Key Features of Regulation under Extreme Uncertainty
Consider first contracting under uncertainty between private parties (which as we will see in a moment closely approximates the ARPA case). Under stable conditions, each party can specify precisely what it expects in exchange with the other—do ut des.
Precision, moreover, is often unnecessary, because in stable circumstances the same parties often contract repeatedly with each other, and these relations give rise to shared norms and expectations that guide performance even when there are gaps and ambiguities in formal agreements.But under uncertainty, the very trajectory of technology is unforeseeable and solutions in any domain are often found by applying ideas that arise far afield. It is neither possible to specify obligations in advance nor to rely on shared norms as supplements or substitutes for detailed agreements. Operating at the edge of established solutions, neither party can say exactly what is feasible, let alone what the other should contribute to the joint effort. When solutions are in view, they will often involve collaboration not with familiar partners but with strangers, with norms and expectations of their own.
Under these circumstances, the nature of the contract itself changes. Instead of defining precisely each party's obligations, the agreement establishes broad goals and a regime for evaluating achievement of them. As observed in domains as diverse as biotechnology, information technology (IT), and advanced manufacturing, this regime establishes regular, joint reviews of progress toward interim targets or milestones, procedures for deciding whether and with what exact aim to proceed or not, and mechanisms for resolving disagreements. The information exchanged under such a regime allows the parties to develop a more and more precise idea of the shared goal while allowing each to assess with increasing reliability the capacities and good faith of the other: to observe if the capable stranger can become a reliable partner and the long-trusted partner is capable of innovative tasks.
As collaboration progresses, each party comes to rely increasingly on the capacities of the other, deterring opportunistic defection and generating or activating norms of reciprocity. Joint regular review and deliberate consideration of the interim results thus create the conditions in which informal norms and selfinterested calculations bind the parties to continue promising collaboration in good faith. Trust and mutual reliance are the result of agreement to collaborate, not its precondition, just as the precise aims of cooperation are the outcome, not the starting point ofjoint efforts (Gilson, Sabel, and Scott 2009).Regulation under extreme uncertainty arrives at a closely related solution from a somewhat different starting point. Under stable conditions, mitigation of externalities is mandated by legislation and given precise form in consultation between the regulator and the regulated parties (subject to judicial review in case of continuing, insistent disagreement). The costs of mitigation are known to the regulated party but not (or at least not easily) to the regulator. Addressees of regulation try to use this information asymmetry to minimize their costs of adjustment while regulators devise ways of eliciting serviceable cost information without being captured by the actors that provide it. The upshot is a fixed set of limits on permissible behavior and a schedule of fines for exceeding them.
Under uncertainty, neither the regulator nor the regulated parties have reliable information on the possibilities and costs of adjustment in the medium term, and only conjectures regarding the possibilities that will open—or not—upon further investigation. Again the response—seen in food safety, civil aviation, and pharmaceuticals, among many other industries—is the creation of an information-exchange regime that ties ongoing specification of goals (here regulatory standards) to continuing exploration of new solutions. Typically the regulator, acting as before under a legislative mandate and after extensive consolations, establishes an ambitious, open-ended outcome: for example, “good water,” as measured by minimal deviation from the pristine state of a particular type of body of water such as an alpine stream or Mediterranean river, or a dramatic reduction over an extended period in vehicular emissions from various sources.
The regulated entities—private parties, states, or member states and their subdivisions in the US or the EU—are obligated to make plans to achieve the goals and to regularly report their results. Penalties in this regime are not calculated to deter infraction of clear rules but rather to incentivize cooperative production of the information from which standards will eventually be derived. Thus penalties are imposed as a rule only for failure to report or to report honestly, or for persistent failure to achieve results whose feasibility is demonstrated by the attainments of others in like positions; though infrequent, those penalties can be dauntingly severe, often amounting to exclusion from the market or (for public addressees) severe limits on decision-making autonomy. In contract law, such information-forcing sanctions are often called penalty defaults, and we adopt that term here.The combination of ambitious and open-ended goals, planning obligations, and the threat of potentially draconian penalties for obstinately uncooperative behavior encourages investigation of new possibilities, including contextualized variants of general solutions and collaboration among regulated parties and between them as a group and the regulator. As long as some actors are looking to set new standards though their innovations—creating markets for innovative technology they develop, or simply putting competitors under pressure to match their performance—others will be less willing to cling to the status quo at the risk of being caught out when methods advance. In an environment where the development of technology is uncertain precisely because of the continually surprising abundance of opportunities it affords, the expansive search for innovation is likely to feed on itself, with inquiry generating more inquiry, if only to minimize the chances of being surprised by developments. Search is likely to be collaborative either because projects are interdisciplinary and require the combined efforts of different specialists or because any one approach, interdisciplinary or not, is likely to fail and many actors will consider it prudent to pool the risks of exploration through various forms of collaboration.
Taken together, many concurrent searches will yield a stream of surprises, unsettling the understanding of what is technically possible and raising questions about what regulation can and should reasonably require. “Notice and comment”—the one-time consultation of stakeholders required in rulemaking by regulatory agencies in the US—gives way to regular, organized exchanges as regulators and addressees seek to establish common expectation in the face of rapidly evolving knowledge. Mutual ignorance and fear of surprises further bolsters information sharing between public and private actors. By making it risky to bet on the status quo and potentially rewarding to try to surpass it, this regulatory regime turns uncertainty itself from an obstacle to demanding standards into a spur to collective learning that shows, cumulatively, how to realize them.