Conclusion
East Asian states have been widely studied for their developmental trajectories and their different state-led strategies as well as to determine whether innovation was led by startups or big business. The region of developmental states offers two archetypal examples of large firm–led systems (Japan and Korea), a small firm–centric approach (Taiwan), and a mix of state-owned enterprises and private entrepreneurs (China). In their antecedent periods, which largely correspond to their classic developmental state eras, this large- or small-firm focus was reinforced with complementary institutional dynamics across employment, finance, innovation, and social purpose.
Looking at the four cases in comparative terms, how can we understand the continuity and change experienced from the antecedent era to their pursuits of startup capitalism? To inform the analysis, figure 6.1 illustrates all four cases’ antecedent conditions in terms of the radar chart used in the case chapters. The Mark II type constitutes the starting point (0) for the quintessential developmental state, which Japan and Korea approximate most closely in terms of firm size central to innovation (large), the financing of innovation (debt), the employment (lifetime), and the nature of innovation (incremental). In contrast, the Mark I setting is depicted as constituting the opposite ends of the spectrum in terms of the firm size central to innovation (small), the financing of innovation (equity), employment markets (fluid), and the nature of innovation (radical).
In figure 6.1, Japan and Korea have the same starting scores, indicated by their full black lines touching on 1s for each arena other than social purpose. The antecedent period in Taiwan was closest to the Mark I variety and thus is the most distant from the Japanese and Korean depiction in figure 6.1. Taiwan’s antecedent score reflects the fact that financing (equity), employment (fluid), and efforts to compete in the technological frontier while advancing process-based innovations (mix of incremental and radical innovation) combined with an emphasis on small firms as essential engines. In between these two shapes are China’s antecedent characteristics. China is in the middle because of its mix of large state-owned firms and small-scale entrepreneurs, the availability of a combination of debt and equity financing, the relatively fluid labor (given early promotion of entrepreneurial careers), and its pursuance of “catch-up” innovation (incremental). This gave it a mix of Mark I and II attributes and thus scores ranging between 1 (for its catch-up innovation), 2 (for firm size and social purpose), and 2.5 (for financing and employment).
FIGURE 6.1.Antecedents in Japan, Korea, Taiwan, and China
From these distinct starting points, did East Asian states converge on an idealized version of Mark I? Does this examination of startup capitalism serve as further evidence of a convergence on the LME variety of capitalism? Or do East Asia’s approaches to startup capitalism align better with the reality of Silicon Valley? By this, we mean that startups are largely perceived as resources for big businesses that have only expanded their competitive positions. Do the trajectories mostly suggest continuity in each state’s own approach?
Let us start with Japan. Since the Heisei recession in 1990, the Japanese developmental state has increasingly promulgated startup largesse. But rather than a rejection of its large firm–led developmental state apparatus in which pension savings and main banks reinforced lifetime employment and incremental innovation prowess, there has been continuity. Japan’s startup capitalism is one of fostering open innovation systems in which the keiretsu and main banks have emerged as crucial partners, investors, and beneficiaries. Prominent programs like the 2018 J-Startup Initiative and the 2020 Startup Ecosystem Consortium center on the participation of keiretsu and main banks, which play vital roles in selecting startups and are recognized as essential investors and potential acquirers. Prime Minister Kishida’s 2022 “Startup Development Five-Year Plan” even emphasized that large, existing companies engage with startups in open innovation as a means of sustaining their business operations.
Thus, Japan’s startup capitalism constitutes some change but not a radical departure from its developmental state starting point, as illustrated with a modest shift outward in figure 6.2. More equity-based investment vehicles have been promoted, lifetime employment has been modestly eroded, and the nature of innovation has, for many years, been at the world’s technological frontier. However, the size of firms still regarded as essential to national innovation capacity remains large. The keiretsu are beneficiaries of startup activities, including talent. The financiers of the previous epoch are involved in startup capitalism as active investors, judges, and ecosystem actors. It cannot be said that the keiretsu fear competition from startups nor that the Japanese government has discontinued its backing of big business.
Korea was the second country examined, given that it is also often characterized as an archetypical developmental state. In the chaebol-dominated era, the state acted to encourage the upgrading of the technological capacity of SMEs that could provide high value-added employment in addition to being suppliers to the internationally competitive chaebol. The Korean context closely aligned with an open innovation variety of the Mark II mode. Since the EAFC in the late 1990s, policies have publicly attested to a rising dedication to startups as employers and innovators. Every government since the crisis, whether liberal or conservative, has stated its championing of startups. High-profile initiatives, such as President Park Geun-hye’s implementation of the Creative Economy Action Plan and President Moon Jae-in’s establishment of the MSS in July 2017, have amplified those messages in a very public way.
While institutional change has occurred, particularly in the extent to which innovation is now radical, equity is more prevalent, and the employment market is more fluid, the Korean startup capitalism mode remains best depicted as fitting the open innovation variant of the Mark II type. Startup initiatives often foster partnerships between startups and chaebol, with the aim of injecting innovative DNA into the country’s leading firms. Startup capitalism offers a different, and less publicly visible, way of backing large firms than the credit-centric tools linked to exports in the classic developmental period. Sentiment expressed across the media suggests that the chaebol are aware that the promotion of startups by successive governments aims to reinforce them and expand their offerings through the integration of new technologies rather than foster their replacement by new firms. While startups are presented as key beneficiaries of innovation policies, we observe that the chaebol often act as vital partners and ultimate recipients of this help. The K-Startup Grand Challenge, for instance, sees a licensing agreement with a chaebol as a KPI. The Korean startup capitalism mode, then, has seen movement outward in several institutional arena, as illustrated in figure 6.2, but remains most closely aligned with the Mark II variety (one fueled by open innovation logics).
In the antecedent period in Taiwan, the state stands out relative to the others in East Asia in its early orientation toward small technology-centric firms as central to technological upgrading capacity. The government’s policies focused on enhancing the capabilities of small firms by incentivizing R&D and export activities. Entrepreneurship was encouraged as it provided a form of employment; early-stage equity markets for high-growth startups were developed already in the 1980s, and government entities facilitated the competitiveness of nascent firms by acquiring foreign technology and transferring it from public institutes like ITRI. However, by the turn of the twenty-first century, the government publicly advocated for startups in emerging sectors and technologies, and not in collaboration with, or necessarily for the benefit of, large firms (like TSMC) or their technologies. The primary focus has been on widening the entrepreneurial pool by attracting foreign talent and directing attention to emerging technologies. The government routinely asserts that the success of Taiwan’s personal computer and semiconductor-focused miracle has prompted them to encourage startup activity in emerging technology sectors such as biotechnology, LCDs, and green tech.
Thus, Taiwan’s startup capitalism began as the closest to the Mark I ideal type and remains the most oriented toward startups as innovators in their own right. Taiwan’s social purpose in helping startups remains to encourage more new entrants who can disrupt firms in world markets—but not domestic firms.1 The aim of backing startups is not to prop up domestic industries or firms, nor is it to challenge them. The aspiration to achieve more radical innovation has grown, shifting from acquiring foreign technology and catching up in manufacturing to competing as front-runners in design capabilities in the semiconductor industry. However, large firms like TSMC, despite its essential position in the world’s critical technology, are not considered the ultimate beneficiaries, unlike such an orientation in the two East Asian developed counterparts. Collectively, Taiwan’s movement from antecedents to startup capitalism reflects the outward shift toward (more) radical innovation and the persistence of its focus on small firms, equity financing, fluid employment, and external security threats as its great motivator.
China’s political economy is often depicted as opposing the neoliberal approach of the United States, with the state actively shaping firm-specific capabilities and allocating resources toward critical technologies. Its antecedent period is characterized by its mobilizing of resources to bolster catchup technological capabilities across a combination of firms, including state-owned enterprises and large private firms. This gave the Chinese antecedent more of a Mark II character. Even in the antecedent period, however, startups were already conceptualized in playing a role, and clusters of innovative startups were developed across major cosmopolitan areas. So while the antecedent period was closer to the Mark II ideal, it did not fully align with that ideal type.
The striking finding in the Chinese context is the magnitude of the change. While all cases experienced movement in the direction of Mark I, the size of the evolution in the Chinese case is the greatest. Yet, like the other cases, it comprises a mix of Mark I and II elements. This is seen as policy makers not only encourage new entrants to foster capabilities in emerging technologies but also strive to dismantle the dominant positions held by giant corporations like Alibaba. In the consumer internet or platform economy, efforts encourage China’s entrepreneurs to fight against each other as well as against industry leaders for their place. This consternation has been mirrored in market dynamics. The Economist (2024a) notes that “capital for the consumer internet has all but dried up,” while, in the same period, “hard-tech developers have collectively raised about 550bn yuan through initial public offerings.”
In technologies like semiconductors, renewable energies, or 6G, the state finances its large would-be national champion firms, such as SMIC, which was given a US$282.1 million subsidy in 2022 (Cao 2023). This is a telling example, as it reveals that SMIC was far from being the sole recipient; more than 190 Chinese companies operating in the semiconductor industry received funding as part of a US$1.75 billion package. This semiconductor package is in some ways emblematic of China’s startup capitalism; large companies receive financing, but so do would-be startups operating in the space. What is more, China’s crackdown on the oligopolies of large private firms is consistent with China as a Mark I variety only in the sense that the destruction of dominant positions occurs. However, this is not fueled by startups as engines of this existential threat. At the same time, the Chinese approach includes the marshaling of state resources in key technologies, which represents a continuation of some of its Mark II attributes from the antecedent period. As a result, the Chinese evolution toward startup capitalism contains a mix of contradictory elements from both the Mark I and II varieties.
Collectively, the continuity and change trajectories reveal movements of interlocking institutions underpinning East Asia’s antecedents through to the startup capitalism modes in each country. Overall, relative to figure 6.1’s concentration by Japan, Korea, and China in the center of the radar chart, figure 6.2 shows that each country has adapted such that employment markets are now more fluid, innovation aims are more radical, and financial systems are more comprehensive, with equity offerings alongside the provision of credit. There are varying degrees of outward movements toward small firms as the locus for innovation activities, with Taiwan remaining so and China moving more in that direction. For most countries, the motivating social purpose has remained a mix of external and domestic elements, but the national security imperative has grown more noticeably for both China and Taiwan compared to Japan and Korea. In this sense, there has been a convergent trend across the countries, with hybridity moving in the direction of, but not coming close to reaching, Mark I paradigms.
What should we make of this continuity and change? One read is that there is an overall convergence toward a Mark I mode across the cases. None of the cases have seen their trajectories shift inward, toward the Mark II ideal type (e.g., no scores have gotten closer to zero). What is more, we do not observe a mere continuation of previous approaches. No country’s scores remain at the same position across their antecedent and startup capitalism eras, as illustrated in figures 6.1 and 6.2. There have been shifts away from Mark II logics in all four countries. Each case has seen movement in the direction of Mark I in terms of radical innovation and the inclusion of equity financing and startups as innovation engines. In Taiwan, there has not been a retreat of its small firm antecedents, even though the economy is dominated by just a handful of firms that operate in a small set of industries, with TSMC’s position in the fabless manufacturing of semiconductors being the most prominent and well-known. Startup initiatives strive to bolster the widening of entrepreneurial pools in emerging industries, like biotech and green energy, not as innovative DNA to inject into the semiconductor giants.
FIGURE 6.2.Startup capitalism in Japan, Korea, Taiwan, and China
However, it would be wrong to conclude that there is a broad convergence on the Silicon Valley–styled mode (Mark I). As we showed in the Japan and Korea cases, startups fuel the pursuit of large firms’ capacity to compete at the world frontier in a more complex and competitive international environment. Incumbents themselves provide equity financing for the startups that benefit them and serve as the beneficiaries of more fluid labor markets, where startups help them develop their long-term access to talent. In at least these two countries, the expanse away from 0 and toward 5 on the radar chart is in collaboration with, and primarily for the benefit of, large firms, not instead of them. Startups are construed as engines of innovation and talent creation in incumbent-led open innovation systems, where it is big businesses that are key investors and acquirers. Thus, the size of the shift in Japan and Korea is from 1s to 3s, rather than 1s to 5s. This augurs for the pursuit of an open innovation variety of startup capitalism rather than convergence on creative destruction logic.
In fact, even in cases where startups are scored as essential agents for innovating in new technologies (e.g., Taiwan’s score of 3.5 for startup centrality), none of the cases strongly position startups as engines of creative destruction. Startups are portrayed as agents capable of spurring economic growth through the enhancing of capabilities for incumbents or as essential stewards of emerging technologies. Initiatives incentivize the widening of entrepreneurial pools who are pursuing emerging technologies. These startups complement, rather than challenge, big businesses. They are not creative destruction agents in any case.
Given that startup promotion is not necessarily striving to transform the system, we conclude that startup capitalism can be best understood as a form of institutional coevolution (Breznitz 2007; Edmondson et al. 2019). One reason for coevolution is that the organizations that were once responsible for developmental state policies are often the same organizations leading startup initiatives. In Japan, METI plays a central role. In China, the State Council, the NDRC, and MoST lead entrepreneurship and startup promotion initiatives. In Taiwan, what was the CEPD—then the MoST, and then, since summer 2022, the National Science and Technology Council—is a leader in orchestrating startup advancements in emerging industries. Pauline Debanes (2017) calls this “institutional layering” in Korea, as organizations such as the MSS have been merely changed in name, toward startups and the technological frontier rather than shepherding the country’s less-productive small firms. The ministry became the MSS in 2017, which served as an explicit indicator that the movement toward viewing small firms as innovators that had been underway since the EAFC was officially enshrined in the governance structure (Klingler-Vidra and Pacheco Pardo 2019). But these organizations are not new; they are new iterations of those that helped propel growth in the past and from which they evolved.
Across the four countries, there has been a shift outward in the context of the nature of innovation, when comparing figures 6.1 and 6.2, from incremental (score of 0) to radical (score of 5) innovation. Startups are increasingly envisaged as driving radical innovation rather than participating in the catch-up, incremental manufacturing that was the focus in the antecedent period. In Japan, Korea, and Taiwan, shifts were already well underway by the 1990s; the first developer in this context, Japan, was pushing (though not always successful in reaching) for the technological frontier by at least the 1980s, with personal computing software, semiconductors, and even AI and machine learning (Callon 1995). In the case of Korea, this move toward the technological frontier was clearly underway by the 1990s, when both the government and the chaebol became aware that the phase of catch-up development was over. In the aftermath of the EAFC, the ambition of producing radical (open) innovation with startups as a crucial component became more open and obvious. Taiwan’s Hsinchu Park, for its part, was already on the world map of semiconductor foundries in the same decade, with TSMC being catapulted to international innovation importance shortly after its establishment in the 1980s. Startup policies’ aims in Taiwan have clearly been, since the dawn of the twenty-first century, to cultivate entrepreneurship in emerging technologies rather than to achieve incremental innovation capacity for giants in established industries. In China, the thrust of efforts shifted toward indigenous innovation capabilities at the technological frontier later, especially in the early 2000s and then even more so since Xi took office in 2012, enabling it to compete in specific world-leading technologies, again bringing it toward Mark I in this respect.
As each country moved to incorporate the promotion of startup-centric innovation at the technological frontier into its suite of industrial policies, the institutional foundations have adapted in step with one another, having an interlocking effect. The institutional complementarities of finance and labor have moved, in varying degrees, toward equity funding (VC and stock markets) and more fluidity (less lifelong employment).
To different extents, the financing of innovation had already expanded to include equity funding means, especially VC and stock market access for high-growth firms, in the antecedent period. Beginning in Taiwan in the early 1980s, governments have themselves acted as VC investors by running VC funds that invest in startups. They have also increasingly improved the regulatory environment for VC by ensuring that legal structures such as the LP are available and offering tax incentives. In this way, each country has acted as a “venture capital state” (Klingler-Vidra 2018), seeing early-stage equity financing as essential to startups’ ability to innovate. Stock markets have been launched in each country that cater to high-tech startups that do not have many years of profitability under their belt, which is typically required to list on a stock exchange. In Japan, JASDAQ was launched (with the backing of SoftBank); the Emerging Stock Board was created within the Taipei Exchange in 2002; KOSDAQ was established in 1996 and KONEX in 2013 in Korea; and ChiNext, the Shanghai Stock Exchange’s Sci-tech Innovation Board, or STAR market, and the Beijing Stock Exchange were launched in China.
The varying moves toward a greater use of equity-based financing has not necessarily come as separate from large firms, banks, or national pension funds. The rise of startup-focused equity investment activity has reflected an evolution in the comprehensiveness of the country’s financial systems, not in opposition to debt financing. In Japan and Korea, the keiretsu and chaebol, respectively, are active VC investors openly courted by the government, investing along with, and in addition to, government programs and national pension funds. In Taiwan, the government leveraged financing from big businesses, like Acer and Formosa Plastics, to initially seed its VC industry in the 1980s. In China, the remarkable rise of equity financing—by some measures, China is the world’s largest VC market—reflects an increasingly enabling regulatory environment and mobilization of state investment for early-stage equity investment. Rather than a rejection of debt financing, then, the embrace of equity investing has been ushered in by stalwarts of the antecedent era, including large firms, main banks, and state funds. In this sense, financing models have not seen a break with the past but instead added new tools to preexisting ones.
Employment has also, broadly speaking, become more fluid. Lifetime employment has been eroded in each case. In Taiwan, policies have long aimed to widen entrepreneurial pools. From the antecedent period, the thrust of employment-related startup policies in Taiwan has been on ensuring skills for what has been deemed the technological frontier or an emerging industry. With the onset of crises and recessions in Japan and Korea in the 1990s came a challenge to the ubiquity of the permanent employment system; companies faced such dire situations that they were forced to conduct layoffs (Pempel 1998; Vogel 2006, 2018) and reduce numbers through natural attrition. Crucially for labor market fluidity, pension fund regulations were relaxed, so employees did not have to stay with their company for the duration of their career. Over time, secondment programs and the normalization of midcareer moves contributed to a modest increase in the fluidity of the Japanese and Korean labor markets. In China, the most notable change in the employment arena toward fluidity is the way in which government efforts, particularly the hukuo system and talent incentives, began to target entrepreneurs. Entrepreneurship has—from 1998—counted as a form of employment that could help secure access to social services. It has been accepted in party circles, and high-tech startup activity has been promoted among graduates and returnees.
Yet, labor market fluidity, in practice, continues to fall short of the Mark I ideal. There are numerous drivers of this. For one, practices have helped enable flexibility within the existing system. For instance, in Japan, secondments have enabled permanent employees to gain startup experience without leaving their job for life. In Korea, while startup experience is more socially acceptable, the path of attending a SKY university and then working for a chaebol remains the most sought after. In Taiwan, startup policies encourage widening pools of entrepreneurial activity in emerging technologies. China’s intense “007” technology work culture, which refers to working midnight to midnight seven days a week and is the newer and even more intense version of the “996” pattern (working 9:00 a.m. to 9 p.m. six days a week), does see movement across firms as well as mavericks who leave to set up their own startups.
Social purpose does not necessarily align with external or domestic drivers, according to our articulation of the Mark I and II types. Some scholars do expect that “creative insecurity” (Taylor 2016) instigates innovation prowess, so a relative focus on external concerns could motivate more investment in radical innovation. As we saw with the four cases, the need for innovation—even radical innovation—does not necessarily align with a focus on either startups or large firms. As a case in point, external determinants have been essential to Taiwan’s pursuit of its Mark I version of startup capitalism for decades, while China’s investment in critical technologies has centered in some ways on startups as resources for large firms. However, innovation in critical technologies has become a more explicit social purpose underpinning government efforts. In this sense, startups and economic statecraft have converged. Economic statecraft is not only the domain of large military contractors. Interestingly, given the region’s geopolitical context, China and Taiwan are similar in the emphasis on techno-national security, whereas Japan and Korea show more of a hybrid aim of delivering on domestic and external arenas.
More often, there has been a mix of external and domestic social purpose motivating efforts that have varied over time. Here Japan offers a clear example. The Japanese government has aimed, through startup promotion, to generate high-quality employment opportunities, first motivated by job losses suffered in the 1990s. Entrepreneurial experience has also been regarded as a means of training talent for the keiretsu. Secondly, Japanese startup assistance seeks to enhance the technological capabilities of large firms as a means of propelling economic growth and competing in key world technology niches. Collectively, these efforts are aimed at advancing Japan’s relative technological security (external) and fostering job creation and economic growth (domestic). Another domestic motivation has come to motivate startup promotion across the cases. Startup initiatives have increasingly been positioned as means of improving social inclusion, especially in terms of gender participation rates in Japan, Korea, and Taiwan and in the context of regional inclusion in the Chinese case.
Thus, across all four cases, a domestic source of social purpose has emerged: a means of fostering social inclusion. In China, there has been a strong thrust for decentralization, with the promotion of mass innovation and entrepreneurship aiming to better distribute high value–added activity across China, especially to small towns and villages away from the glittering lights of Beijing, Shanghai, or Shenzhen. Through the government’s bankrolling of infrastructure alongside the e-commerce platforms of Taobao and Pinduoduo, local “makers” are encouraged to become digital entrepreneurs (Li 2017). However, there remain debates about the extent to which such efforts have reduced inequality, as China’s coastal, urban hubs continue to dominate (see Klingler-Vidra et al. 2022). In the other cases, especially Japan, employment in technological startups is a solution to demographic challenges (aging population, gender-based exclusion, and regional inequalities) as well as social exclusion concerns. Korea’s startup policies speak of addressing gender inequalities, with the aim of increasing the participation of female entrepreneurs. In the Taiwanese context, dynamic startups have long been construed as important employers; this has continued to be the case, with greater efforts to drive social inclusion in terms of who is employed by, or able to establish, these firms, especially in gender terms.
The Emperor’s New (Startup) Clothes
In this book, we contribute to the “developmental state: dead or alive” debate. We do so by showing that even in empirical arenas that suggest a decline in state direction, the state (the emperor) has not retired but instead has adapted the means of intervening (the new clothes in the form of startup capitalism). The fact that there are outward movements toward Schumpeter’s Mark I type, which aligns in many ways with a stylized version of Silicon Valley, does not signify a death of the developmental state. Instead, we argue that the developmental state is very much alive. It has adopted equity financing tools, enabled more fluid labor markets, and worked to compete at the world’s technological frontier by embracing startups as engines of innovation capabilities. In other words, the developmental state has adapted to the realities of today’s innovation paradigm.
As a result, we argue that an assessment of the persistence (or not) of the developmental state’s focus on large firm is too narrow of an analytical lens. In a similar way, studies of the entrepreneurial state are too narrow, as they miss the role of large firms, and see startup help as something distinct or even at odds with long-standing industrial policy. As evidenced here, these two modes of analysis—one focused on state–large firm relations and the other on state-entrepreneur interactions—are both incomplete. Thus, we contend that political economy scholars need to assess interactions between the state and the wider innovation system, one that is more of an open innovation system composed of startups as well as national champions, universities, research institutes, and multinationals, to understand the contemporary innovation mode.
We contend that East Asia’s startup capitalism constitutes modern forms of their respective developmental state orientations rather than a decline or death. For some, they have continued along a variety of Mark I logic in which startup capitalism boosts high-growth entrepreneurs in emerging industries (Taiwan); for others, oligopolistic competition remains the organizing logic (Japan and Korea), bringing them closer to a Mark II type. We are not alone in our finding that startup capitalism constitutes continuity and change. Debanes (2017) refers to this as institutional layering in Korea, and Ulrike Schaede (2020, ix) depicts this as Japan’s “new corporate culture that foster[s] coexistence of mature and new businesses.” China’s developmental state operates according to some Mark I logics, as it actively encourages startups and breaks the centrality of some large firms (most famously, Alibaba) while also funding startups alongside large firms competing in even critical technology markets, which aligns more with a Mark II approach.
We reveal that through startup capitalism, developmental states pursuing an open innovation variety of a Mark II approach have found a way to continue backing national champions. The nature of the assistance is less direct and more nuanced than in the classic developmental state era. Rather than organizing consortia and endowing preferential credit access, which is direct and highly visible, governments can invest in and give tax breaks to startups that then partner with, or are acquired by, large firms. This tack helps limit political contestation over state–large firm relations by framing startup efforts as open innovation initiatives (Hsieh 2018; Klingler-Vidra and Pacheco Pardo 2022). Korea’s CCEIs make for one clear example. Each of the country’s nineteen centers have a chaebol partner that helps finance the center, which appears like the large firm’s contribution to society. In reality, the CCEIs are aligned with the strategic direction of the chaebol; they fit with its business lines and help it to source solutions to “what keeps it up late at night.”
The CCEIs embed the chaebol at the heart of the entrepreneurial ecosystem, which endows the large firms with access to entrepreneurial talent and ideas. By the chaebol advancing their sector-specific positions in Korea’s open innovation system, they can showcase their innovativeness to their workforce and also to their customers and suppliers (Pacheco Pardo and Klingler-Vidra 2019). Similarly, in Japan’s J-Startup Initiative, select main banks and keiretsu work with METI bureaucrats to choose the startups that can participate in the program. In the process of judging and then collaborating with METI in the running of the program, the large firms have insight into new ideas, talent, and prospective customers. The keiretsu, as a result, align their brand with startup-fueled innovation. Startup policy, in both cases, is designed in collaboration with—and for the benefit of—large firms.
Startup Capitalism’s Global Implications
The story goes that the phrase “what’s good for GM is good for America” was uttered by the president of General Motors, Charles Wilson, when he was answering questions about his ability to hold company stock while serving as the US secretary of defense in 1953.2 What Wilson meant was that big businesses like GM were essential employers, taxpayers, and contractors for national security aims. For Wilson, there was a mutually beneficial relationship between big business, the US government, and society that propelled his company’s alignment with the American people. This story speaks to long-standing relations between big business and the state in even the quintessential LME and the home to Silicon Valley.
This raises questions: Is our observation that startups can be construed as resources for large firms in East Asia specific to the region? Why could the interests of large firms not feature in startup capitalism in other countries, even in the United States? We return to that crystalizing moment for us in the early days of writing this book, when the manager of the CCEI in Seoul explained that a key aim was to inject innovative DNA into the country’s largest employers, producers, and taxpayers. It makes sense, from that lens, that startup capitalism does not necessarily manifest as separate from big businesses, let alone have the aim of disrupting them.
At the same time, the societal position of big business is now more contentious. Unlike the description of the alignment of GM and American interests, big businesses are criticized for their dominance of markets and exploitation of consumers. They are more likely to be described as strictly operating according to self-interest than as being motivated by what is good for America. Congressional hearings comprise business leaders like Mark Zuckerberg testifying to their good intentions in the face of antitrust cases and scandals. Playing on this sentiment, Jonathan Tepper (2023, 15) asserts in The Myth of Capitalism that “what is good, right, and logical for the corporation is not good, right, or logical for the economy as a whole.” The critique being that the US economy is increasingly concentrated in the hands of oligopolies that build “moats” that protect their competitive advantage.
Thus, in contrast to popular depictions of the United States as a—the—neoliberal, free-for-all market economy and the home to the world’s leading startup cluster, Silicon Valley, which policy makers across the world seek to replicate in their own home countries, there is a growing chorus of critiques. US markets, some critics posit, are less competitive, with sector after sector being dominated by a handful of companies protecting their oligopolistic, and even monopolistic, practices. “Not only are the big companies gobbling up the small,” according to Tepper, the United States has “not seen a new wave of startups coming in to compete with the Goliaths” (Tepper 2023, 11). This sentiment about worrying levels of industrial concentration has been echoed in popular media, too. For instance, the Economist found that in the 900 sectors it tracks, “the number where the four biggest firms have a market share above two-thirds grew from 65 in 1997 to 97 by 2017” (Economist 2023d).
Why is this happening? One explanation is that large firms are simply acting in line with prevailing business strategy logic, such as Michael Porter’s “five forces” and Warren Buffet’s sage investment advice. The economics discipline has linked this to a decline in economic productivity, even in the face of the supposed rise of Silicon Valley and its brand of startup-centric innovation (e.g., see Decker et al. 2016a).
A second reason is that the nature of technology today may lend itself more to monopolistic positions. Silicon Valley platform economy firms like Amazon and eBay achieve dominance over markets, and there are little means for the public—and even the state—to hold them accountable (Culpepper and Thelen 2020; Moore and Tambini 2021). Antitrust laws and calls for decentralizing the internet giants have proven difficult to action, as these companies continue to accumulate massive amounts of data, expand their physical infrastructure, and enter new market domains (Lehdonvirta 2022). The US government has, however, made some initial advances. Notably, the Federal Trade Commission is suing—along with seventeen state governments—Amazon for an illegal protection of its monopoly in online retail (McCabe 2023).
A third driver of the increased market concentration and declining business dynamism may be the widespread embrace of open innovation. As evidenced in our examination, even public policy makers tasked with promoting startups are often working to connect startups as customers of, and resources for, large firms. Startups are linked with large companies and, as such, do not instigate creative destruction’s productive forces. Rather than viewing them as competitors to be feared, big businesses see startups as customers, as outsourced R&D, and as marketing partners.
The aim of integrating startups as customers is neatly exemplified by the case of the Amazon Web Services accelerator. Early cohorts of their accelerator received coaching, mentorship, and a monetary investment. By the early 2020s, startups selected for the AWS Startup Loft Accelerator received AWS credits rather than money. The accelerator program, in this way, helped establish AWS as core to startups’ tech stack. The head of Microsoft’s corporate VC unit spoke of a similar shift toward a more strategic, rather than financial, orientation, as she “transformed the CVC’s startup investments so that they more closely align with the parent company’s business” (Rivera 2023). These approaches fit Linus Dahlander and David M. Gann’s (2010) depiction of open innovation as potentially taking inbound and outbound orientations, with startups being capable of constituting sales channels and R&D inputs. This reality—that startup-incumbent engagement in the form of accelerators is a potential sales boost for big businesses—raises important questions about whether this form of open innovation is subduing, rather than fueling, business dynamism.
Simultaneous to the growing dominance of oligopolist firms in open innovation systems, a culture of celebrating disruptive entrepreneurs has become ubiquitous. Successful startup founders, like Jack Ma or Steve Jobs, are idolized. University graduates aim to start their own world-changing business rather than work for a large company. The connotation of startups is that of exciting, nimble innovators intent on displacing the unloved, pedestrian, and sometimes even mocked big businesses. The cultural salience of startups is enshrined in hit shows like Silicon Valley, the proliferation of documentaries, and, in some prominent cases, Hollywood movies on the lives of tech founders.
Researchers have begun to lament the entrepreneurial obsession. Some have noted the problem of conflating gazelles and unicorns with the wider population of entrepreneurs (Aldrich and Ruef 2018). Books like Big Is Beautiful (Atkinson and Lind 2019) contend that large firms do bring societal value as employers and innovators, and as a result, government efforts should adopt innovation-centric and size-neutral policies rather than favoring firms based on their small size. Perhaps unsurprisingly, technology industry leaders have bemoaned the hollowing out of large-scale, cutting-edge manufacturing in the US. Already in 2010, Andy Grove, the former head of Intel, made the case for American policy that would incentivize high-end manufacturing in the semiconductor industry, and in so doing, create quality jobs and bolster US prowess in end-to-end chip capabilities (Grove 2010).
The critique of prioritizing entrepreneurship has been boosted by studies that question the relationship between startups and economic performance indicators. For instance, one analysis (Fairlie et al. 2023) of US Census Bureau data revealed that job creation and survival rates of startups are much lower than otherwise reported by the government. This adds to the call to arms around policy makers’ need to clearly define and distinguish different types of entrepreneurship and the policies required to aid their growth (Acs et al. 2016). The implication here is that there is a growing need for public policy makers to reconsider how to best encourage innovation.
Bringing these different strands together, we distill two simultaneous but paradoxical observations. The first is the reality of large firms dominating the economy. Incumbents have a greater share of markets and the ability to shape the information and products available to society, now more than ever. Business dynamism is down, in part because of the inability of high-growth firms to realize their potential due to the entrenched position of large firms serving as barriers to scale in numerous markets (Decker et al. 2016b). The second observation is that the myth of this being the era of the audacious entrepreneur persists. Within this guise, governments publicize their startup promotion efforts, like Start-Up Chile and Japan’s J-Startup Initiative, showcasing their intentions of building Silicon Valley–style clusters. Now more than ever, governments around the world communicate their desire to bolster aspiring entrepreneurs and their startups. Large companies often do the same by organizing a range of corporate venture programs that excite their employees and show external stakeholders that they are innovative.
What, then, can be said about the aims of the ubiquitous international efforts to foster local Silicon Valleys? On the surface, the ambition of startup capitalism is to build cohorts of high-growth startups that will create high-quality employment, launch innovative products, and help drive economic performance and national security. But is the goal to encourage disruptive startups that challenge large companies or instead to service the competitive positioning of the domestic economy’s Goliaths? We show that startup capitalism should not be viewed as separate from the treatment of large firms, from San Jose, to Seoul, to Tokyo. We call for a critical discussion around what really is the aim of public policies with Silicon monikers around the world. The point of governments fostering local Silicon Valleys is especially poignant, as Silicon Valley itself is under attack for being incumbent dominated.
We reveal that startup capitalism is not necessarily about creative destruction. Schumpeter’s (1934) Mark I understanding of the industrial dynamics that best foster innovation and economic growth emphasizes the value of the existential threat that new entrants bring. The conventional understanding of the Mark I pattern of innovation hinges on the notion of creative destruction in which new entrants displace the position of large firms and their technologies. Even the competitive threat of new entrants is enough to spur innovation from existing firms, for fear of being replaced by more nimble and hungrier startups.
However, we find that in startup capitalism, this rationale does not necessarily underpin startup policies, which instead strive to widen entrepreneurial pools in emerging technologies. What is more, we find that policies often take a Mark II tack, as they—sometimes explicitly but more often implicitly—aim to benefit big businesses rather than challenge them. The underlying logic is that large firms, especially those that are important to society’s job creation, economic output, and national security, such as Japan’s keiretsu and Korea’s chaebol, but also many large American firms that are too big to fail, should benefit from startups. Rather than fear startups that could displace them, leaders of large firms, such as those represented by Japan’s powerful and decades-old business group, the Nippon Keidanren, lobby the government for more investment in startups. As an example, in July 2023, Tomoko Namba, the vice chairperson of the Keidanren, told the Japan Times that “it is crucial to produce more globally successful startups” (Nagata 2023). Startups, the Keidanren chair contends, will not compete with them directly, as they will instead offer them new technologies to integrate into their business operations, products, and services. As such, even the Keidanren were “urging the government to boost support” to startups, stressing “the importance of Japan attracting foreign talent and major overseas venture capital to reinforce global networks” (Nagata 2023).
In Schumpeter’s work, incumbents aim to preclude new entrants, seeking to instead deepen their position. This can be understood as the human self-preservation instinct applied to the firm level, with large firms seeking their own continuation in business. We find that industry giants—and policy makers—today engage startups as external (technology-centric) resources or customers that benefit large firms. Thus, rather than preventing startups from coming into existence or scaling up in the name of open innovation policies, big businesses collaborate with their local governments to help them access new ideas and talent from startups. Policy makers often pursue startup capitalism as a hybrid of Schumpeterian thinking, as they strive to create new cohorts of unicorns while also ensuring the competitive positioning of large firms.
Creative Destruction or Oligopolistic Competition?
Schumpeter was ultimately interested in understanding which industrial dynamic was optimal, given the attributes of the historical setting, such as the maturity of the industry and technology. At the dawn of the twentieth century, he contended that it had been dynamic new entrants that incentivized or directly caused innovation advances. But a few decades later, he opined that large, well-capitalized companies like US Steel were able to take big bets that their smaller competitors could not consider. The human resources and deep pockets of oligopolies meant that they were capable of driving society’s transformative innovation. The two modes of innovation operated at different historic moments, according to distinct logics.
In this book, we contend that startup capitalism is underpinned by the widespread acceptance of the following three presumptions: (1) startups are essential motors of economic growth and innovation; (2) large firms, which are important actors in open innovation systems, bring value to fledgling startups, and incumbents likewise benefit from startups’ ideas and talent; (3) markets are competitive, with high-growth startups capable of disrupting the position of big businesses.
In open innovation systems in which large firms deepen their position by leveraging external resources, incumbents do not fear startups; rather, they seek out collaboration with startups to help boost their competitive position. As Baslandze (2023) argues, this orientation toward seeking out collaborations with new entrants makes sense at the level of the individual big firm. Large companies want to work with startups for inbound and outbound activities (Dahlander and Gann 2010), to know what challenges may loom, to access new product ideas that may escape them, to sell to new markets, and to find new talent. The complementarity, rather than contention, of startups from the eyes of incumbent firms aligns with Henry Chesbrough’s expectation that the proliferation of the knowledge economy and ICT fundamentally changed the nature of innovation; digital prowess can be perceived as an exogenous resource to be layered in rather than an existential threat to an established firm (see Chesbrough and Bogers 2014, 16). However, more cynically, it could be that large firms pursue open innovation strategies in a cloaked attempt to suppress threats to their business by acquiring teams and developing startup ideas as their own. Regardless of the motivation, we note that it is rational for the individual incumbent firm to lobby the government to facilitate their engagement of startups (Bombardini et al. 2023), repurposing startup assistance in a way that benefits, rather than threatens, their business. For big companies, open innovation offers multidimensional access to startups in ways that can engender innovation and simultaneously undermine their direct competition. Their motivation to shape startup help is clear.
It remains to be seen whether startup capitalism is optimal for innovation in the twenty-first century. There are various approaches taken, depending on the novelty of the technology and its accompanying industry. Policy makers strive to widen the entrepreneurial pool (closer to Mark I), and some aim to deepen the innovative capabilities of the large firms (closer to Mark II). Startups may be invoked as essential engines in areas of high uncertainty, when radical innovation is likely; oligopolies take the fore in more mature settings, and thus innovation advances are more incremental. Kai-fu Lee’s (2018) assessment of who will win the AI competition—the United States or China—invokes this logic, as he contends that AI is a mature technology, given that its major breakthroughs occurred in the mid-twentieth century. He asserts that large datasets will now dominate, not nimble entrepreneurs writing genius code. The implication is that the government should encourage scale and therefore an innovation pattern more aligned with an open innovation orientation.
Pure Mark I approaches that strive for creative destruction could challenge or alienate some companies, which would be met with pushback from business leaders—potentially the same leaders who offer crucial support for politicians and political parties. Japan’s powerful business group now lobbies for more state bankrolling for startups under the veil of open innovation aims in which startups are not threatening, but it would campaign strongly against startup policies that sought to challenge its position. This does not mean that it is politically impossible.
The Taiwan chapter opened with a public debate that unfolded for years between the TSMC founder Morris Chang and Taiwanese President Tsai Ing-wen. TSMC is central to the Taiwanese economy, but the state’s startup help is dedicated to a hybrid version centered on widening entrepreneurial pools in emerging technologies, not on fostering open innovation to benefit TSMC and its semiconductor dominance. High-profile initiatives, such as the 5+2 innovative industries strategy or Asia’s Silicon Valley Plan, continue to focus on widening entrepreneurial pools around emerging technologies. Morris Chang (2017) has been outspoken about his frustration about the lack of aid. The Tsai administration did not retract its position, but future governments may. This means there is precedent for policy makers pursuing efforts that is primarily in service of startups. However, this is not creative destruction per se.
Our analysis revealed that one of the motivations is that startup support is politically cleaner than directly underwriting long-established firms that may be perceived as cronies. This is clear in the Korean case, as successive governments work to distance themselves from headlines about their cozy relations with big business. More broadly, assistance for entrepreneurs is more acceptable to the public, as big business has come to have negative connotations, whereas the public roots for individual entrepreneurs and their growing firms. The techlash—US society’s strong negative reaction to the growing power of Silicon Valley’s large tech firms (Wladawsky-Berger 2020) and now the Amazon antitrust case—serves as evidence of this sentiment opposing more help for large firms. Unlike the GM refrain, what is good for Amazon is not widely thought to be good for America.
The Social Purpose of Startup Capitalism
We conclude by returning to the big so what? Schumpeter was interested in. He asked: What is the pattern of innovation that is best for society? He posited that either oligopolies or new entrants could best propel economic growth and innovation outcomes. But what if we cast a wider understanding of the aim—or what we have called here social purpose—of public largesse to encourage startups? What if the social purpose of startup capitalism had to do with the (re)distribution of benefits, decreasing inequality, and incentivizing technological innovations that benefit the environment and underrepresented members of society? To what extent does the pursuit of startup-led innovation to compete at the world’s technological frontier align with, or contrast with, stakeholder capitalism, which emphasizes a distribution of gains across people and planet (Schwab 2021)?
E. F. Schumacher (1973), a leader of the Appropriate Technologies movement in the 1970s, argued that Small Is Beautiful. This in some ways counters the emphasis on innovation at the technological frontier and in pursuit of economic growth. Schumacher argued that we should organize innovation capabilities in a way that prioritizes our communities and technologies to solve local challenges rather than to compete at the world’s technological frontier. The contemporary lexicon for this idea is inclusive innovation, which acknowledges that innovation has a direction and that efforts should consider the environmental and societal aims as well as who participates in and benefits from innovation (Klingler-Vidra et al. 2022).
In the social purposes analyzed across the cases, we saw job creation as central in the antecedent period and even more strongly invoked in startup capitalism. Employment opportunities, rather than competing at the technological frontier, are often named in public communications of startup policies. Our analysis of social purpose also reveals the rise of inclusion and redistribution aims especially focused on women and rural populations. Distribution considerations, in spatial terms, were most prominent in the Chinese context, acknowledging that the participants—and beneficiaries—of startup-fueled innovation are often urban populations. So, efforts strive to encourage greater regional distribution of opportunities. Women’s participation in startup clusters has adorned initiatives in Japan, Korea, and Taiwan, with the aim being to increase the number of female founders as well as the level of equity financing available for their startups.
There has, however, been less evidence of explicit links to startups as a mechanism for addressing environmental degradation. This is interesting, as elsewhere, evidence suggests the state—including in East Asia—has taken a central role in environmental action. Elizabeth Thurbon and colleagues (2023) call this “developmental environmentalism,” based on their study of China and Korea. While the state takes a strategic role in the green energy shift, our examination suggests that large firms are likely at the center of its open innovation efforts. Startups as a means of instigating environmental innovation did not feature prominently in our analysis. Perhaps policy makers instinctively see large firms are better resourced and capable of leveraging their scale to drive widespread impact. If this is the case, it explains why environmental degradation and protection efforts focus on the activities of large firms rather than incentivize startups to develop solutions.
The challenge for wider society and especially for bureaucrats is to determine the relative priority of the different social purposes underpinning startup capitalism. Which domestic priorities should motivate efforts? Are economic growth, stability, and job creation the priority? Should redistribution and environmental stewardship be the focus? Or should it be supremacy at the world’s frontier in critical technologies to advance national security that is primary? Should startups be bankrolled as a purposeful constraint on the growing dominance of platform businesses (Srnicek 2016; Thelen 2018; Kenney et al. 2020; Boyer 2022), like Amazon and Alibaba, which otherwise stand to hold massive amounts of data and control over market activities and society (see Cioffi et al. 2022)?
No single startup initiative can deliver cutting-edge innovation, national security, equitable growth, job security, and environmental benefit. Depending on the motivating social purpose, startup capitalism will take different forms. We argue that startup capitalism should be viewed as a contemporary mode of socioeconomic management well beyond the confines of high-tech startups. Startup capitalism—in its different forms, which involve large firms as well as a myriad of government bodies—is how governments are delivering economic statecraft in the twenty-first century.