In 'The Lighthouse in Economics' (Coase, 1974), Coase reached the conclusion that in England there existed a relatively efficient privately financed lighthouse system, which would refute economists' traditional statements concerning the production of public goods. The purpose of this paper is to challenge his conclusion. We first show that, from a methodological and theoretical perspective, 'The Lighthouse' is consistent with 'The Problem of Social Cost' (Coase, 1960). Then, applying Coase's own method (historical case studies), we attempt to reexamine the respective roles and efficiencies of private initiative and government.

This article aims at understanding Coase's apparently paradoxical attitude towards his eponym 'theorem'. On the one hand, he judges as excessive the attention devoted to an assertion that makes the assumption of zero transaction costs. On the other, he has never stopped reasserting its largely questioned validity. We explain this puzzle by identifying three roles of the 'Coase theorem' in his works: heuristic (to bring to light the role of transaction costs), critical (of the Pigovian tradition), and normative (Coase derives policy prescriptions).

One of Ronald Coase's insights was to extend the economic theory of choice to include the policy choice among institutional arrangements, which had to be analyzed with the same framework as the producer's choice. Both choices, he argued, are amenable to an opportunity-cost approach. The similarity he points to, however, is somewhat limited: while some of his articles from the 1930s stressed the subjectivity of producers' decisions, his later criticisms of standard policies, as well as the method he suggests for the design of policy, are based on the idea that costs are objective and measurable. Are the subjective aspects of the production decision reconcilable with the objective aspects of the policy decision in Coase's analysis? I shall argue that the framework he adopts is objectivist or subjectivist depending on the nature of the criticism he is leveling against standard theory and on the type of decision he is studying. Eventually he did propose a univocal analysis—an objectivist one—of the producer's decision between making and buying and the policy decision among institutional arrangements. This essay initiates a study of Coase's theory of decision. It returns to his subjectivist account of choice and contributes to solving the apparent contradiction between the subjectivist young Coase and the more mature objectivist scholar. It thereby sets out the diversity of the criticisms that Coase levels against standard theory and shows the evolution of his strategy. Ultimately, the problem of the difference between Coase's analyses of production decisions and policy decisions is more subtle than simply being an apparent contradiction: it turns on the subjectivity of individual decisions having no consequence for his analysis of policy.

The usual internalization of externality 'by the market' can be thought of through two different exchange modes: competitive markets, with Kenneth J. Arrow ( 1969 ); or bargaining, with Ronald H. Coase ( 1960 ). Although, in both cases, 'externality' refers to a non-exchanged effect that produces suboptimalities, these authors are working with two different, implicit conceptions of externality, rooted in different analytical worlds and calling for different institutions-parametric prices for the former but not for the latter. Moreover, while both start out with different theoretical frameworks, the authors share a concern for realism and unite when they introduce transaction costs, both advocating a policy design that calls for taking into account the costs of the different solutions. Nevertheless, this introduction of transaction costs does not itself escape consistency problems, since they do both maintain a reference to their respective ideal worlds.