I. Introduction: Regulators in a Knowledge Based and Global Economy A. Renewing the 'New Economics' of Regulation Economists traditionally see regulation as a way to fix 'market failures', while it consists more broadly of building the infrastructure and foundations of markets (Joskow 2003, Brousseau and Nicita 2009, Glachant and Perez 2008, Brousseau and Glachant 2011). This interpretation encompasses the establishment of property rights, management of both negative and positive externalities (since complete systems of property rights are out of reach), management of long-term investments in common infrastructures and standards (especially because the lack of shared information on the future may lead to inadequate and under-investment). Upstream, to make choices in these matters, regulation implies the design of legitimate mechanisms to operationalize the notion of 'public/general' interest (Brousseau and Glachant 2010). Indeed the simple aggregation of individual preferences in the presence of externalities and public goods, notwithstanding uncertainty, information asymmetries and bounded rationality, does not guarantee the most efficient use or resources or even the targeting of a Ch.2-EB/ JMG-final 14.01.10 2 satisfactory social outcome. Therefore regulations imply a wide set of interrelated choices that are both highly complex and subject to strong pressures, because all individual interests have incentives to influence these essential collective choices. The 'New' economics of regulation based on incentives, as exemplified by Laffont and Tirole (1993), consists of attempting to obtain the results of competition when competition cannot be implemented. This is a way of dealing with the many 'regulatory' failures hindering the efficiency (and even the feasibility) of public attempts to correct ex-ante market failures. The above quoted contribution and others considerably renew the practices of regulatory agencies and the 'regulation philosophy' of many governments. As perfect markets, efficient regulations are out of reach. So the challenge is to design the less imperfect regulations, and compare the benefits and costs of regulatory failures with market failures in order to decide implementation, given the cost and the actual impact of regulation (Glachant and Perez 2009). However designing the least biased regulation-i.e. regulations taking into account information asymmetries and their consequences-calls for in-depth knowledge from the regulator, who should anticipate the reactions of regulated firms to the incentive schemes it designs. Moreover, assessing ex-ante the cost and benefits of feasible regulations compared to the absence of regulation, requires extensive knowledge of cost, supply, demand and customers' preferences. What is feasible, thanks to information revelation mechanisms and learning in a (relatively) stable world, becomes impossible to implement in an innovation-based economy. When the set of available products and services change, when new usage emerges, when disruptive technologies render...