knowledge is too raw, unrefined, and far from market to provide clues about how scientific discoveries might translate into real-world commercial applications.51 Similarly, on the demand side, tremendous economic uncertainty surrounds potential commercial applications of new knowledge and technologies. Demand-side uncertainty rests in the difficulties by which market actors are able to forecast the economic value of basic research. Economists have long shown that the appropriability and economic valuation of new upstream discoveries are inherently prone to market failure.52 The long distance to market of new knowledge confounds prospective market demand.

Primary uncertainty is not simply a matter of higher and greater risk, however. Rather, my point is that this sort of uncertainty eludes from the start the calculability of risk. Conditions of uncertainty in first-order innovation, unlike risk propositions in technological upgrading, are those for which very little data, theory, or even a sense of eventual value exist from which to infer the probability of certain outcomes. Stakeholders in Korea, Taiwan, and Singapore have scant sense of the probability of success in commercial biotech, never mind any confidence about how to increase that probability. Biotech innovation thus entails conditions of uncertainty that are qualitatively different from those of risk scenarios, demanding qualitatively different bets by stakeholders and decision makers. Put simply, risk can be mitigated while uncertainty cannot. Picking winners in risk scenarios is an informed bet based on calculable probabilities, whereas picking winners in completely uncertain circumstances is more or less a game of chance, or as Knight put it, “pure guesswork.”53 Success in new science-based industries is about “possibility rather than probability.”5 4

Uncertainty in biotech commercialization is exacerbated by the underwhelming performance of the global biotech sector to date. The promise of health biotech first emerged during the 1970s, when scientists anticipated that the biological rather than the synthetic chemistry “heuristic” would allow researchers to better understand the biomechanisms of illness and disease. This turn marked the beginning of what innovation scholars call a new “technological paradigm.”55 New ways of diagnosing disease were explored, and it was thought that genomic techniques would enable researchers to identify and match drug targets and compounds. The completion of the human genome project in 2003 unveiled the most comprehensive map of the genetic make-up of humans. The further development of recombinant DNA (rDNA) technology meant that proteins could be synthesized in the lab for drug therapies. Knowing the biological structure of disease and illness raised the possibility that drug screening could be rational, targeted, and efficient, an approach believed to be superior to the existing model of random screening, or “molecular roulette.”56 Health biotech and the biological heuristic were envisioned as the new platform or enabling technology for rapid advancements in health and the health care industry. The health care industry was to be revolutionized, and governments, researchers, entrepreneurs, venture capitalists, and popular hype fueled the commercial biotech race.

Despite such promise, biotech has failed to live up to the lofty expectations of the 1970s and 1980s. Both R&D and industrial performance in the sector have been underwhelming. For instance, R&D productivity and new drug discovery through biotechnology techniques have been much lower than expected. Biotech-facilitated candidates have not fared as well in clinical trials as initially predicted. Biotech’s clinical efficacy also remains uncertain. While the prospects of commercial biotech prompted the restructuring of the pharmaceutical industry—notably the rapid growth of small biotech firms providing leads for established drug firms—the commercial returns from biotech have been much slower than expected and very small compared with the rate and size of investment in the sector. In fact, as of 2005, relatively few biotech products had come to market and profitability in the American-based industry had flatlined.57 As Michael Hopkins and his colleagues put it, “widely held expectations about the impact of biotechnology are over-optimistic.”58

Global trends in commercial biotech development mirror trends in industrial Asia. As relatively recent entrants in the life sciences field, Korea, Taiwan, and Singapore have understandably struggled in their efforts to grow their bio-industries. The cutting edge of bio-industry there is contracting. Innovative firms are not surviving. Risk capital has become very cautious, a position that was accentuated after the 2001 dot-com bust. And stakeholders there, increasingly frustrated with the slow rate of return, are beginning to question the viability of the long-term bet on biotech. Indeed, the underwhelming performance of the sector globally and domestically has inspired less and less confidence among stakeholders, exacerbating the uncertainties surrounding innovative commercial biotech. Prospects are bleak even after two decades of investment. Stakeholders have little sense of which R&D and business models work best. They have little to go on with respect to due diligence or even the means for benchmarking and measuring progress. There are few models to emulate, given the lack of blockbusters in the industry. In other words, the poor performance of the sector has provided so little clarity about the sector that decision makers are not able to get any feel for the probability of success, and they have even less sense of how to increase that probability.

The challenges of managing uncertainty in commercial biotech innovation are derived, I contend, from the nature of the science itself. As with other science-based industries, biotech’s uncertainties are technological, economic, and temporal. The basis of technological uncertainty in biotech innovation stems from the unintegrated, multidisciplinary nature of the field. Biotech innovation requires expertise from different scientific fields (biology, chemistry, physics), engineering (applied problem solving), and industry, including intellectual property management, venture firm strategy, legal policy, and regulatory sciences. Its multidisciplinarity also means that specialized skills and expertise are institutionally disparate, divided among many stakeholders.59 And yet innovative biotech R&D demands interaction and collaboration among these disparate groups of expert stakeholders. The fact that biotech is considered an enabling or platform technology for both upstream researchers and downstream entrepreneurs means that the biotech R&D process is neither linear nor unidirectional. Rather, bio-industrial R&D continually moves up, down, and across various R&D agendas, making the likelihood of a technological breakthrough even more uncertain.60

Biotechnology’s demand-side economic uncertainties are equally pronounced. As with all science-based industries, commercializing biotech, a technology that is far from mature, encounters the inherent problems of valuing knowledge, identifying market demand, and forecasting market potential for such knowledge.61 Gary Pisano notes: “Consider the problem of valuation. Existing financial modeling methodologies are of little help: they rely on an analysis of historical earnings and earnings potential. Most biotech companies have no earnings (let alone an earnings history), and with nearest products several years away and facing enormous technical and commercial uncertainty, it is hard to construct any reasonable valuation model.”62 What Pisano means is that the commercial biotech market remains undeveloped. The array of industrial applications of biotech is vast. The potential value chain in commercial biotech is exceptionally long and, as yet, relatively undefined. Identifying commercially viable markets thus remains an open-ended endeavor.

Furthermore, because of the complexity and unintegrated nature of biotech innovation, the structure of the commercial biotech industry is one in which firms and labs are highly specialized and narrow in their R&D operations. On their own, these firms create little commercial value. Successful commercialization rests on the integration of these different bases of expertise with bits and pieces of knowledge. As an enabling technology in health care, for instance, biotech is useful insofar as there are applications that it can help deliver. As a platform technology, biotech can realize its integrative potential only if there are firms and labs working on biotech applications that can use this platform. Commercializing innovation thus requires not only the cultivation of disparate specialized knowledge but also the eventual integration of decentralized expertise in ways that “fit.” The reality is that much of the innovative work in commercializing biotech takes place long before the technology is even considered market-ready. Biotech’s distance to market exacerbates economic uncertainty.

The story of Celera Genomics illustrates the problem. Based in the United States, Celera was the first private firm to map the human genome. The firm developed in tandem, and indeed in competition, with the U.S. government’s efforts in leading the worldwide human genome project. When the human genome map was completed, Celera was sure customers would demand access to this new genetic knowledge. Celera quickly learned, however, that its original business plan to sell its proprietary genomic information was flawed. Though the firm was on the cutting edge of life sciences R&D, too few customers were buying Celera’s product. The information business of life sciences, which made good commercial sense in the abstract, failed. Despite forecasts to the contrary, Celera’s product commanded little market value. In 2000, Celera shares sold for US$200. By the end of 2002, share prices had fallen to just US$20, one-tenth the value. Even what seemed to be the most commercially certain of enterprises, a good bet backed by good science, succumbed to commercial biotech’s economic uncertainty.63

Biotech’s long distance to market is a source of temporal uncertainty. Commercializing innovative biotech is a very long-term enterprise, and success is a long shot. And for recent entrants to the sector such as Korea, Taiwan, and Singapore, the temporal horizon looms even longer. The implications of distance to market exacerbate biotech’s uncertainties in several ways. For one, the regulatory demands of the sector mean that premarket regulatory regimes governing ethics, standards, R&D, and product development not only lengthen biotech’s distance to market but also add to the potential obstacles to the commercialization of new biotech products and services. The problem of regulatory obstacles is compounded, I argue in Chapter 5, if regulatory policies are not coherent and consistently enforced. Second, and from a more political point of view, the sector’s long distance to market continually tests stakeholders’ commitment to growing innovative bio-industries. As alluded to above, the sector’s poor performance in Korea, Taiwan, and Singapore has dampened people’s confidence about its prospects and exacerbated its uncertainties. Coping with such uncertainty indefinitely is inconceivable and, more precisely, as I show in Chapter 4, politically unfeasible. Thus, biotech innovation requires political strategies to sustain a nation’s appetite for uncertainty over the very long term. It requires that strategic stakeholders manage expectations, rationalize sunk investments and year-over-year losses, and, most pressing, normalize failure in bio-industrial endeavors. All the while, stakeholders are confronted with the question, “How long is long enough?” Betting on biotech means that decision makers have to gauge when the time is right to cut their losses, admit failure, and move on, a choice that has rarely been explored in the experience of the postwar developmental state.

Betting on the biotech sector more broadly is one thing, but making specific bets in what is a technologically, economically, and temporally uncertain industry is another. As I have argued, conceiving of biotech innovation in terms of primary uncertainty implies that decision makers have little to go on in how they make strategic choices. Which technologies in the life sciences ought to be targeted and how? What R&D capacities need to be facilitated? What regulatory policies should be implemented and by what criteria? How should resources be allocated and according to what rationale? Primary uncertainty obscures clarity with respect to these core questions of policy, resource allocation, and industrial strategy. In other words, managing uncertainty is a tall order because primary uncertainty precludes such manageability.

Economists tell us that one plausible way out of this conundrum is to leverage the economies of scope and scale. Economic actors should allocate resources broadly and indiscriminately, diversifying the scope of activity across an entire industrial field and diffusing the costs of such activity among many actors. From this process, universes of “like cases” can be assembled, out of which relevant data and knowledge are gathered. Specialization, diversification, diffusion, and then consolidation in uncertain ventures provide a history that decision makers can use to infer ways to manage uncertainty. Scale is essential, however. The ability to bet indiscriminately is also critical. But the reality is that it is also inconceivable in many economies, the exceptions being perhaps the United States and giant latecomers such as China. Rather, the challenge for Korea, Taiwan, and Singapore, and virtually all other aspirants in the biotech sector, is to bet discriminately and thus strategically. But how?

Despite an enormous—and still growing—multidisciplinary literature on technology innovation, theory provides few answers about how decision makers can go about dealing with the uncertainties inherent in first-order innovation. The national innovation systems framework provides considerable insight into the process of connecting disparate actors and experts to bridge the gaps in translating science into applied technologies, in long-term financing of new inventive products and services, and in people-to-people social networks and clusters.64 Innovation, we know, is an intensely interactive enterprise, and one that is determined by people and the institutional complexes that shape people’s behaviors, their incentives, and their interactions.65 Interactions are also local and global.66 However, innovation systems theory tends to be empirically rich and analytically descriptive rather than predictive-theoretical. To correct this, the varieties-of-capitalism framework developed by Peter Hall and David Soskice attempts to provide some causal claims between certain institutional designs and innovation outcomes.67 And while their assertions intuitively make sense—for instance, the basis of financial markets determines incremental versus radical innovations—the empirical evidence does not hold up across all sectors and all national contexts.68 In other words, theory does not inform confident claims about how to bet on innovation and specifically on biotech innovation.

Indeed, the inherent uncertainties of first-order innovation obscure theoretical and prescriptive precision. Instinctively, for example, intellectual property (IP) protection might seem to be an institutional foundation for any innovation system, though empirical studies have shown that the effect of IP regimes varies by sector. The presumed positive effects of the 1980 Bayh-Dole Act, a landmark policy initiative that permits closer commercial linkages to be formed between the academy and industry, have also been brought into question.69 In the commercial biotech sector specifically, stakeholders are increasingly questioning the relevance and applicability of the venture financing model developed in the information technology sector, just as strategic planners inside global pharmaceutical firms increasingly wonder about firm best practices regarding the development or acquisition of biotech leads. Susanne Giesecke captures this conundrum when she observes that

[biotechnology’s] specific features include the length and especially the uncertainty of races to find new compounds. This search involves extensive time horizons that are due to strict safety regulations and thereby cause high capital intensity. At the same time, biotechnology is a high-risk technology, economically as well as technically. Thus, an enabling institutional arrangement has to provide for the maximum containment of those economic, technical and safety risks. These specific features—time intensity, uncertainty, capital intensity and risks—are inherent to the path of biotechnical development.70

The point is that biotech’s uncertainties—and the inherent uncertainty of science-based industries more generally—beget further organizational, institutional, and strategic uncertainties. Theories thus provide important analytical guideposts, but in the end, adapting to the realities of uncertainty and the innovation economy involves continual experimentation and learning among decision makers and stakeholders.71 As Bengt-Ake Lundvall puts it, it is the “ongoing processes of learning, searching and exploring which result in new products, new techniques, new forms of organization and new markets.”72 Dani Rodrik similarly asserts that technology innovation systems and innovation best practices cannot simply be imported from abroad and transposed and replicated at home. Rather, he argues, the strategic and institutional bases of an innovation system are ultimately shaped “locally, relying on hands-on experience, local knowledge and experimentation.”73 To experiment and to gain experience means that strategic decision makers have to do something and they have to rationalize their decisions. Doing nothing is not an option; they have to make choices .

The rest of this book examines what “doing something” has actually entailed in Korea, Taiwan, and Singapore since they ramped up efforts to make it in the commercial biotech sector. The book essentially tells two stories, both of which I briefly foreground here. The first is a story of continuities. Decision makers in Korea, Taiwan, and Singapore have continued to make policies and allocate resources in ways not dissimilar to the earlier periods of industrial upgrading in the developmental state. In some respects, stakeholders have attempted to manage biotech’s uncertainties by adhering to past practices in mitigating risk. The second story, however, is one of significant change and discontinuity. More specifically, the evidence from Korea, Taiwan, and Singapore reveals a fundamentally changing role of the developmentally oriented state, reflecting the ways in which first-order innovation in commercial biotech is a qualitatively different political economic endeavor than in the past and demonstrating that betting in bio-industrial development requires different strategic rationalizations.

Decision makers in Korea, Taiwan, and Singapore are very aware of the technological, economic, and temporal uncertainties of commercial biotech innovation. Invariably, stakeholder informants in all three places (and elsewhere as well) express to me their frustration over these sorts of uncertainties and what to do about them. The fact that Korea, Taiwan, and Singapore have invested billions of dollars in the sector—an investment that in the past would have inspired great confidence for success—and have seen, as yet, relatively little economic return only exacerbates such feelings of frustration. Still, they recognize that they need to do something to facilitate commercial biotech innovation. As in the postwar period of industrial upgrading, biotech stakeholders continue to believe that the state can help create an environment that is conducive to the distant and unpredictable possibilities of breakthrough innovations.

The state in all three places has similarly increased its financial commitment to upstream life sciences R&D, not only with research dollars but also with major investments in infrastructure. The government in Korea, Taiwan, and Singapore has, for instance, paid significant attention to regulatory reform, from intellectual property protection to clinical regulation. It has directed resources to encourage industry to take on a larger share in the biotech R&D pie. The state has also explicitly encouraged technology transfer and the deepening of linkages between the academy and industry to narrow the research gap in technological innovation. Specifically, it has relaxed regulations that had earlier impeded the flow of knowledge and people between the public and private sectors, and the governments in all three places have taken an active role in creating new technology transfer mechanisms inside public R&D institutions.

With respect to growing firms, the state in Korea, Taiwan, and Singapore has provided myriad incentives, including fiscal and credit subsidies, to lure private sector investment into science-based industries. In fact, as I argue in Chapter 4, the state has invested heavily in creating biotech “stars” and national champions in industry and R&D. In all three places, public resources have also been allocated directly and indirectly to the venture capital sector, and state regulations have been relaxed specifically to encourage bio-venture investment. Meanwhile, financial policies have similarly been reformed in order to channel more private sector investment into high-risk technology sectors: secondary stock markets have been created, mergers and acquisitions regulations have been revised, and the initial public offering (IPO) process has been reformed, thus providing early-stage venture investors a larger menu of exit options.

Variations in past practices among Korea, Taiwan, and Singapore have also persisted into the era of biotechnology. The evidence presented in Chapters 2 and 3 shows how strategic decision makers have not veered far from existing strategic repertoires of policy interventions and models of industrial organization. Stakeholders in Korea, for example, have continued to pin the future of commercial biotech innovation on the chaebol sector and its scale advantages. Managing uncertainty in Taiwan has continued to rely on a hit-and-miss strategy for identifying potential leads inside midstream technology development centers such as the ITRI. And decision makers in Singapore aim to continue attracting multinational firms, rather than creating homegrown innovators, by spending on and strengthening the city-state’s existing locational advantages. In each case, decision makers have approached the challenges of managing uncertainty by reverting to existing scripts that had helped them mitigate risk so well in the past. The legacies of the postwar developmental state have persisted into the present, and the specific imprint of each country’s unique developmental experiences has also endured.

But this should not be surprising. An influential body of empirical research shows that under conditions of primary uncertainty, decision makers rely on heuristic biases—beliefs, values, cognitive shortcuts, repertoires—to help them navigate what they otherwise know little about. According to economics Nobel laureate Daniel Kahneman, decision makers cope with uncertainty by making decisions as though they can in fact manage it. Heuristic biases, scripts, or strategic repertoires allow decision makers to cognitively transform otherwise unmanageable realities of primary uncertainty into what they believe to be more manageable risk propositions.74 Biotech stakeholders in Korea, Taiwan, and Singapore have reverted to what they know best and what they know worked in the past to help them make strategic choices in the face of what in reality is a tremendously uncertain technical and commercial endeavor. Therefore, one way to think of this first story is that strategic decision making in the era of science-based industrialization in Korea, Taiwan, and Singapore is merely “old wine in new bottles.” There is obviously some truth to that characterization, as we would be naive to think that decision makers today are working from a strategic tabula rasa.

But the evidence presented in this book also suggests that amid such continuity, there is something more fundamentally transformative going on in Korea, Taiwan, and Singapore. Decision makers, particularly those inside the state, have had to make significant adjustments and to articulate new strategic rationalizations about the role of the state in the face of such uncertainty. What has occurred over the past twenty years or so in Korea, Taiwan, and Singapore in terms of science-based industrial development is therefore not merely old wine in new bottles. The enduring scripts, strategic repertoires, and heuristic biases described above notwithstanding, there has also been a fundamental transformation in the very idea of the developmental state. The strategic logic of the postwar experience in state-led economic development has been altered.

First, in all three cases, the state has retreated from its past leadership role in directing industrial upgrading, and economic development more generally, from the top down. It continues to allocate resources to facilitate new industrial endeavors. But frankly, this sort of government intervention is not unique to places such as Korea, Taiwan, and Singapore. All states—even the leanest of the liberal market economies such as the United States—pour massive amounts of public resources into new industry development. Rather, what always stood out among the Asian cases was that the developmental state proactively picked winners and, in doing so, bore the heavy costs of risk inherent in industrial upgrading. But this, I suggest, is no longer the case. The state has strategically refrained from picking winners in commercial biotech. In fact, it has proactively shifted uncertainty to other stakeholders.

Second, the long-term realities and related uncertainties of biotech innovation have compelled decision makers in Korea, Taiwan, and Singapore to significantly scale back past practices in actively coordinating industrial activity. The hallmark of the developmental state was its capacity and willingness to coordinate actors in productive ways, using various state incentives, its close interactions with industry, and strategic industrial policies. The developmental state also actively forged linkages among public and private actors. The founding of Taiwan’s TSMC is a perfect example of the developmental state’s coordinative capacity for growing firms and its willingness to use it. The approach to biotech innovation, however, rests more on the long-term potentiality that such linkages will form over time than on strategic decision makers coordinating such linkages from the top down. Put another way, the state has helped put the pieces of the biotech sector in place but has left putting those pieces together to others.

Third, the retreat of the state and its diminishing capacity to coordinate reflect a more general pattern in the diminished coherence and fragmentation of the state apparatus. The postwar developmental state was vertically organized, characterized by what Evans calls its “corporate coherence.” However, as the following chapters show, this is no longer the case when it comes to life sciences innovation. Biotech stakeholders in Korea, Taiwan, and Singapore in fact comprise a wide range of actors with specialized knowledge and expertise, interests and priorities, both inside and outside the state apparatus. Corporate coherence has given way to greater fragmentation, institutional disparateness, contested policy agendas, and varied bases of expertise, which have, I argue, reshaped the processes of decision making.

The main objective of this chapter was to lay out a conceptual distinction between risk and uncertainty. Some may think the distinction, as I have argued it here, is a bit overstated. It is not as though we know nothing about biotechnology; and in reality, biotech’s prospects are not what Knight theoretically saw as primary uncertainty. This is true. But as the basis for a framework, the distinction provides tremendous analytical utility. The reality for decision makers in Korea, Taiwan, and Singapore is that they themselves understand their current circumstances to be vastly different from what they confronted during earlier modes of industrial upgrading in the postwar period. They have experienced the prospects of growing bio-industry as profoundly uncertain. Whereas in the past, industrial upgrading centered on mature technologies with defined applications and market demand, biotech innovation entails an as yet undeveloped technology with very unclear market signals. The framework I have presented here thus helps illuminate how stakeholders and decision makers in Korea, Taiwan, and Singapore have had to make choices, make strategic bets, and articulate new rationalizations of those bets in their efforts to succeed in commercial biotech innovation. Indeed, without such a conceptual distinction between risk and primary uncertainty, one might see the empirical evidence in Asia as solely path-dependent continuities of past practices in mitigating risk—old wine in new bottles. And while this is one of the stories of biotech innovation in Korea, Taiwan, and Singapore, the distinction between risk and uncertainty permits a deeper story about a more fundamental transformation beyond the developmental state in the current era of science-based industrialization.

1 . In the context of capital markets, Jeffrey Winters distinguishes between risk and uncertainty in the following way: “When the stakes are high, information is crucial. Investors operate well with risk, for which probabilities of different outcomes can be calculated. They do not operate well with uncertainty, which means the absence of quality information on which to base investment decisions.” Winters, “The Determinants of Financial Crisis in Asia,” in The Politics of the Asian Economic Crisis, ed. T. J. Pempel (Ithaca: Cornell University Press, 1999), 87.

2 . Linda Weiss, “Guiding Globalisation in East Asia: New Roles for Old Developmental States,” in States in the Global Economy: Bringing Domestic Institutions Back In, ed. Weiss (Cambridge: Cambridge University Press, 2003).

3 . Robert Wade, Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization (Princeton, N.J.: Princeton University Press, 1990); Gregory Noble, Collective Action in East Asia: How Ruling Parties Shape Industrial Policy (Ithaca: Cornell University Press, 1998).

4 . Peter Evans, Embedded Autonomy: States and Industrial Transformation (Princeton, N.J.: Princeton University Press, 1995).

5 . Dan Breznitz observes, rightly in my view, that theorists of the developmental state and the neodevelopmental state essentially “advance an argument about the need for specific state structure that enables emerging economies to utilize a particular strategy of development.” Theories of the developmental state thus privilege the structure of the state and state-society relations as a priori conditions to effective strategic decision making and implementation. See Breznitz, Innovation and the State: Political Choice and Strategies for Growth in Israel, Taiwan and Ireland (New Haven, Conn.: Yale University Press, 2007), 14.

6 . In 1974, earnings from the top ten chaebols equaled 15% of Korea’s national product; ten years later, the top ten chaebols accounted for over 67% of GNP.

7 . Linsu Kim, Imitation to Innovation: The Dynamics of Korea’s Technological Learning (Boston: Harvard Business School Press, 1997); Kim, “Crisis, National Innovation and Reform in South Korea,” MIT Japan Program, Working Paper 01.01, 2001; Larry Westphal, Linsu Kim, and Carl Dahlman, “Reflections on the Republic of Korea’s Acquisition of Technological Capability,” in International Technology Transfer: Concepts, Measures and Comparisons, ed. Nathan Rosenberg and Claudio Frischtak (New York: Praeger, 1985).

8 . Alice Amsden, Asia’s Next Giant: South Korea and Late Industrialization (New York: Oxford University Press, 1989), 116–130.

9 . Jung-En Woo, Race to the Swift: State and Finance in Korean Industrialization (New York: Columbia University Press, 1991); Karl Fields, Enterprise and the State in Korea and Taiwan (Ithaca: Cornell University Press, 1995); Evans, Embedded Autonomy.

10 . According to Eun Mee Kim, nearly 50% of all commercial loans invested during the 1960s and 1970s were allocated to just six targeted industrial sectors, most of which were guaranteed by the government in terms of repayment. The six sectors were steel, nonferrous metals, machinery, shipbuilding, electrical appliances and electronics, and chemicals. See Eun Mee Kim, Big Business, Strong State: Collusion and Conflict in South Korean Development, 1960–1990 (Albany: SUNY Press, 1997), 110.

11 . Evans, Embedded Autonomy, 141; Mariko Sakakibara and Dong-Sung Cho, “Cooperative R&D in Japan and Korea: A Comparison of Industrial Policy,” Research Policy 31 (2002), 685; William Keller and Louis Pauly, “Crisis and Adaptation in Taiwan and South Korea: The Political Economy of Semiconductors,” in Crisis and Innovation in Asian Technology, ed. William Keller and Richard Samuels (Cambridge: Cambridge University Press, 2003).

12 . John Matthews and Dong-Sung Cho, Tiger Technology: The Creation of a Semiconductor Industry in East Asia (Cambridge: Cambridge University Press, 2000), 121–123.

13 . Evans, Embedded Autonomy, 142.

14 . Linsu Kim, “National Systems of Industrial Innovation: Dynamics of Capability Building in Korea,” in National Innovation Systems: A Comparative Analysis, ed. Richard Nelson (New York: Oxford University Press, 1993), 377.

15 . Yongping Wu, A Political Explanation of Economic Growth: State Survival, Bureaucratic Politics, and Private Enterprises in the Making of Taiwan’s Economy, 1950–1985 (Cambridge, Mass.: Harvard University Press, 2005); Breznitz, Innovation and the State; Fields, Enterprise and the State.

16 . Suzanne Berger and Richard Lester, eds., Global Taiwan: Building Competitive Strengths in a New International Economy (Armonk, N.Y.: M. E. Sharpe, 2005); Vincent Wang, “Developing the Information Industry in Taiwan: Entrepreneurial State, Guerilla Capitalists and Accommodative Technologists,” Pacific Affairs 68 (1995).

17 . Tun-Jen Cheng, “Political Regimes and Development Strategies: South Korea and Taiwan,” in Manufacturing Miracles: Paths of Industrialization in Latin America and East Asia, ed. Gary Gereffi and Donald Wyman (Princeton, N.J.: Princeton University Press, 1990); Fields, Enterprise and the State.

18 . Matthews and Cho, Tiger Technology, 167; see also Constance Squires Meaney, “State Policy and the Development of Taiwan’s Semiconductor Industry,” in The Role of the State in Taiwan’s Development, ed. Joel Aberbach, David Dollar, and Kenneth Sokoloff (Armonk, N.Y.: M. E. Sharpe, 1994).

19 . Despite firms’ initial reluctance to invest in Taiwan’s IC manufacturing sector, the state resisted forming a wholly state-owned firm, opting instead to use its “direct influence” to organize a broad investment consortium of private firms, which in the end accounted for 51% equity in UMC. See Breznitz, Innovation and the State, 107.

20 . Breznitz, Innovation and the State, 109.

21 . Doug Fuller, Charles Sodini, and Akintunde Akinwande, “Leading, Following or Cooked Goose? Innovation Successes and Failures in Taiwan’s Electronics Industry,” in Global Taiwan: Building Competitive Strengths in a New International Economy, ed. Suzanne Berger and Richard Lester (Armonk, N.Y.: M. E. Sharpe, 2005).

22 . Philips, the multinational Dutch electronics firm, made up 28% of TSMC’s initial investment, which, combined with the government’s 48%, meant that less than one-quarter of the firm’s investment came from local private sources.

23 . Fuller, Sodini, and Akinwande, “Leading, Following or Cooked Goose?,” 77; Annalee Saxenian and Jinn-Yuh Hsu, “The Silicon Valley–Hsinchu Connection: Technical Communities and Industrial Upgrading,” Industrial and Corporate Change 10 (2001).

24 . Henry Wai-Chung Yeung, “State Intervention and Neoliberalism in the Globalizing World Economy: Lessons from Singapore’s Regionalization Program,” Pacific Review 13 (2001), 146; see also David Chang, “Nation-Building in Singapore,” Asian Survey 8 (1968).

25 . Mike Hobday, Innovation in East Asia: The Challenge to Japan (Brookfield, Vt.: Edward Elgar, 1995), 140; Linda Lim, “Singapore’s Success: The Myth of the Free Market Economy,” Asian Survey 23 (1983), 757.

26 . Lim, “Singapore’s Success,” 755.

27 . W. G. Huff, “The Developmental State, Government and Singapore’s Economic Development since 1960,” World Development 23 (1995), 1428.

28 . According to Lee Kuan Yew, “the only reason the government moved in was that no entrepreneur had the guts and the gumption and the capital to go in on his own. . . . And we are prepared to go into the more high-risk areas where Singaporean entrepreneurs are unable to carry that risk, either for lack of daring or for lack of capital.” Cited in Florian von Alten, The Role of Government in the Singapore Economy (Frankfurt: Peter Lang, 1995), 200.

29 . Linda Low, “The Singapore Developmental State in the New Economy and Polity,” Pacific Review 14 (2001), 146. See also Alexius Pereira, “Whither the Developmental State? Explaining Singapore’s Continued Developmentalism,” Third World Quarterly 29 (2008).

30 . Huff, “Developmental State,” 1425.

31 . Gavin Peebles and Peter Wilson, Economic Growth and Development in Singapore (Cheltenham, U.K.: Edward Elgar, 2002), 14.

32 . Kai-Sun Kwong, “Singapore: Dominance of Multinational Corporations,” in Industrial Development in Singapore, Taiwan and South Korea, ed. Kwong et al. (Hackensack, N.J.: World Scientific Press, 2001), 10.

33 . Peebles and Wilson, Economic Growth, 171.

34 . David McKendrick, Richard Doner, and Stephan Haggard, From Silicon Valley to Singapore: Location and Competitive Advantage in the Hard Disk Drive Industry (Stanford, Calif.: Stanford University Press, 2000), 37.

35 . Alice Amsden and F. Ted Tschang, “A New Approach to Assessing the Technological Complexity of Different Categories of R&D (with examples from Singapore),” Research Policy 32 (2003).

36 . Matthews and Cho, Tiger Technology, 211–220.

37 . In their study of the hard disk drive sector, David McKendrick, Richard Doner, and Stephan Haggard make a distinction between “operational” and “technological” clusters. Singapore was selected to be an operational cluster for the HDD sector, while the industry’s technological cluster by and large remained in the United States. See McKendrick, Doner, and Haggard, From Silicon Valley to Singapore. This observation is also emphasized in Amsden and Tschang, “New Approach.”

38 . Peter Evans, “Predatory, Developmental and Other Apparatuses: A Comparative Political Economy Perspective on the Third World State,” Sociological Forum 4 (1989), 562–563.

39 . Yun-Han Chu, “Surviving the East Asian Financial Storm,” in The Politics of the Asian Economic Crisis, ed. T. J. Pempel (Ithaca: Cornell University Press, 1999), 201.

40 . Frank Knight, Risk, Uncertainty and Profit (New York: Houghton Mifflin, 1921).

41 . As Lewis Branscomb and Philip Auerswald explain, “it is not possible . . . to talk meaningfully about a given project having a ‘ten percent probability of success’ in the absence of some accumulated prior experience (such as a sample of similar projects of which nine in ten were failures).” See Branscomb and Auerswald, Taking Technical Risks: How Innovators, Executives and Investors Manage High-Tech Risks (Cambridge, Mass.: MIT Press, 2001), 44.

42 . Alice Amsden and Wan-Wen Chu, Beyond Late Development: Taiwan’s Upgrading Policies (Cambridge, Mass.: MIT Press, 2003).

43 . Marie Anchordoguy, Reprogramming Japan: The High-Tech Crisis under Communitarian Capitalism (Ithaca: Cornell University Press, 2005).

44 . Amsden and Chu, Beyond Late Development.

45 . Vincent Wang explains that Taiwan’s entry into the information and communication technologies sector was “an outgrowth of previous successive development strategies,” not a “detached “new” sector . . . without any links to the economy at large. Wang, “Developing the Information Industry,” 552.

46 . Poh-Kam Wong, “Singapore’s Technology Strategy,” in The Emerging Technological Trajectory in the Pacific Rim, ed. Denis Fred Simon (Armonk, N.Y.: M. E. Sharpe, 1995).

47 . Breznitz, Innovation and the State, 13.

48 . Alice Amsden, Asia’s Next Giant: South Korea and Late Industrialization (New York: Oxford University Press, 1989).

49 . Kim, “National Systems of Industrial Innovation,” 376–377.

50 . Gary Pisano, Science Business: The Promise, the Reality and the Future of Biotech (Boston: Harvard Business School Press, 2006), 7–8. Pisano elaborates: “In science-based business, R&D confronts fundamental questions about technical feasibility. Is it possible to express a protein in a bacteria cell? Is it possible to culture mammalian cells in vitro? What genes are involved in depression? Which biochemical pathways are involved in inflammation? . . . These are the types of questions with which science-based business in biotechnology have had to grapple. Not only are such questions difficult to answer, but the attempt to answer them leads, in all likelihood, to more questions—or to unexpected results.”

51 . Johan Brink, Maureen McKelvey, and Keith Smith, “Conceptualizing and Measuring Modern Biotechnology,” in The Economic Dynamics of Modern Biotechnology, ed. Maureen McKelvey, Annika Rickne, and Jens Laage-Hellman (Northampton, Mass.: Edward Elgar, 2004).

52 . See John Kay, “Technology and Wealth Creation: Where We Are and Where We’re Going,” Other, December 19, 2000; J. S. Metcalfe, “Science Policy and Technology Policy in a Competitive Economy,” International Journal of Social Economics 24 (1997), 728; Kevin Murphy and Robert Topel, eds., Measuring the Gains from Medical Research: An Economic Approach (Chicago: University of Chicago Press, 2003); Partha Dasgupta and Paul David, “Toward a New Economics of Science,” Research Policy 23 (1994), 490; Kenneth Arrow, “Economic Welfare and the Allocation of Resources for Inventions,” in The Rate and Direction of Inventive Activity, ed. Richard Nelson (Princeton, N.J.: Princeton University Press, 1962); Evan Berman, “The Economic Impact of Industry-Funded University R&D,” Research Policy 19 (1990).

53 . Knight, Risk, 250.

54 . Seongjae Yu, “Korea’s High Technology Thrust,” in The Emerging Technological Trajectory of the Pacific Rim, ed. Denis Simon (Armonk, N.Y.: M. E. Sharpe, 1995), 85.

55 . G. Dosi, “Technological Paradigms and Technological Trajectories: A Suggested Interpretation of the Determinants and Directions of Technological Change,” Research Policy 11 (1982).

56 . Michael Hopkins, Paul Martin, Paul Nightingale, Alison Kraft, and Surya Mahdi, “The Myth of the Biotech Revolution: An Assessment of Technological, Clinical and Organisational Change,” Research Policy 36 (2007), 568.

57 . Pisano, Science Business, 112–115.

58 . Hopkins et al., “Myth of the Biotech Revolution,” 584.

59 . Branscomb and Auerswald, Taking Technical Risks.

60 . See Maureen McKelvey, Annika Rickne, and Jens Laage-Hellman, eds., The Economic Dynamics of Modern Biotechnology (Northampton, Mass.: Edward Elgar, 2004); Robert Kaiser and Heiko Prange, “The Reconfiguration of National Innovation Systems: The Example of German Biotechnology,” Research Policy 33 (2004); Susanne Giesecke, “The Contrasting Roles of Government in the Development of Biotechnology Industry in the US and Germany,” Research Policy 29 (2000).

61 . The fact that, according to Luigi Orsenigo, market predictions during the 1990s for the biotech sector ranged from US$10 billion to US$60 billion reflects the problem of forecasting biotechnology’s market potential. Orsenigo, “The Dynamics of Competition in a Science-Based Technology: The Case of Biotechnology,” in Technology and the Wealth of Nations, ed. Dominique Foray and Christopher Freeman (New York: Pinter, 1993), 43.

62 . Pisano, Science Business, 143.

63 . See Ingrid Wickelgren, The Gene Masters: How a New Breed of Scientific Entrepreneurs Races for the Biggest Prize in Biology (New York: Times Books, 2002).

64 . See Richard Nelson, ed., National Innovation Systems (New York: Oxford University Press, 1993); Donald Stokes, Pasteur’s Quadrant: Basic Science and Technological Innovation (Washington, D.C.: Brookings Institution, 1997); Branscomb and Auerswald, Taking Technical Risks.

65 . For a more thorough review of institutional effects on technological innovation, see J. Rogers Hollingsworth, “Doing Institutional Analysis: Implications for the Study of Innovations,” Review of International Political Economy 7 (2000).

66 . See Steven Casper, Creating Silicon Valley in Europe: Public Policy towards New Technology Industries (New York: Oxford University Press, 2007); Daniele Archibugi and Simona Iammarino, “The Policy Implications of the Globalization of Innovation,” Research Policy 28 (1999); Michael Porter, “Clusters and the New Economics of Competition,” Harvard Business Review, November–December 1998.

67 . Peter Hall and David Soskice, eds., Varieties of Capitalism: The Institutional Foundations of Comparative Advantage (New York: Oxford University Press, 2001).

68 . See, for instance, Mark Zachary Taylor, “Empirical Evidence against Varieties of Capitalism’s Theory of Technological Innovation,” International Organization 58 (2004); Dirk Akkermans, Carolina Castaldi, and Bart Los, “Do ‘Liberal Market Economies’ Really Innovate More Radically than ‘Coordinated Market Economies’? Hall and Soskice Reconsidered,” Research Policy 38 (2009).

69 . David Mowery, Richard Nelson, Bhaven Sampat, and Arvids Ziedonis, eds., Ivory Tower and Industrial Innovation: University-Industry Technology Transfer Before and After the Bayh-Dole Act in the United States (Stanford, Calif.: Stanford University Press, 2004).

70 . Giesecke, “Contrasting Roles of Government,” 209–210.

71 . In her study of the origins of life sciences technology transfer at Stanford University, Jeannette Colyvas contends that the early institutionalization of Stanford’s program resulted from experimentation with four contending organizational models, reflecting different interests and understandings of research collaboration. See Jeannette Colyvas, “From Divergent Meanings to Common Practices: The Early Institutionalization of Technology Transfer in the Life Sciences at Stanford University,” Research Policy 36 (2007), 474.