1
The Venture Capital State
The road to the free market was opened and kept open by an enormous increase in continuous, centrally organized and controlled interventionism.
âKarl Polanyi, The Great Transformation
Rumors abounded as twelve North Korean government officials spent two weeks in Silicon Valley during the spring of 2011. What we know of their northern California trip is that the delegatesâsurrounded by security officers in a visit enshrouded in secrecyâspent a hundred minutes at the Googleplex in Mountain View and joined a lunch seminar at Stanford University. Their itinerary was reportedly motivated by a desire to build up North Koreaâs âown internet expertise.â The North Korean policymakers wanted to learn what the Silicon Valley cluster was all about, and what aided its ephemeral success. We could presume that they came to study Silicon Valley with an eye toward developing Pyongyang as a hub of innovation and entrepreneurship.
The North Korean delegates are perhaps the least likely in a long list of policymakers, acting on behalf of cities, provinces, and countries around the globe, to study the drivers of Silicon Valleyâs success. Policymakers visit the valley to learn how to create their own system of innovation (Nelson and Winter 1982; Lundvall 1992) and to build a bridge to Silicon Valley (Rapo and Seulamo-Vargas 2010; Saxenian 2008). As an expression of their aspirations, policymakers go on to craft Silicon monikers for their locale. The British governmentâs promotion of the Silicon Roundabout in London, Taiwanâs use of the Silicon Island, and Skolkovo being created as Russiaâs Silicon Valley are just a few examples.1
Many of these policymakers study one particular component of the Silicon Valley cluster: its venture capital (VC) industry. They learn about the clusterâs VC element for two reasons. First, VC epitomizes the neoliberal essence of Silicon Valley: private investors allocating money to high-growth, but high-risk, start-upsâin exchange for an ownership stakeâthat strive to disrupt existing markets (Lerner 2002).2 If a local Silicon Valley is to be created, then a VC marketâakin to the one that financed and nurtured the meteoric rise of start-ups in Silicon Valleyâis a necessary component of the free market fabric.
Second, venture capital is widely recognized as an engine of innovation and entrepreneurship (Gompers and Lerner 1999; OECD 1996; National Venture Capital Association 2012). In the postâglobal financial crisis world, as policymakers in advanced economies work to resuscitate economic activity, they hope that venture capitalâs âsmart moneyâ for start-ups fosters innovation and high-quality job creation. Policymakers in emerging markets see venture capital as a tool for advancing the knowledge-based economy capabilities essential to competing in the services and technology-centric segments of the global economy. Policymakers in emerging market countries hope that an escape from the âmiddle income trapâ can be attained through competition in high value-added sectors, such as those that venture capitalists invest in (Paus 2014). Venture capital promotion is conceptualized as one of the tools capable of driving advanced and developing statesâ ability to compete on the innovation frontier. Such innovation capacity brings greater productivityâand, according to mainstream economic theoryâsustainable growth and competitiveness (OECD 2007).
In light of its positive associations with innovation and entrepreneurship, venture capital holds a unique position in the minds of the public and policymakers. Unlike other investment classes, such as hedge funds and sovereign wealth funds, which are criticized for their dubious impact on firms and markets, venture capital is conceptualized as part of the financial sector that society cannot afford to live without. The Wall Street Journal went so far as to dub venture capital âHumanityâs Last Great Hopeâ for its purported ability to identify and nurture transformative technologies (Mims 2014). National Venture Capital Association research corroborates this sentiment: 40 percent of all the companies that have gone public in the United States between 1974 and 2015 received VC funding in their ascent. Those 556 companies account for 85 percent of all research and development (R&D) spending, 63 percent of the public equities market capitalization, and employ more than 3 million people.
Concurrent to the rising exubaerance for venture capital is the increased restraint on public R&D coffers. Downward pressure on spending since 2008 stems from the promulgation of austerity measures, leaving governments increasingly unable to fund basic research. Yet R&D spending is often viewed as central to national industrial strategy, particularly in the technology sector (Spencer and Brander 1983). European Union member states have struggled to achieve the 3 percent R&D spending targets set in 2002, as they spend, on average, 2 percent (Eurostat 2015). This shortfall fuels claims that there is not enough investment in innovation. Venture capitalists are put forward as the answer, as they are private investors flush with money, networks, and expertise. It is in this context that policymakersâ interest in deploying purposive action to build VC clusters along the lines of Silicon Valleyâs venture capital industryânow conceptualized to be singularly capable of advancing entrepreneurship and innovation, which is believed to be central to spurring economic growthâhas never been greater.
Puzzle
At least forty-five countries have utilized public policies aimed at creating local Silicon Valleyâlike venture capital markets. These âventure capital statesâ span a remarkable range, from Russia and Canada, Finland and the United Kingdom, and China and Chile. The diffusion of VC policy transcends ideological lines, as neoliberal and socialist states, liberal market economies and coordinated market economies, and left- and right-wing governments alike pursue policy actions.3 Efforts are not only attempted by advanced economies; policymakers in numerous emerging economies (including all of the BRICS) have also acted. Figure 1.1 below illustrates the breadth of venture capital policy diffusion, specifying the cumulative number of states initiating policies since the United States first catalyzed VC activity by providing loans to venture capital managers via the SBIC program in 1958.
North Korean policymakersâ interest in Silicon Valley, along with the VC policy diffusion trend depicted in figure 1.1, paints a picture of states converging in their efforts to replicate the Silicon Valley model. More specifically, the diffusion of the VC policy model appears to be driving economically, spatially, and culturally different states to reach âuniversal convergenceâ on yet another component of the American capitalist model (Kuczynski and Williamson 2003, 325).
Diffusion scholars expect this type of result as the temporal clustering of economic policies and government regime types comes from states having increasingly less scope for distinctive policy choices due to globalizationâs advance (Elkins and Simmons 2005; Braun and Gilardi 2006; Busch and Jörgens 2005; Holzinger and Knill 2005; Haggard and Maxfield, 1996; Goodman and Pauly 1993). In 1979, Kenneth Waltz contended that policymakers are socialized into following similar paths through emulation, praise, and ridicule mechanisms. Organizational sociologists such as Paul DiMaggio and Woody Powell (1983) assert that the preponderance of a âworld cultureâ explains the ânon-rationalâ adoption of similar policies across states. Policymakers pursue similar, highly acclaimed policies like that of trying to create a Silicon Valley as âsocial proofâ that they âbelongâ (Axelrod 1986). Particularly relevant to this study of the diffusion of a highly revered model, Florini (1996) argues that diffusion is likely when the model emanates from a âprominentâ state that is viewed as successful. Canonizing the conceptualization of mechanisms that drive diffusion processes, Simmons et al. (2006) distill the mechanisms by which convergence occurs: coercive forces, competitive pressures, learning, or imitation processes. The overarching orientation of this scholarship is the assumption that, at some critical âtipping point,â international influences âbecome more important than domestic politicsâ (Finnemore and Sikkink 1998, 902), trumping the potential for domestic context in repudiatingâor reshapingâthe internationally disseminated object.4
Figure 1.1 State launches of VC policy efforts (cumulative number) 1957â2017
Sources: European Venture Capital Association, Latin American Venture Capital Association, European Union, OECD, World Bank, and individual country sources. Methodology: Chart indicates each countryâs initial VC policy launch date, and the curve represents the cumulative number of VC policy launches. Sample is all OECD, G-20, BRICS, and Asian Tiger countries. All forty-five states launched VC policies by 2014. The forty-five states included in the sample are Argentina, Australia, Austria, Belgium, Brazil, Canada, Chile, China, Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hong Kong, Hungary, Iceland, India, Indonesia, Ireland, Israel, Italy, Japan, Korea, Luxembourg, Mexico, Netherlands, New Zealand, Norway, Poland, Portugal, Russia, Saudi Arabia, Singapore, Slovak Republic, Slovenia, South Africa, Spain, Sweden, Switzerland, Taiwan, Turkey, United Kingdom and the United States. The sample also includes the European Union as an entity.
In light of the widespread expectations for convergence, this study reveals a puzzling result: systematic diversity underlies the veneer of the convergent VC policy trend. Policies do not converge on replicating the Silicon Valley VC policy model. Policymakers do not replicate the American tax and regulatory policies to enable the private supply of investment capital into local venture capital funds. Even states of similar population and economic sizes, that are geographically proximate and at comparable levels of industrialization, do not pursue VC policies similar to one another. In fact, no two of the countries studied in this book pursued identical VC policy paths. This book explains why policymakers, who meticulously study the Silicon Valley model in an effort to build their own VC market, come to pursue distinctiveâand interventionistâVC policy formula. It reveals why they use different âmeansâ to reach the same âendââa Silicon Valleyâlike venture capital cluster.
Argument
This book argues that this resultâof varied, interventionist strategiesâis not puzzling at all. Diffusion frameworks expect convergence because they are designed to explain similarity, not variety. They emphasize the strength of the model, not the importance of local context, in explaining why policies diffuse. In so doing, this scholarship depicts policy information diffusing into contextless realms. This results in diffusionâs analytical tools only being able to explain part of an interesting phenomenon: the occurrence of interdependent, but diverse, policies (Klingler-Vidra and Schleifer 2014). Diffusion theorization is ill equipped for investigating the occurrence of variety within patterns of diffusion. It, instead, excels in explaining top-line convergence.
Diffusion researchâs quantitative bias contributes to this insufficient ability to articulate how, in actuality, the diffusion process leads to only limited degrees of convergence (Dobbin et al. 2007, 463). Research designs and methodologies consist of Large-N studies that impute explanations for the occurrence of broad but shallow patterns of convergence (Levi-Faur 2005a; Levi-Faur et al. 2011). They reveal correlationâthat states adopt policies at the same timeâwithout tracking the causal process by which the policy information diffused and, then, the local policymaking process that shaped its implementation. Diffusion studies only rarely employ qualitative case study examinations to trace causal processes (Poulsen 2014).5
For example, in her research on the diffusion of liberal market reforms, Meseguer (2009, 1) acknowledges that there were âdifferences in the timing of reforms, in their speed and intensity, and in their results,â but states that the aim of her study was ânot to explain those differences.â Research on norm internationalization has also not been oriented toward explaining âvariation in state behaviorâ; it has instead been âpuzzled by the degree of similarity or âisomorphismâ among states and societiesâ (Finnermore and Sikkink 1998, 904â5). Focused on surface trends, diffusion scholarship does not delve into the causal chain by which elements of the diffusion process contribute to the pursuit of varying policies (Weyland 2006, 14). Simply said, they donât pay âattention to diffusion itselfâ (Solingen 2012, 631). In fact, to date only a few scholars have systematically examined the role of domestic factors in driving variation in international diffusion processes (Yeo and Painter 2011; Acharya 2009; Lenschow et al. 2005).
State-of-the-art diffusion scholarship has formulated expectations about the degree of convergence based on properties of the model (the level of specificity of the model and the number of models) and the mechanisms (see Klingler-Vidra and Schleifer 2014 for a review of the literature). The existence of multiple policy models and vague policy ideas are expected to drive diverse outcomes (Falkner and Gupta 2009; Weyland 2006). In contrast, the diffusion of models that are precise blueprints are expected to result in policymakers closely mimicking the source model, and therefore the universe of adopters converging on a similar policy strategy. Competition, coercion, and emulation mechanisms are expected to propel high degrees of convergence (Simmons et al. 2008). Conventionally rational learners converge on replicating the most consistently successful policy; they undergo full Bayesian updating processes, meaning that their prior beliefs (about the consequences of a policy) are overridden such that all policymakers come to hold the same posterior beliefs (Meseguer 2009), and therefore choose the same policy. Boundedly rational learners converge on replicating the âanchorâ policy models of the countries they deem to be leaders and peers (Weyland 2006).
These analytical approaches constitute progress in the theoretical tools available to sources of variance in the diffusion process. However, they still leave much to be desired in terms of systematically accounting for whether, how much, and in what way policies vary. They do not offer analytical tools sufficient for investigating how the diffusion process results in adaptation and simply not enough analytical muscle for exploring the local contextâs role in the process.
Contextual, Not Conventional or Bounded, Rationality in Learning Processes
This book addresses this lacuna by developing a âcontextual rationalityâ approach for exploring diversity amid convergence, which elucidates how local normative contexts shape rational learning in international diffusion processes. This approachâwhich conceptualizes rationality as contextually basedâposits that learning is rooted in policymakersâ constitutive and regulative norms rather than in some acontextual environment. How policymakers see themselves (constitutive norms inform their identity) and the appropriate policies available to them (regulative norms) shapes their preferences. Norm-rooted preferences, rather than exogenously derived preferences, inform how they value policies in the diffusion process, and then, how they design policies for local use.
The premise is that political economy scholars need to focus on the local normative contexts in which diffusion occurs in order to understand patterns of policy diffusion. The contextual rationality approach builds on the work of scholars who argue that domestic contexts are not âblack boxesâ into which models diffuse (Yeo and Painter 2011). It takes seriously the contention that âstate identity fundamentally shapes state behavior, and that state identity is, in turn, shaped by the cultural-institutional context within which states actâ (Finnemore and Sikkink 1998, 902). The approach taken here presumes that cultural similarities and differences across local contexts affect the ârate and formâ of diffusion (Hall 1993; Strang and Soule 1998; Lenschow et al. 2005, 799), rather than result in a binary acceptance or rejection of the diffusion model. It holds that local environments transform the object being diffused in a similar way to what Acharya (2004, 240â41; 2009) refers to as the âlocalizationâ process in which adopters reinterpret an external model in order to increase its âfitnessâ with prevailing local norms. It extends our understanding of how the relationships between foreign information and local contexts affect the resul...