Comments on Ismael Rafols et al.

In essence, Rafols et al argue that the diffuse nature of “nanotechnology” (flexibility in applications, wide distribution of innovation) means that direct regulation will be less effective than a bottom-up approach. They associate direct regulation with risk governance, and in its place advocate innovation governance. They also call this “downstream” engagement. They contrast this with not only risk appraisal approaches, but also with public deliberation approaches. By looking at the way “industrial dynamics” shapes innovation (e.g. nanotechnology’s special features as an emerging set of industries), they seek to shift attention towards a mode Arie Rip calls “de facto governance” (2009) in a dispersed industrial system that can “scope to broadly modulate innovation processes towards generic goals” (4). This “modulated” innovation governance will occur by directing policy intervention not only at R&D priorities – ‘upstream’ in the value chain, (Wilsdon and Willis 2004)- or even midstream (Fisher et al. 2006; Joly and Rip 2008), but also downstream, closer to the end-user.”

Arguing for bottom up, downstream innovation governance involves arguing against upstream engagement. By the latter they mean “A variety of activities, such as the discourse on responsible development of nanotechnology and its associated corporate codes, expert committees in the International Organization for Standards, OECD working parties, etc., have been presented as examples of the emergence of a distributed form of governance for nanomaterials (RCEP 2008).” (7) What’s wrong with this? It leaves out the huge role of higly distributed industrial firms (the yellow and reddish circles in Fig 1) and overemphasizes public actors (blue, turquoise, purple).

Mature industries can also be distributed, as in pharmaceuticals or telecommunications, but there “large firms operate as system integrators than bring together the different pieces of the value chain and in practice also heavily shape the governance” (12). In contrast, in the case of most nanomaterials, with an overwhelming multiplicity of innovation sites and actors (distributed innovation), and wide diversity of outcome options and societal functions (flexibility of application), our knowledge is too incomplete and the web of interactions too complex to assume that efforts to ‘control’ the directions of technology will succeed in driving innovation towards the desired options” (15-16).

The necessarily distributed nature of governance also follows from the non-linear form innovation takes. They offer a nice summary:

Heterogeneity of nanomaterials in end-user products means that nanotechnologies cannot be governed at the end-product level. Instead, it is in each of the branching ‘junctures’ of the value-chain, that a given nanomaterial is subject to multiple pressures and constraints that push, pull and shape its development in certain directions. These influences include the scientific and technical paradigms and routines that frame researchers’ thinking (Dosi, 1982); dedicated infrastructures that make substitution with alternatives difficult (Jacobsson and Johnson, 2000); practices that enjoy greater economies of scale and positive network externalities (Arthur, 1989); prevailing social practices (Shove, 2003); dominant policies, legal frameworks and professional association lobbying (Walker, 2000); and hegemonic discourses that inform socially acceptable performance criteria.6 Hence, at every stage in the value chain – up-, mid- and downstream, these various factors lead to specific commitments that might result in different types of lock-in (13).

Though we must seek better knowledge of the nanomaterials system via “an empirical, systematic research on the industrial dynamics of nanomaterials,” we also need to acknowledge that we need a different understanding of governance, one that fits with the fragmented, complex situation.

This is because the combined properties of distributed innovation and flexibility exacerbate the multiplicity of actors, complementary technologies and linkages involved in developing or using nanomaterials. Historical examples of other emergent technologies displaying general purpose characteristics, such as chemicals, ICTs and biotechnology, provide vivid illustrations of the sheer unpredictability of social outcomes, and challenge the assumption that collective action can purposefully shape technological directions (Williams 2006). Yet accepting the incapacity to exert control over the specific outcomes of nanomaterials, does not negate the possibility of modulating their development more broadly towards generic goals (16).

What does “innovation governance” mean in practice?

* apparently it means intervention at every stage and location where innovation occurs.¬† They don’t say this, but if “upstream” is too limited, and public engagement (a version of downstream) is also too limited, the alternative seems to be an intervention that matches the non-linearity of innovation by being everywhere.

* at the same time, innovation governance combines the “promotion” of R&D with the “regulation” of its effects rather than allowing their ongoing separation. They note astutely¬† that promotion is limited to upstream R&D (” We found no promotional policies targeted further downstream, where value-chains and markets unfold.’) Promotion needs to move downstream too.

* Downstream promotion should not demand specific outcomes, which are unpredictable, but “address generic frameworks that will have a modulating effect on the development of nanomaterials” (19). This seems to mean that “policies mid- and downstream (and sometimes upstream), need to support certain social functions such as transportation, energy provision and health,
without being necessarily specific about the technologies that can provide them.”

* The practical result is the list on pp 20-21 of a familiar set of policy-driven innovation incentives – targeted R&D, tech assessment, loan guarantees, public procurement (cf.Mowery), clearing-houses for knowledge flows in networks, consumer incentives (e.g. tax breaks), multi-level engagement .

My conclusions are as follows. It is odd that a conceptually sophisticated, learned, and promising theoretical shift winds up with examples of already pervasive public interventions of mixed effectiveness. The main outcome seems to be that because innovation is non-linearly everywhere, intervention must be everywhere – there is a real escalation of points of “goverance.” There is also an implicit traditional split between private (industry) and public governance, the latter being the place where social needs (“desirable social functions” 22) (as opposed to economic incentives?) are articulated. But I see two still-missing links: first are the systemic inequalities among parties that keeps the downstream from full participation or even learning (the mention of secrecy on 14 is important). The authors are no doubt aware of these personally, but they aren’t present in the analysis. Second are the modes in which the various elements of the network would communicate in order to articulate social needs cultural positions that are not reduced either to technological or economic pressures. Public forums, etc. are mentioned, but they have largely failed to engage anyone – scientists, companies, or the public – which is why we are still writing these articles. This piece’s persuasive description of nanomaterials’ distributed science and industrial systems – especially since it is not limited to nano-anything – should push us towards stronger ideas of co-generation – narratives, gaming, and other more explicitly egalitarian and mobile forms of contact.

I actually like the piece very much, and my impatience is more with the overall situation than with the article itself.


Back to reader.

1 Comment

Leave a Reply

Comments are closed.