356In summary, a mix of public ﬁnancing and private provision does not preclude CIBs, but they are likely to be insufﬁciently coalesced to generate innovations in the long run. Still, permitting private provision is better than not doing so. Indeed, it has been shown that opening previously monopolized markets to private providers has led to impressive performance of high-growth ﬁrms. This suggests that there is large, untapped potential in sectors such as health care, education, and care of children and the elderly (Andersson et al., 2019). Sweden offers several illustrative examples in this respect, e.g., the voucher system for school choice introduced in the early 1990s, which paved the way for several high-growth ﬁrms in the area. At about the same time, local governments began to outsource health care, spawning several highgrowth ﬁrms, some of which have become multinational (Blix & Jordahl, 2021).

4.3 From Fearing Failure to Welcoming Experimentation

Here, Mazzucato (2018, p. 807) embraces what in Harford’s (2011) view (and ours), should guide private and public initiatives of all kinds. “Because innovation is extremely uncertain, the ability to experiment and explore is key for a successful entrepreneurial state,” she writes. “Therefore, a crucial element in organizing the state for its entrepreneurial role is absorptive capacity or institutional learning . . . Governmental agencies learn in a process of investment, discovery, and experimentation that is part of mission-oriented initiatives.” Yet, it is one thing to say that actors should experiment and learn, and another to appreciate how this is done; and to appreciate how learning differs between private actors staking their own money and public actors staking taxpayer money.

CIBs are experimental at their core, with frequent failure being inevitable and sometimes even desirable. Unsuccessful projects are not necessarily a waste of Collaborative Innovation Blocs and Mission-Oriented Innovation Policy: An. . .

357resources; failures provide actors with valuable information on a business model’s viability. This “process of learning by trial and error . . . must involve a constant disappointment of some expectations” (Hayek, 1976, p. 124). The process will be quicker and less costly if entry and expansion, as well as contraction and exit, are easy. Indeed, empirical research shows that a higher turnover of companies leads to a more competitive economy both nationally and regionally, boosting the number of high-growth ﬁrms (Brown et al., 2008; Heyman et al., 2019). Conversely, business failures can stimulate ﬁrm-founding by opening new opportunities, enabling knowledge spillovers, and making additional resources available (Hoetker & Agarwal, 2007). Indeed, more lenient bankruptcy laws are associated with higher rates of venture formation (Peng et al., 2009), to the point where “lowering barriers to failure via lenient bankruptcy laws encourages more capable—and not just more—entrepreneurs to start ﬁrms” (Eberhart et al., 2017, p. 93).

How, absent turbulence driven by markets and ultimately by what citizens qua consumers want, do mission-driven innovation agencies determine what is a failure and what is success? Innovation prizes (Sect. 4.1) may be one way to do so. Other hints may be found in what Azoulay et al. (2019) label the “ARPA model” of mission-oriented research to generate breakthrough innovations. These authors argue that successful examples of such ambitious initiatives are characterized by decentralization, active project selection, tolerance for failure, and organizational ﬂexibility. Essentially by mimicking the way markets work, we should add. While it is difﬁcult to see how actors are to have the incentives to alter plans when they lack market actors’ skin in the game, government agencies are likely to be more successful in doing so when embracing and maintaining an experimentally oriented political and bureaucratic culture that lauds experimentation.

But how? To us, at least, the current democratic and media-driven system appears highly intolerant of public sector failure, although we disagree with Mazzucato that this entirely inhibits politicians from taking risks (with someone else’s money). Politicians do take risks; however, while they are usually ready to take credit for risky projects when they succeed, they are also ready to blame a scapegoat, usually a bureaucrat, an agency, or “the market,” when projects fail. Mazzucato would likely counter with her juxtaposition between Solyndra, seen as a government failure, and Tesla, seen as a private success, even though both ﬁrms received government money.2 While that narrative exists, so do narratives blaming private actors for virtually every ﬁnancial crisis that has ever happened.

2Solyndra was a California-based manufacturer of thin ﬁlm solar cells. The company was once touted for its unusual technology, but falling silicon prices made the company unable to compete with conventional solar panels. Solyndra ﬁled for bankruptcy in 2011 and the U.S. government lost more than US$500 million based on a loan guarantee (Groom, 2014).

358N. Elert and M. Henrekson

From a Focus on Quantity of Finance to a Focus

on Quality

Whereas Mazzucato (2013, p. 40) argues against subsidies to R&D, this seems to be a matter of how. According to Mazzucato (2018, p. 808), several mission-driven institutions “have been critical to basic research,” and continue to be so today, with the rise in R&D expenditure, e.g., by NIH, being “a deliberate and targeted choice on where to direct public R&D funding.” She is certainly not alone in seeing R&D as key to innovation; indeed, this is a core assumption in much of the mainstream entrepreneurship and economics literature considering innovative activities as the result of systematic and purposeful efforts to create new knowledge by investing in R&D, followed by commercialization (Audretsch et al., 2006; Chandler, 1990). From the CIB perspective, the ancillary idea that more R&D spending is the tool that will promote innovation reveals an overly mechanical view of how the economic system works, neglecting other means of innovation, such as learning-bydoing, networking, and combinatorial insights (Braunerhjelm, 2011). Bhidé (2008) even argues that turning a high-level idea (available to anyone once produced) into a commercially viable product seldom involves much high-level R&D.

Although high R&D spending can be a necessary component of a thriving economy, it is far from sufﬁcient, and a policy of increased government R&D spending or subsidies will not necessarily result in more economically valuable knowledge (Da Rin et al., 2006). Spillovers, after all, do not need to be positive. Public R&D can crowd out private R&D, as attested to by the fact that the share of R&D in the business sector that is directly or indirectly funded by the government is lower in countries with high R&D spending by business enterprises and higher in countries with low business spending (Elert et al., 2019). Furthermore, R&D is an input in the production process; the desired output from the CIB perspective is higher value creation, which depends on many more steps along the way.

This is not to neglect the role of the state, but to nuance what the state does: A broad policy program conducive to innovative entrepreneurial venturing will likely spontaneously increase R&D spending and allocate it efﬁciently as a side effect. In contrast, if a healthy system of CIBs is not already in place, a government R&D push becomes a waste of resources, directing focus and resources toward factors that would have found better use elsewhere. It should be clear by now that the CIB perspective judges it virtually impossible for a bureaucracy to “pick the winners,” which is why spontaneous, demand-driven increases in R&D expenditures should be preferred to any top-down designed alternatives. Thus, policies and reforms should aim to mobilize and incentivize the available resources, including R&D, to ﬂow to their most productive use. This implies that R&D—and ultimately, scientiﬁc knowledge and innovation—is most effectively promoted through the pull of demand rather than the push of supply. So, what happens to the CIB when the government nevertheless opts to stimulate R&D? Both tax incentives and subsidies appear to promote this, as well as policy measures increasing the supply of skilled labor (whether through freer migration or