Mission-oriented innovation policies and bureaucracies in East Asia

Abstract

This article tackles two “how” questions of mission-oriented innovation policies in East Asia: How do these economies legitimize their mission-oriented innovation policies in the public discourse? and How do they create dynamic capabilities for implementing these policies? The article shows that although on the level of policy choices and design, the global–Western discourse has also entered the East Asian mission-oriented innovation policy rhetoric, this has not stopped the policymakers in Japan, South Korea, and Taiwan from relying on old “developmentalist” logics for legitimizing these policies. This also influences actual policy implementation styles: as opposed to Western models of peripheral Schumpeterian agencies offering dynamic capabilities, the East Asian innovation bureaucracies seem to rely on more visible and “politicized” organizations for such capabilities. We also discuss the strengths and weaknesses of this model.

1. Introduction

Parallel to the growing Western debates on mission-oriented innovation policies (Foray et al., 2012; Mazzucato, 2013,, 2016), some of the leading East Asian economies—Japan, the Republic of Korea (South Korea), and the Republic of China (Taiwan)—have also introduced a similar mission orientation to their respective policy discourses and strategies. These missions tend to be twofold: first, to speed up innovation in the areas of national specializations and global bets on future growth drivers [biotech, information and communication technologies (ICT), and renewable energy]; second, to use the concept of societal challenges as mission-creating and coordinating devices for directing national innovation efforts. Although the policy rhetoric of these economies has many overlaps with the Western debates both on what kinds of missions to set and how to design supportive policies (e.g., focus on emulating the Defense Advanced Research Projects Agency [DARPA] model), these countries also have their unique policy and governance legacies of “developmentalism” (Fields, 2012; Thurbon, 2014) and “developmental state” (Johnson, 1982). This article tackles two “how” questions: How do the East Asian economies legitimize their mission-oriented innovation policy efforts in the public discourse? and How do they create dynamic capabilities for implementing these policies, i.e., what type of governance systems and organizations do they use?

The article is structured as follows. The next section provides the analytical framework that assumes that the design and implementation of effective mission-oriented innovation policies depends on how different countries manage to achieve complementarity between effective ways of legitimizing policies, which is often an issue out of the hands of innovation policymakers, and ways of implementing policies, which is a “choice” innovation policymakers are more likely to be able to make, but within the broader politico-economic, politico-administrative, and techno-economic contexts (Karo and Kattel, 2018). The empirical section focuses on three East Asian economies that still have institutional memory from the 1960s to 1980s when rather successful catching-up, or global leadership in the case of Japan, was achieved by relatively top-down agenda-setting—prioritizing economic development over social development concerns—and the implementation of industrial and innovation policies. The contemporary relevance of this historical developmental state model has been challenged by several developments in the respective economies and globally: the emergence of more pluralist political systems and competition between different political and economic interests; liberalization and opening of economies; emergence of global value chains that dominate over national policies and economic dynamics; emergence of new science-based industries and new technologies; the growing demand for science- and innovation-based (as opposed to emulation-based) development strategies (Wong, 2011; Yeung, 2014; Chu, 2016). Yet, this article shows that although on the level of policy design, the global–Western discourse has also been entering the East Asian mission-oriented innovation policy rhetoric, this has not stopped the policymakers in Japan, South Korea and Taiwan from relying on old “developmentalist” logics for legitimizing mission-oriented innovation policies, and this has an impact on how specific organizational solutions are actually used for creating dynamic capabilities for implementing these policies.

2. Legitimization and design of mission-oriented innovation policies

Innovation policy scholars who focus on mission-oriented innovation policies seem to agree that the rationales for mission-oriented innovation policies need to go beyond both historical lessons from the post-WWII era missions, such as the Apollo program (Mowery et al., 2010; Foray et al., 2012), and traditional models of market or system failures focusing on improving the general framework condition for firms and industries (Weber and Rohracher, 2012; Mazzucato, 2013,, 2016). Overall, there seems to be a growing consensus that modern “boundary spanning” societal challenges (Arundel et al., 2011; Hicks, 2016; Ulnicane, 2016) and “socio-technical transitions” toward more sustainable techno-economic environments (Geels and Schot, 2007; Markard et al., 2012), possibly delivered by mission-oriented innovation policies, may require policy and governance approaches that balance between the top-down direction giving role of the state and the maintenance of spaces for more bottom-up experimental search. We can also reframe this challenge as how to build innovation bureaucracies that encompass “dynamic capabilities” to coherently both come up with (to explore) novel innovative policies as well as provide patient and long-term visions and policy implementation capabilities to implement (to exploit) these (Mazzucato, 2014; Karo and Kattel, 2016, 2018; Kattel and Mazzucato, this issue).

To date, most analytical work on “how” to create such dynamic capabilities has focused on the analysis of change agents tasked with developing and piloting novel and often experimental policy instruments and/or implementation designs. Existing research of Western experiences has proposed different types of change agents: charismatic policy entrepreneurs triggering changes in public organizations and policies (Leyden and Link, 2015), small peripheral and flexible agencies focusing on experimentation and policy innovations (Breznitz and Ornston, 2013,, 2016), and autonomous mission-oriented and merit-based agencies leading the way to socio-technical breakthroughs (Bonvillian and Weiss, 2015). The archetypical change agent is DARPA, whose organizational design—flat organization with limited formalistic rules recruiting the “best” people to work as mission-oriented (one-term) project managers who select and fund (generously) public and private research groups most likely to find solutions to particular challenges—has been studied at length (Bonvillian and van Atta, 2011; Bonvillian and Weiss, 2015; Bonvillian, this issue). It has been established through broader comparative studies of similar change agents (Breznitz, 2007; Breznitz and Ornston, 2013,, 2016; Breznitz et al., this issue) that the peripheral status of such an organization vis-à-vis traditional ministries and government departments may create the space—with less political interference and public scrutiny of all processes—for policy, organizational, and technological experimentation and learning. The merit-based recruitment and dynamic internal management principles and cultures—rewarding, or at least not punishing, risk-taking and experimentation—create the needed missions and drive for pursuing social and technological innovations. There are at least two critical limitations in the current state of this literature.

First, the focus of this line of research has been predominantly on change agents themselves and not the broader context of innovation bureaucracies as sets of policy actors who are tasked also with implementing the novel solutions developed by these change agents on sufficient scale and scope. As a result, we often tend to focus on how the “explorative” capabilities for innovation are created but neglect the context of how public sectors “exploit” the outcomes of these explorations for systemic impact (Karo and Kattel, 2016).

Second, one has to also recognize that the growing “politicization” of innovation policy—partly created by the expectation of the societal challenges and mission-oriented innovation policy rhetoric (Kattel et al., 2018) —has limited the effectiveness of such an approach, as these change agents are under growing political and public scrutiny and can often deliver only “partial success” before their autonomy gets constrained and mission-blurred (Breznitz and Ornston, 2016).

This brings to the center of discussions the context where such change agents emerge and evolve and how to legitimize the creation and existence of such organizations. DARPA, emerging through trial and error, could eventually rely on the “island-bridge” model whereby the organization itself was protected and thus legitimized by the broader US military system, and it also benefited from the direct “bridge” to the military leadership and its power of public procurement for diffusing its innovations (Bonvillian and Weiss, 2015). In other areas, the United States arguably solved the legitimacy issues by “hiding” its dynamic innovation agents and programs from political and public scrutiny inside the complex technocratic web of the US federal system (Block, 2008).

This approach may not suffice for innovations in more complex and established legacy sectors and for tackling the modern boundary-spanning challenges where there is no single destination (customer) to build the bridge to. Most modern challenges require the participation of different actors, from global and local users and producers to infrastructure owners and regulators, in the diffusion of new innovations (Bonvillian and Weiss, 2015). Although the Obama administration placed societal challenges and missions as one of the core pillars of US innovation policies and tried initially to pursue these through extending the modified DARPA model to other areas (e.g., ARPA-E for energy innovations; Bonvillian and van Atta, 2011; Bonvillian, this issue), in the mid-2010s, it proposed to develop more bottom-up and participatory approaches for government innovation, i.e., co-creating solutions through “innovation labs” and “social innovation” practices, such as citizen science and makerspaces (NEC and OSTP, 2015).

Angel and Rock (2009) argue that such bottom-up processes may be context-specific and suit a few Western systems, if at all, whereas most other countries, especially in Asia, strive to tackle modern challenges by only gradually reforming their existing policy and governance approaches. For example, the South Korean “green growth” initiative—one of the most ambitious green growth initiatives to date—adopted much of the ideational and rhetorical toolbox of Western innovation policy discourse, but in its institutional approach, the policies were arguably pursued through the logic of “developmentalism,” or “developmental environmentalism,” which still prioritized economic development concerns over other social goals and maintained a state-centric approach to policy design and implementation (Han, 2015; Kim and Thurbon, 2015; Seong et al., 2016).

In sum, although such change agents with their specific organizational missions may be necessary for mission-oriented innovation policies to trigger search and change on the organizational level, this article departs from the idea that for understanding how these organizations should be built and how they can deliver the needed dynamic capabilities, we need to first understand the broader contexts and landscapes of different governance systems and how the broader (mission-oriented) innovation policy efforts are legitimized and systematically implemented.

Conceptually, governance scholars approach the question of political and policy legitimacy from two key perspectives (Mazepus, 2018). First, political scientists distinguish between input, throughput, and output legitimacy (Schmidt, 2013). Following this perspective, innovation policies may be among the most difficult to directly legitimize: policy issues and challenges are often too complex to engage nonexperts in policy design (to achieve input legitimacy) and implementation phases (to achieve throughput legitimacy), and the outcomes and effectiveness of policies (output legitimacy) are as difficult to prove (Edler et al., 2016). Second, from a more institutionalist perspective and focusing on a broader context of governance, we can use the Weberian notion of “legitimate domination” to better understand the different ways governments may try to impose their authority and initiatives on society. Weber (1922/2013: 215) distinguished between three pure types of legitimate domination based on:

• rational grounds: “resting on a belief in the legality of enacted rules and the right of those elevated to authority under such rules to issue commands”;

• traditional grounds: “resting on an established belief in the sanctity of immemorial traditions and the legitimacy of those exercising authority under them”; and

• charismatic grounds: “resting on devotion of the exceptional sanctity, heroism or exemplary character of an individual person.”

Most mainstream approaches to innovation policy have been built on the assumptions of “rational” processes of policymaking—policies are designed and implemented through impersonal and impartial processes of discovering specific “failures” in existing markets/systems that provide commonly acceptable and analytically replicable “rationales” for government interventions. This technocratic impersonality and analytical and procedural transparency sustain the legitimacy of the system: input, throughput, and output legitimacy is provided by carefully selected “experts” following the set rules and regulations of policymaking and economic analysis and/or by international comparisons and benchmarking exercises. The current focus on mission-oriented policies and policy experimentation seems to imply that solving persistent societal challenges, which existing policies have not been able to solve, and transformative changes driven by partly noneconomic interests (e.g., sustainable development) may require more radical shifts in policy directions (on directionality, see Mazzucato, 2017) than possibly delivered by such technocratic approaches. Following Weber, such economically uncertain and high-risk policy shifts may have to be legitimized on nonrational and noneconomic grounds (in his view, on charismatic grounds), and this should also affect the governance arrangements emerging from this context. As we show below, the analyzed East Asian countries represent an interesting case of how different forms of legitimization of policies have co-evolved with the set-up of and changes in respective governance systems.

The following empirical sections are based on the following methodological strategy. The next section provides an overview of existing literature on the emergence of innovation policies and bureaucracies in East Asia. Given the space constraints, the discussion focuses on the key “change agents” and their immediate policy and administrative contexts. For our original research on current developments of innovation policy, a desk analysis of key strategic documents drafted since the 1990s (science, technology and innovation strategies, and economic development plans) was carried out to map the emergence and evolution of mission-like priorities and key organizations tasked with specifying and implementing these missions. This was supported by the synthesis of secondary literature on these key organizations (academic sources, documents and reports by organizations, and websites) to get a sense of their specific role in innovation bureaucracies (missions, functions, and interaction and coordination with other public and private actors) and their internal organization. Finally, field trips to the three countries took place in 2016 to interview representatives of key organizations in charge of designing and implementing innovation policies. The goal of these interviews was to corroborate the findings and interpretations of the two phases of prior desk research. The interviews were organized as semi-structured discussions.

In Japan, interviews were conducted with current and past high-level officials (members of boards and director and deputy director generals) from the Council for Science, Technology and Innovation (CSTI); the Ministry of Economy, Trade and Industry (METI); the Ministry of Education, Culture, Sports, Science and Technology (MEXT); the Japan Science and Technology Agency (JST); and the New Energy and Industrial Technology Development Organization (NEDO, in written form). Meetings were carried out with six officials who had each worked in different capacities in two to three of these organizations (normally, one of the key ministries, METI or MEXT, and its respective agencies). These were further supported by discussions (throughout a 1-year research visit in 2016) with academics at the University of Tokyo, Meiji University, and the National Graduate Institute for Policy Studies (who were often former government officials or currently working closely with them). In Taiwan, formal interviews and informal meetings (due to the elections and shift of power, many experts were unable to meet in official capacity) were carried out with representatives of the National Development Council (NDC), the Ministry of Science and Technology (MOST), the Ministry of Economic Affairs (MOEA) and its departmental agencies, the Bureau of Energy, the Industrial Development Bureau (IDB), the Industrial Technology Research Institute (ITRI), and the Science and Technology Policy Research and Information Centre of the National Applied Research Laboratories. In total, 16 officials participated in these meetings. These were further supported by discussions with academics (some of them former officials) at the Academia Sinica and the National Dong Hwa University during two research visits in July and September of 2016. In the case of South Korea, rather similar research had already been published by the Science and Technology Policy Institute’s experts (Hong, 2011; Hwang, 2011; Lee, 2011; Seong, 2011a; Yang, 2011; Seong and Song, 2014; Seong et al., 2016). These findings were corroborated and updated through five interviews (conducted in September 2016) with past and current officials of the Korea Institute of S&T Evaluation and Planning and the Korean Academy of Science and Technology and innovation and governance scholars from the Seoul National University and Korea University.

3. Industrial policy and governance legacies of the developmental state era

After WWII, the starting points were rather similar for Japan, South Korea, and Taiwan: natural resources were lacking and post-war reconstruction challenges were huge. Once the political systems became more stable, all countries developed their early models of industrial science and technology (S&T) policies. Japan set the stage from 1950s onwards (Watanabe and Honda, 1992) and became the source for emulation (next to the United States) for South Korea (by the 1960s) and Taiwan (by the 1970s). These policies were linked to the national strategic efforts to maintain national security and independence through, among other things, S&T autonomy and export-oriented industrialization (especially in South Korea and Taiwan; Wong, 2011). Thus, already these policies had overarching “missions” and were not driven solely by economic rationales but also broader goals of national security and development.

The traditional accounts of East Asian development (from Johnson, 1982, to Wade, 1990, and Evans, 1995/2012), or at least the mainstream reading of them, emphasized Weberian rational bureaucratic models of policymaking and legitimization: capabilities for coherent industrial policies, and autonomy to implement these, were created through Weberian pockets of excellence (key nodal organizations based on merit-based recruitment and career systems; Evans and Rauch, 1999) that were “embedded” in broader industrial networks for policy intelligence and feedback (Evans, 1995/2012). Yet, looking back at the early set-up of these developmental states, we can note much more diverse sets of organizations providing capabilities for both policy implementation and also more dynamic focuses on policy innovation.

In Japan, the famed Ministry of International Trade and Industry (MITI; Johnson, 1982) was not the sole policy planning nodal agent, but through organizations such as the National Institute of Advance Science and Technology (AIST, which was separated from MITI into an independent body only in 2001), it encompassed within its organizational structure also dynamic technological and research capabilities. Its advisory and deliberation councils and soft tools of policymaking (white papers and administrative guidance) fused the lines between public and private as well as political and technocratic initiatives (Watanabe and Honda, 1992; Evans, 1995/2012).

In South Korea, although the Economic Planning Board (EPB) became, next to presidential coordinating bodies, the central coordinating body of economic and financial policies, the country also developed its early S&T bureaucracy. Between 1967 and 1969, the Science and Technology Agency (STA), which in the 1990s became MOST, was set up as a rather flexibly organized umbrella department of the Prime Minister’s Office (freedom to hire top-level staff outside normal rules) to plan and coordinate overall S&T strategies across ministries (Seong, 2011a). Further, for transferring and localizing foreign knowledge and supporting the selected firms, South Korea relied throughout the 1960s–1970s on an extensive network of government research institutes (GRIs), such as the Korea Institute of Science and Technology and the Korean Advanced Institute of Science, established as flexibly organized and formally independent foundations under the leadership of President Park (Hwang, 2011; Moon, 2011).

In Taiwan, although ITRI is often highlighted as its secret weapon (Breznitz, 2007), it was only part of a broader and emerging system of national planning and coordination bodies. Initially, the system centered around the science bureaucracy led by the National Long-Term Science Development Council (NSTC) that was reformed into the National Science Council (NSC) in the late 1960s (Greene, 2008). Eventually, the main logic of industrial S&T policy became to support the export-oriented small and medium-sized enterprise (SME) sector by socializing the innovation-related risks of SMEs. For this, the country relied on IDB under MOEA to provide analytical capabilities and key national research institutes—especially ITRI—to license in promising technologies from abroad and develop and transfer them to firms that would develop export products (Breznitz, 2007; Wong, 2011). This formal innovation bureaucracy was complemented by the network of “parastatal” networks and agencies for technology testing, quality assurance, etc. (Hsieh, 2016).

Importantly, although the concept of “embedded autonomy” was originally linked to the rational/technocratic accounts of policymaking, politicians insulate the developmental bureaucracies and public–private networks from competing/predatory networks to allow them to design and implement coherent development-oriented policies. More recent research emphasizes that the emergence of the developmentalist policy initiatives and key change agents can also be linked to specific charismatic development-oriented politicians who would not only insulate these organizations from competing interest but would also provide charismatic visions and impetus for establishing specific policies and organizations (i.e., Park Chung-hee in South Korea—see Kim and Vogel, 2013—and K. T. Lee in Taiwan—see Greene, 2008). In the case of the more developed Japanese system, one should emphasize not only the economic bureaucracy but also the broader close networks of politicians, bureaucrats, and business elites often making key policy choices outside the formal industrial and innovation bureaucracy (Evans, 1995/2012; Schwartz, 2001).

Given the impact of oil crises and geo-political shifts, the 1970s witnessed the emergence of new types of challenges and missions. Especially the energy dependence and environment-related pressures to transform toward domestic sources, and eventually toward low carbon and renewable energy, became extremely acute in East Asia (Angel and Rock, 2009; Liou, 2010; Dent, 2012; Kim and Thurbon, 2015; Seong et al., 2016). This resulted in intensified S&T policy efforts to create further dynamic capabilities in industrial and innovation bureaucracies, which had to tackle more complex challenges requiring systemic and better organized responses.

In Japan, instead of traditional industrial policy tools [import quotas, foreign direct investment (FDI) restrictions, etc.], support of high-tech R&D consortia (for developing integrated circuits, supercomputers, and similar) became one of the main tools for MITI. According to Callon (1995), these were not necessarily effective, as firms lacked many incentives to join such efforts and MITI’s focuses overlapped in some cases with other organizations such as the STA. In this context, the NEDO was established under MITI in 1980 with a mission to implement the so-called “industrial ecology” approach (Watanabe, 1995): to substitute the key constraining production factor, imported energy, with an unlimited production factor, technology. In 1988, NEDO’s mission and name were extended to include also industrial technology R&D. Compared with traditional generalist ministries at the center of S&T policy efforts, NEDO was established as a mission-oriented, dynamic, and open organization: its funding came mostly from the newly established national energy account (independent from traditional fiscal policy and budgeting processes), and it could recruit private sector experts outside the traditional bureaucratic career model.

In South Korea, the 1979 crisis and assassination of President Park led to a significant re-thinking of the development model and a shift from “modernization” through export-led growth toward the crisis-based narrative of “saving the economy” through “technology drive” (Hong, 2011: 21). S&T grew into a self-standing policy domain next to the EPB-led overall economic planning (Yang, 2011). The newly strengthened STA consolidated the GRI sector and initiated the long-term national R&D program and tried to create incentives for public–private research consortia between GRIs and chaebols (Hong, 2011). Other policy actors—the Ministry of the Postal Service (later the Ministry of Information and Communication) and the Ministry of Trade and Industry—introduced in the late 1980s their own information & communication technology and industrial technologies program (Yang, 2011).

In Taiwan, the global oil crises, loss of international recognition, change of leadership and gradual Taiwanization and technocratization of the political elite, and accumulating pressures from scientists and foreign advisors led to the restructuring of the basic approach to S&T: according to Greene (2008), politicians made bureaucratic planners not only planners but also S&T “leaders.” In 1976, the Council for Science Development was replaced with the Committee of Applied Technology to better coordinate national plans with a clear focus on technology development and diffusion. In 1978, national S&T conferences were initiated that combined foreign, public sector, academic, and industrial expertise to discuss and coordinate these plans. In 1979, the Committee proposed a new long-term national S&T plan that foresaw a more active role for state institutions (research organizations and state-owned enterprises) in supporting private industries and international technology cooperation and transfer. The Department of Industrial Technology (DoIT) of MOEA was created in 1979 as the Science and Technology Advisory Office (and minister-led advisory group; it became DoIT in 1993) that together with IDB sought to plan and coordinate these policies and gradually move from improving industrial infrastructure (developing basic infrastructure and establishing ITRI and science parks in the 1970s) to industrial upgrading (in the 1980s and 1990s). The consultative role of foreign advisors was institutionalized by the creation of the Science and Technology Advisory Group (STAG, 1980) under the Executive Yuan that comprised foreign and local experts to be appointed by the premier for 2–3 years to evaluate policies and programs, facilitate cross-sectoral cooperation, and act as global diplomatic liaisons.

From these brief overviews, we can derive some common patterns for the emergence of East Asian industrial and innovation bureaucracies. The policies and bureaucracies emerged from rather fluid periods of institution building, which combined strong/charismatic political leadership and vision setting and legitimization supported by more flexible and dynamic innovation bureaucracies than often described in the developmental state literature. These bureaucracies included dynamic organizations staffed with people from different backgrounds (from military planners to generalist lawyers and managers and engineers and researchers) acting within different bureaucratic rule sets, or consciously outside of them. In addition, they were complemented by different forms of dynamic networks, such as both domestic and international public–private consultation bodies and close links between politico-administrative and business elites. From the late 1970s onwards, these initial fluid forms of innovation bureaucracy became more institutionalized as specific ministries and designated agencies started to emerge and take the central stage in policy planning, requiring more patient and specialized focuses. As a result, also the charismatic legitimization of policies by few key political actors became less relevant next to the more prevalent rational-grounds-based technocratic style of policymaking and structuring of innovation bureaucracies.

4. Emergence of modern innovation policies and bureaucracies

By the 1990s, all three countries faced increasingly complex policy environments. Democratization in South Korea and Taiwan and administrative reform attempts in Japan increased the political competition for resources and weakened the public–private developmental networks. Liberalization, development of firms, and their growing integration into global networks increased the autonomous innovation capabilities and power of firms vis-à-vis governments and made policy challenges more complex. Furthermore, innovation policies and bureaucracies evolved in the context of recurrent crisis narratives: Japan was most influenced by the asset bubble collapse in the early 1990s and the 2011 Great East Japan Earthquake; the 1997 Asian financial crisis had the biggest impact on South Korea; for Taiwan, the most influential events have been the 2008 global financial crisis (GFC), environmental disasters, and the broader impact of the 2011 Great East Japan Earthquake.

The 1990s and early 2000 were characterized by the emergence of Western-like, or global-mainstream, innovation policies and bureaucracies. In the mid-1990s, all three countries introduced laws and routines for drafting R&D and innovation strategies that continue to this date. In 1995, Japan adopted the S&T Basic Law and started to adopt the 5-year Science and Technology Basic Plans. The third plan (2006–2010) was further complemented by “Innovation 25” as a long-term strategic guideline looking into 2025 (Stenberg and Nagano, 2009). In 1997, South Korea adopted the Special Law for Science and Technology Innovation and introduced the Science and Technology Innovation 5-Year Plan. In the early 2000s, it adopted Vision 2025: Korea’s Long-Term Plan for Science and Technology Development, which became the basis for subsequent 5-year Science and Technology Basic Plans. This led to one of the highest global growth rates of public R&D investments between 1997 and 2006 (Lee, 2011). In Taiwan, the 1999 Fundamental Science and Technology Act required the organization of national broad-based S&T conferences every 4 years and revisions of national policies every 2 years through adopting either national S&T Plans or white papers. This was complemented by a significant increase of S&T funding and the organization of national R&D programs, e.g., in telecommunications, energy, and different strands of biotechnology (MOST, 2014).

In all countries, the first innovation strategies (until the mid-2000s) were trial-and-error-like and rather technocratic efforts to introduce new priorities to innovation policies. The first common goal was to either re-take global innovation leadership (Japan) or become one among the core group of global innovation leaders (South Korea and Taiwan). The second goal was to increase the overall public funding of R&D and innovation. Third, all countries tried to define technological priorities, but this often led to “bucket lists” of tens of overlapping and competing technologies supported and funded by different government agencies and difficulties in setting clear priorities (on Japan, see Stenberg and Nagano, 2009; on South Korea, see Choi and Choi, 2015). In Japan and South Korea, these formal policies and strategies were paralleled throughout the 1990s and 2000s by several administrative reforms of new public management influenced by Western best practices.

In Japan, both the 1990s and especially the early 2000s saw attempts to redesign the Japanese administrative system to formally reduce the power of bureaucracy vis-à-vis political institutions (Takenaka, 2008): MITI was reformed into METI, the Ministry of Education and STA were merged into MEXT, and many of the existing bureaus and agencies (such as AIST, but also research centers and universities) were reformed into arms-length administrative agencies. The political coordination and oversight was to be provided by two innovation-related high-level councils established under the Prime Minister: the Council for Economic and Fiscal Policy (which sets overall targets for the national budgetary process) and the Council for Science and Technology Policy (CSTP; before acting as the advisory Council for Science and Technology) to take the role of the “control tower” for S&T policy through constant monitoring and coordination of priorities and activities of different actors, especially METI and MEXT.

In South Korea, the EPB was dismantled in 1994 and since then, there have been continued attempts to bring back such coordination capabilities and top-down planning under the new innovation and governance logics. Between 1998 and 2008, this was attempted through the newly established MOST and the presidential National Science and Technology Council (NSTC; composed of public and private sector members), which became an important coordinator of national innovation funding ceilings (Hong, 2011; Seong, 2011a). Between 2004 and 2008, a new administrative unit—Office of S&T Innovation—existed inside MOST to create a new policy level and a relatively dynamic change agent (staffed with 40% MOST employees, 40% from other ministries, and 20% from the private sector) supporting, together with the Science and Technology Ministers Meeting, the work of NSTC in a logic quite similar to the EPB, but with a clear mission and a focus on innovation. Yet, local experts claim that such top-down planning was not very effective in the context of uncertainty of innovation (Seong and Song, 2014). In parallel, the by default one-term presidents became more involved and interested in S&T policy by setting their own top-down innovation agendas, i.e., “innovation drive” by Kim Dae-jung (1998–2003) and “S&T Based Society” by Roh Moo-hyun (2003–2008).

Taiwan has followed a peculiar non-Western path by not establishing classic R&D funding agencies, but by coordinating funding and selection processes at the level of political coordination bodies and ministries. Over the course of the late 1990s and early 2000s, it was recognized that the NSC-centered coordination and control of predominantly bottom-up R&D and innovation initiatives may be insufficient, especially for pursuing nationally critical priorities, such as transition to knowledge economy, environmental sustainability, and quality of life. The early strategic documents sought for different mechanisms to improve top-down planning and coordination of R&D and innovation projects through different high-level meetings and committees in areas like biotechnology, energy, industrial development, more pro-active review of bottom-up drafted innovation programs, and budgets. The most important organizational reform was opening national science and technology programs in priority areas. Since 1998, Taiwan has implemented—with different degrees of success—13 large-scale national R&D programs, which have sought to combine public and private needs as well as existing R&D and academic capabilities in a more coherent and coordinated way.

In sum, the 1990s were characterized by further attempts to build Western-like rational-grounds-based systems of innovation policymaking and bureaucracies (with representative politicians and business elites providing oversight and legitimization) and an increasing emphasis on market-based developments and governments limiting their roles to fixing market failures and improving framework conditions of innovation systems. Yet, the outcomes of these attempts seem to not have satisfied policymakers and broader innovation communities and since the late 1990s, attempts emerged to return to more top-down styles of priority setting and planning.

5. Shift toward mission-oriented innovation policies and bureaucracies

The mid-2000s saw a shift toward policy rhetoric emphasizing societal challenges—aging of societies and debates on the contradiction between economic versus social and environmental developments—and gradual return of politicians-led top-down developmentalist ideas and initiatives, which differed from the Western patterns of continued technocratization and rationalization of innovation policies. Many of these ideas were jotted down already in government strategies and documents drafted in the early 2000s, but it took different external events—changes of political leadership, financial crisis, and Great East Japan Earthquake—to trigger more systemic shifts in innovation policies.

5.1 Japan

The “Innovation 25,” a vision document led by Prime Minister Abe (Government of Japan, 2007), framed the main challenges of Japan for the next 20 years: aging and population decline in Japan; explosive advancement of globalization and knowledge- and information-based society; and threats to the sustainability of earth posed by population growth, resource and energy use patterns, climate change and environmental impacts, global water and food shortages, and threats from terrorism and infectious diseases. Furthermore, it explicitly emphasized the “social context” of innovation, or how innovations can solve the problems of people in Japan and globally. In this context, Stenberg and Nagano (2009) show how the Cabinet Office gradually took a more entrepreneurial policy role. CSTP introduced top priority policy issues (e.g., transformative and low-carbon technologies) as an additional layer of prioritization and used the Special Coordination Fund (about 1% of the total and 10% of competitive STI funding) to initiate cross-ministerial coordination projects. It used the more flexible processes of supplementary budget drafting—in 2009, R&D and innovation funding through supplementary budget equaled 38% of annual public investments—to initiate more proactive and dynamic initiatives (i.e., Transformative Technologies Fund, the Funding Program for World-Leading Innovative R&D on Science and Technology, and FIRST) and coordinate “all Japan efforts” in high-priority innovation projects. Overall, the political leadership became increasingly involved in the coordination and planning processes. For example, Stenberg and Nagano (2009: 84) report that regarding the selection of the principal investigators for the FIRST program, the acting Prime Minister Aso claimed that “I will make final decisions myself when it comes to choosing the central researchers and core research themes.” Although some of these initiatives were relatively short-lived due to the government change, they eventually became core innovation policy ideas of “Abenomics.”

The Fourth S&T Basic Plan (Government of Japan, 2011) was adopted under the new government by the Democratic Party of Japan (in power between 2009 and 2013). Its economic policy program, New Growth Strategy (Government of Japan, 2010), focused on “achieving economic growth by turning the problems faced by the economy and society into opportunities for creating new demand and employment.” In this context, the Fourth Science and Technology Basic Plan (2011) presented a “crisis” narrative regarding both global issues (resource scarcity and political and economic instability) and Japan-specific issues (aging, declining population and birth rate, and economic stagnation). The plan shifted R&D and innovation prioritization even more toward an explicit problem-solving approach with a key focus on two areas of Japanese strengths—green innovation (developing low-carbon energy sources, green social infrastructure, and improving the efficiency of energy resources) and life innovation (medical and nursing care and health services)—and tried to integrate previously prioritized technology domains (environment, energy, and health) and policy logics (solving societal challenges while sustaining international economic competitiveness).

The return of Prime Minister Abe to power resulted in the revision of most economic policies under “Abenomics” and the growth-oriented Japan Revitalization Strategy “Japan Is Back” (Government of Japan, 2013). The Industrial Revitalization Plan of the strategy sought to induce structural reforms (in universities, regulatory environment for innovation) to create opportunities for existing industries to find new growth paths and to pursue “all-Japan-efforts” in new innovation frontiers. The Strategic Market Creation Plan of the strategy was based on the logic of creating new domestic and global markets by tackling current social issues. The Fifth S&T Basic Plan (Government of Japan, 2016) formalized these ideas and proposed a new paradigm for innovation, i.e., “realizing a World-leading ‘super smart society’,” or Society 5.0, focusing on long-term state-coordinated development of different socio-technical systems and the integration of different technologies from internet of things (IOT) and artificial intelligence (AI) to bio- and nanotech.

Paralleling these strategic shifts, there have been several attempts to make the Cabinet Office the new strategic and dynamic “headquarter” of Japanese innovation bureaucracy by establishing different mission-oriented and dynamic units to oversee policy initiatives from supporting ICT diffusion and medical innovations (e.g., Office for Healthcare and Medical Strategy) to innovation policy in general. In 2014, the CSTP was renamed the CSTI, and there have been attempts to increase its role in budgeting processes (e.g., coordinating cross-ministerial priorities before these are submitted to the Ministry of Finance), initiating and overseeing a “new high-risk and high-return” innovation project as part of the Impulsing Paradigm Change through Disruptive Technologies Program (ImPACT), and in coordinating the cross-ministerial national challenge projects as part of the Strategic Innovation Program (SIP). SIP continues the top-down cross-ministerial coordination initiatives of the CSTP (funded by the Special Coordination Fund), and ImPACT carries forward the ideas of the FIRST program. The ImPACT program tries to emulate the DARPA approach with a program director-based model for investing in the “high-risk high-impact” topics (Government of Japan, 2014; ImPACT, 2016). An important difference is that under the CSTI, these programs and projects are potentially under much stronger public and political scrutiny. The SIP program seeks to implement specific projects in national priority areas agreed by the CSTI (e.g., energy and infrastructure development; SIP, 2016) through comprehensive coordination of funding, regulatory reforms, and market support between different ministries to speed up the processes of innovation and diffusion. Both programs seem to be legitimized through not only high-level political participation (Cabinet Office) but also the active advisory and managerial roles for high-level business representatives and entrepreneurs as well as by setting the Tokyo Olympics as the symbolic deadline.

5.2 South Korea

The late 2000s and GFC overlapped with significant policy shifts also in South Korea. President Lee’s agenda, summed up in the National Strategy for Green Growth (2009–2050), the Five-Year Plan for Green Growth (2009–2013), and the 2009 stimulus package (Green New Deal), included significant commitments to green growth (Jang, 2010) and an explicit focus on innovation as a tool for bringing about the new development paradigm. In many green growth technologies, the key policy focus was on local content and technological self-sufficiency (Kim and Thurbon, 2015).

These ambitious initiatives were complemented by significant reforms of the innovation bureaucracy. The Lee administration attempted to re-create a new EPB-type centralized change agent through the establishment of the Presidential Committee on Green Growth, which was a dynamic network bringing together policymakers and private sector experts (Kim and Thurbon, 2015). At the same time, it lacked EPB-like control over budgeting processes, and the critics of the agenda (Han, 2015; Seong et al., 2016) have argued that the actual implementation remained to a large extent driven by old legacies (focus on supporting existing industries through infrastructure investments), especially as the civil servants of the committee represented their own ministries (Seong, 2011b).

In the aftermath of the GFC, growing criticism of the elitist policies of President Lee provided a further window for innovation policy changes and “the ‘social’ aspect, which tended to have been overlooked in the policy sphere, began to be debated within innovation policy” (Seong et al., 2016: 5). The incoming President Park Geun-hye (daughter of President Park Chung-hee) soon defined its own developmentalist agenda, “Creative Economy,” which focused on S&T, ICT, and SME-led job and market creation and paid lip service to both the modern open innovation logic and to the “Park Chung-hee nostalgia” (Kim and Thurbon, 2015) and the “Miracle of the Han River” narrative. To implement this agenda, the administration initially attempted a strongly top-down approach formalized in the Science and Technology Basic Plan 2013–2017 and the Three-Year Plan for Economic Innovation (2014). In 2013, the government tried to create a mission-oriented policy level change agent by re-establishing the Ministry of Education and Ministry of Science, ICT and Future Planning (MSIP)—which was merged into the Ministry of Education, Science and Technology under president Lee—to oversee the coordination and implementation of the Creative Economy agenda.

Although the Creative Economy agenda became a central focus of innovation policy (it was estimated that ca. 45% of 2015 government outlays for R&D were related to the agenda: see Cha, 2015: 41), several observers argued that different actors found it difficult to operationalize this emergent philosophy in a logical manner. For example, one of the key organizational innovations of the agenda was the introduction of topic-specific centers for Creative Economy and Innovation in different regions as public–private partnerships between central (MSIP) and local governments, venture businesses, and one key chaebol in each center with the main purposes of nurturing start-ups, boosting innovation of SMEs, and developing innovation ecosystems. Yet, the links of these new centers with existing institutions—from top-level bodies such as NSTC to universities and GRIs and already existing regional centers under Ministry of Trade, Industry and Energy (MOTIE)—remained weak and resulted in overlaps between different networks and limited synergies.

5.3 Taiwan

The 2009–2012 S&T Plan (NSC, 2009) recognized that it will be increasingly difficult to maintain prior rates of economic growth, and the unbalanced economic, environmental, and societal impacts of the “industrialization first” strategy require rethinking the R&D and innovation strategies, and that shifting from a “technology-oriented” to a “needs-oriented” model with “sustainable development” and “quality of life” concerns taking a more important position. This shift in policy rhetoric coincided with the return of Kuomintang (KMT) to power in 2009 (after being in opposition since 2002) and the severe impact of the GFC on Taiwan (Rosier et al., 2016).

Although the 2011 Golden Decade National Vision (integrated into the National Development Plan of 2013–2016) looked at economic, social, and environmental goals from a new perspective, e.g., changing the growth model from efficiency to openness and innovation and substituting the indicator of gross domestic product (GDP) with GNH (gross national happiness), actual policies still “picked” specific industries to prioritize—e.g., biotechnology, green energy, high-end agriculture, tourism, medicine and healthcare, culture, and creation. Given the difficulties of exiting from the impacts of the GFC and the overall debatable performance of the innovation policies and system, the 2013 S&T Plan (MOST, 2013) proposed that “If economic growth could no longer guarantee to secure happiness, the distribution of resources should strike a balance between social welfare and economic development.” The vision proposed in the 2015 White Paper on Science and Technology (MOST, 2015) further departed from the “economic growth first” narrative by proposing the vision of “using intelligent technology to create a prosperous society and achieve sustainable growth” and discussed a new socio-technical concept, “low-carbon intelligent society.” Although the concept was not explicitly defined, it was used as an umbrella term for the strategic activities aimed at establishing Taiwan among global leaders in green technology (MOST, 2015: 166–170). According to local policymakers and scholars, part of this evolution is related to the ongoing processes of democratization leading to more substantive debates over the content and direction of economic development and greater access of public and interest groups to these deliberation and choices (Rosier et al., 2016).

The governance reforms, while at times paying lip service to social innovation and bottom-up participatory approaches, have focused on strengthening the role of traditional innovation bureaucracy and top-down policy planning and control. In 2012, the government formed a new high-level coordinating body—the Board of Science and Technology (BOST) of the Executive Yuan—which is presided over by the Executive Yuan Premier and includes heads of key agencies, members of industry, academia, and research organizations. In principle, BOST institutionalized (explicit organization, own staff, and regulated tasks) prior more flexible and fluid policy coordination activities by STAG and other networks. As a body presided over by the premier, its role and tasks depend on the policy priorities of the premier and government in power. These may also overlap with the NDC that was also created in 2012 as a sort of “mini-cabinet” to debate before formal Executive Yuan meetings and new social and infrastructure investments (e.g., the NDC has deliberated the development of wider prospective industries, such as IoT). In 2014, MOST was established based on the former NSC, but with additional roles and competencies regarding foresight, innovation policy, and academia–university collaboration. The creation of MOST has raised the political salience of S&T and innovation policies because when NSC was under the Executive Yuan, MOST has to be accountable also to the Legislative Yuan. With these two organizational changes, the Taiwanese innovation has become relatively mainstream, spearheaded by political-level coordinating institutions (government meetings, BOST, and NDC) and based on two key ministries and their policy advisory networks (e.g., the National Industrial Development Conference and the Science and Technology Development Advisory Conference, which has proposed new approaches to policymakers, i.e., to adopt the DARPA approach in mission-oriented projects—see MOST, 2015: 26, 27).

A noteworthy element of the Taiwanese system is the lack of R&D and innovation funding agencies under the respective ministries. MOEA’s different departments and bureaus are closely linked to public research organizations, such as ITRI. MOST funds universities that are formally under the Ministry of Education and Academia Sinica, which is under the Office of the President. The government has tried to introduce new policy instruments that adopt more flexible and risk-tolerating organizational principles (from selection to management and performance indicators) to foster cross-sectoral collaboration in high-risk and future-oriented domains. NSC/MOST and MOEA jointly launched the PIONEER Grants for Frontier Technologies Development by Academia–Industry Cooperation to undertake high-risk R&D projects in forward-looking technologies and as defined by industry interests (MOST, 2015: 110–114).

Although most national S&T programs focusing on societal issues have been closed down, the only continuing program—the Taiwan National Energy Program (NEP)—was significantly revised in 2013. Whereas the first phase of NEP—organized under the NSC—was organized in a relatively bottom-up manner and it was largely managed by university scientists themselves, the second phase of NEP has been co-financed by MOST, MOEA, and the industry and implemented under the closer supervision of MOEA (the principal investigator is its vice-minister). Also, the “outsourcing approach” (open calls for project ideas) has been substituted with a more mission-oriented approach whereby the general and domain leaders plan R&D priorities and select teams to fulfill these projects following a clearer phase-out mechanism and performance targets (MOST, 2015). Yet, as energy-related activities have been an important policy focus of MOEA since the late 1960s, it has also developed its own policy strategies, white papers, and initiatives, and representatives of ITRI claim that as opposed to S&T white papers and strategies of MOST/BOST, they still follow energy industry white papers and other related guidelines.

5.4 Discussion

In all cases we can witness, at least since the mid- or late 2000s, attempts to introduce new missions and visions into innovation policies and develop innovation bureaucracies by introducing more dynamic change agents and policy instruments. Yet, as opposed to the Western practices of participatory styles of policymaking and the creation of peripheral change agents, all three countries seem to follow models closer to their developmental state legacies.

On the level of policy legitimization, policymakers in all three countries have tried to leverage the historical state-guided development narratives and try to argue through national strategies that modern challenges require a similar vision-setting role from the state. Thus, instead of relying on technocratic approaches for policy rationalization and legitimization, emerging in the late 1980s and 1990s, they seem to bet on the more historical cultural expectations for state guidance in socio-economic development.

On the level of policy implementation, policymakers also try to rely on developmentalist governance designs centering on authority-based and high-level bureaucratic agencies (offices of the executive leaders and agencies and networks publicly and closely linked to them) that should set the common policy missions for different actors and coordinate their tasks. Given the growing technological capabilities of firms, there is visible engagement of business elites in high-level decision bodies to complement political acumen with business acumen and legitimacy. Thus, as opposed to Western models where peripheral agencies and autonomous “islands” for experimentation and innovation are expected to be the key policy innovators, the East Asian innovation bureaucracies seem to rely on more visible and politicized organizations. Such an organizational design may be effective for triggering short-term changes in innovation policy and complement the prevalent pathways for legitimizing these policies. Yet, building and sustaining long-term dynamic capabilities—i.e., the coherence of explorative and exploitive capabilities necessary for mission-oriented innovations—may be much more challenging in such visible and central organizations than in more peripheral experimentation-focused islands.

Many Japanese observers noted that although the Cabinet Office has attempted to increase its role as a dynamic change agent and strategic steering center, its own organizational and policy capabilities might be too generic and fragmented to actually trigger the desired changes and see them through. In the case of the most high-level SIP and ImPACT programs, CSTI has relied on the agencies of METI (NEDO) and MEXT [Japan Society for the Promotion of Sciences (JSPS) and JST] for actual project management and implementation. At the same time, METI and NEDO pursue their own approach by developing strategic technology roadmaps and related policy tools, which look more incremental innovation-oriented than the CSTI high-risk high-return and experimentation rhetoric.

In the case of South Korea, although the presidents have the power to propose new developmental agendas and missions (which they might also use for legitimizing their own broader power, given the context of the South Korean techno-nationalism; see Wong, 2011) and create high-level dynamic agents to support them, they also face increasingly complex policy tasks and already established innovation bureaucracy with its own priorities and routines. For example, the Third Science and Technology Basic Plan (2013) and overall R&D budgeting processes focused on much broader sets of long-term priorities that are likely to outlive the presidential “Creative Economy” agenda. This creates policy confusion and competing missions between the politically determined and bureaucracy-led parts of innovation policies.

The developments of Taiwan seem to be the most open-ended, as its policy and governance reforms are the most recent, and we can find in parallel the rhetoric of increasing top-down policy prioritization and legitimation and the creation of more participatory spaces in innovation policymaking for rethinking the national developmental model as a whole.

6. Conclusion

The development of mission-oriented innovation policies requires a focus not only on thinking about what types of missions to set, but also on the local context of politics and governance. Current missions of innovation policies embedded in public values, such as sustainable development, seem to require nontechnocratic and nonrational pushes by charismatic-authority-based movements or institutions and supportive change agents for policy innovations. Yet it is also important to make sure that these will be complemented with a supportive bureaucratic apparatus for the actual implementation of the policy innovations. Or, to quote Peter Evans (1995/2012: 30): “If Weber is right, imposing different policies on a state apparatus without changing the structure of the state will not work. Real changes in policies and behavior depend on the possibility of erecting new state structures.”

Despite the growing consensus among innovation policy scholars that modern societal challenges and politico-economic contexts require new types of (mission-oriented) innovation policies, this has not stopped the policymakers in Japan, South Korea, and Taiwan, after a period of relative Westernization and rationalization of their innovation policies around the 1990s, from introducing legacy-driven developmentalist policy logics for legitimizing and implementing their innovation policies. This has resulted in the (re-)emergence of a model of highly politicized innovation policy rhetoric—changing the basic philosophies of national development and societies though government-led innovations—as well as governance and implementation structures based on highly visible and elevated change agents.

Although this style of policy legitimization seems to fit the Asian legacies of politics and policymaking, especially for legitimizing these policy initiatives outside the innovation policy communities, where references to DARPA and other international best practices will probably make little sense, choosing top-down governance systems and politicized change agents remote from the actual policy implementation organizations and capabilities is a more debatable and questionable choice. Without supportive state structures focusing on long-term implementation of the new missions and policy innovations emerging under the umbrellas of these missions, it might easily be that the new ideas of super smart or intelligent societies striving toward ever greener economies remain short-term buzzwords resulting in spectacular policy failures (or just government waste) and subsequent de-legitimization of the role of the state in innovation.

Funding

The work was supported by the Estonian Research Council grant IUT19-13 and the Japanese Society for the Promotion of Sciences KAKENHI grant number 15F15760.