- 1 The mythology is embedded in a rich oral tradition, and the story of the Damascene moment of conver (...)
- 2 For a historical analysis of the cost controversy, see Aslanbeigui and Oakes (2015). For discussion (...)
- 3 For a lucid analysis of “The Problem of Social Cost,” see Medema (2009, 101-124).
1Few essays in the history of economics are embellished with a mythology surrounding their genesis. This was the fate of R.H. Coase’s celebrated article “The Problem of Social Cost” (1960), in which he undertook to eviscerate Arthur Cecil Pigou’s examination of this problem in The Economics of Welfare (1932).1 After the several episodes of the cost controversies had run their course, the most influential critic of Pigou was unquestionably Coase, who launched his attack with full force and extraordinary animus within a year of Pigou’s death. His rendition of Pigou was hardly flawless. More often than not, Pigou was damned for ideas he did not hold and positions he rejected.2 However, these lapses did not touch the premises of Coase’s critique. Irrespective of the extent to which he mangled Pigou’s writings, these premises have an intrinsic theoretical importance. In his Nobel Lecture of December 1991, “The Institutional Structure of Production,” Coase reminisced that in writing “The Problem of Social Cost,” his intentions were modest. “I thought I was exposing the weaknesses of Pigou’s analysis of the divergence between private and public social products, an analysis generally accepted by economists, and that was all.” It was only in the course of the 1960s that he began to see the broader significance of his critique for economic theory. Coase’s own judgment on his article may have been hyperbolic—he expected it to initiate “a complete change in the structure of economic theory,” or at least microeconomics, the structure of which would “ultimately” be transformed (Coase 1994c, 3, 10). However, his insight that the implications of the essay extended far beyond a critique of Pigou was certainly on the mark. If it is a bona fide classic in the literature of economics, this is due to what he called his change of approach.3 The change in question was not merely an alternative to Pigou’s treatment of externalities. The assumptions that Coase introduced in attacking Pigou sketched a theory of economic policy analysis based on his conception of the realities of the economy. He repeatedly stressed the importance of discovering and investigating economic facts, about which he supposed economists knew far too little due to their seduction by dreamlike visions of a purely “blackboard economics.” Perhaps his ultimate pejorative epithet, blackboard economics is a construct of imaginary economic worlds envisioned in the theoretical fantasies—also called ‘models’—of modern economists. “The analysis is carried out with great ingenuity but it floats in the air” (Coase 1988b, 28). An “elegant but sterile” mode of theorizing, it is a “system which lives in the minds of economists but not on earth” (Coase 1994c, 5, 13).
2A cartography of the expansive territory of Coasean economics and a close examination of its methods obviously exceed the limits of a single essay. Here we consider only certain fundamentals, a bedrock of assumptions that underpin his theory of policy analysis. Although he characterized the 1960 sketch of his position as “extremely inadequate” (Coase 1960, 18), in the various amplifications he published over the following three decades, there were no revisions of consequence. If we accept the autobiographical account of the genesis of his ideas, Coase arrived at the basic principles of his theory in 1932, subsequently finding no reason to modify them: “I was then twenty-one years of age and the sun never ceased to shine” (Coase 1994c, 8). Our objective is to interrogate the assumptions of his theory of policy analysis. In doing so, we argue that they entail self-destructive consequences that Coase did not anticipate.
3The main lines of our argument fall into eight parts (sections 2-9). In section 2, we provide a brief account of the basic premises of Coase’s theory of policy analysis: the Total Output Premise, which holds that the aim of economic policy is to maximize the productive value of the economic system as a whole; the Reciprocity Premise, which holds that externalities are symmetrical or bilateral; and the Institutional Premise, which holds that in the real world of transaction costs, total output is a function of the institutional order in which economic transactions are embedded. Coaseana is the political economy that operates on these assumptions.
4In section 3, we follow Coase’s instruction to build economic analysis close to the ground it covers by considering an externality in what he calls the real economic system. Our interest in this case—the Bayshore spillover, a conflict between the economic objectives of a chemical company and those of commercial fishermen on the New Jersey shore—lies in the question of how it might be resolved on Coasean assumptions.
5In section 4, we argue that any attempt to manage the Bayshore spillover on the grounds of the Reciprocity Premise and the Total Output Premise generates a succession of cost-benefit analyses that cannot be terminated within the limits of Coasean principles. The result is an infinite regress: the Coasean Paradox.
6In section 5, we argue that an economy operating on the Reciprocity Premise creates a condition of radical uncertainty. Instrumentally rational economic exchanges cannot be expected, because it is impossible for parties to exchanges to estimate their prospects for profit. As a result, the economy is immobilized.
7In section 6, we argue that even if a complete inventory of institutional regimes were at hand—and Coase provides no grounds to suppose that such an inventory is logically possible—the project of comparing them with a view to determining their impact on total output would be forestalled by a problem of incommensurability. Institutional regimes vary along several lines, perhaps most importantly in regime-specific methods of calculating total output and performing cost-benefit analyses. Because of these differences, what it means to conduct a policy analysis varies across regimes. One consequence of these differences: the inter-regime comparative analyses required by Coase’s Institutional Premise are not possible.
8In section 7, we argue that there are circumstances under which the Total Output Premise and the Reciprocity Premise conflict. The result is an axiomatic inconsistency in Coase’s theory of policy analysis.
9In the final two sections of the article, we define the limits of our critique by countering objections against it. In section 8, we consider an objection to our infinite regress argument, which holds that its real target is not Coaseana but cost-benefit analysis. On this objection, we attempt to prove too much because all economic policy analyses, Coasean or otherwise, employ cost-benefit analysis. Because of the universality of cost-benefit analysis in economics, it is misleading and perhaps also unfair to single out Coase for criticism on this basis. This objection does not touch the validity of the infinite regress argument. In responding, we distinguish the properties of policy analyses that are vulnerable to this argument from those that are not. In section 9, we consider the objection that our incommensurability argument leaves Coase’s project of comparative institutional analysis untouched. On this view, the total outputs of all regimes can be compared in a simple and commonplace fashion: by using a monetary measure of output. We argue that this objection confuses monetary and economic value. In examining regime-specific criteria for economic value, we close by considering the epistemological status of the Coasean economist.
10Several caveats are in order. In the ensuing, ‘rights’ refer to legal rights, and ‘transaction costs’ refer to Coasean transaction costs. Our arguments are intended to hold for complex economies, in which Coasean transaction costs often rule out management of externalities by economic actors operating in firms or markets.
- 4 On Coase’s legacy, see Landes and Lahr-Pastor (2011) and the special issue of the Journal of Law an (...)
11Coase’s theory of policy analysis does not, of course, exhaust the capacious space of Coasean economics. We do not suppose that the theory is his most important contribution to economics or that it lies at the foundation of his work. Although we do not rule out these possibilities, they have no bearing on our arguments. It would be foolish to contest the influence of his thought, which has proven to be enormous even if it has not always taken the directions he intended. One of the founders of the field of law and economics, he also spurred the development of new institutional economics—an interdisciplinary investigation of the origins, development, and operation of states, markets, and firms. Coase was less successful in transforming mainstream economics. “Most economists,” he observed, “have a very different way of looking at economic problems and do not share my conception of the nature of our subject” (Coase 1988b, 1). Here too, however, the impact of his work has been more pronounced than his own assessment would suggest. The reception of his contributions to the study of the costs of private transactions and public policy, realism in economic analysis, and the role of agency and institutions in economic transactions secure his place in the pantheon of economists of the twentieth century.4
12The subtitle of this essay, a riff on the title of Marx’s masterwork, is not a rhetorical ornament. An immanent critique, with the obvious caveat that it is sound, is the most powerful of all refutations. It attacks a position not by introducing extraneous assumptions, but by treating it with dead seriousness, at the same time establishing that it entails consequences that are contradictory, paradoxical, inconsistent with the facts it proposes to explain, or is otherwise self-defeating. Our plan is to subject the three fundamental premises of Coase’s theory of policy analysis to an immanent critique. Thus the essay is narrowly conceived—limited to three metatheoretical premises on which his theory of policy analysis rests and some of the objections to which they are exposed. Do these objections have more general implications that we have not explored? Surely they do. However, the task of spelling out these implications and assessing their validity is work for several other articles.
- 5 If there was a second step, it is not clear that Coase took it. However, he did note that it was “d (...)
13In his essay “Social Cost and Public Policy,” published ten years after “The Problem of Social Cost,” Coase maintained that the most basic problem of economic policy is “to choose that course of action which, on balance, brings the greatest good.” In economics, that good is to “maximise the value of production.” Although the question of what this choice may be is generally a difficult matter, “the nature of the choice is clear” (Coase 1970c, 35). He restated this thesis eighteen years later in his introductory essay to The Firm, the Market, and the Law. The aim of economic policy is to ensure that economic actors choose “that which brings about the best outcome for the system as a whole.” Coase assumed that “as a first step,” this outcome is “equivalent to maximizing the value of total production” (Coase 1988b, 27).5 This is the premise of total output maximization (in the ensuing, the Total Output Premise). The value of total output (in the ensuing, total output) is a function of efficiency in allocating resources, which are best distributed when they are in the hands of economic actors willing to pay most for them. In this fashion, the performance of the economy is decided with stunning simplicity and apparent clarity (Coase 1988b, 12).
14In the Coasean world, economic assets are rights over the disposition of property, and all property rights are reciprocal. Thus the substantial economic weight that Coase ascribed to the legal apparatus of the state. Given two firms A and B, A’s rights are constrained by those of B, and B’s rights by those of A. One consequence of reciprocity: uncompensated damages to third parties without their consent are routine collateral effects of economic enterprise. In Coaseana, spillovers should not be conflated with maldistribution of resources, economic injustice, or unethical business practices; nor are they prima facie grounds for policy interventions (Coase 1988b, 26). What position should economists take on spillovers? Because damages are reciprocal, if A damages B or restricts B’s rights, B also damages A and constrains its rights. It follows that in any Coasean economic transaction, there is no ‘innocent’ party. This is the Reciprocity Premise. In light of his conception of the economy, it is not surprising that Coase’s method of addressing spillovers is to resist the conventional wisdom, according to which the party that initially inflicts damage is required to repair it and right the wrong. To ask how A can be prevented from harming B or to decide on the best procedure for compensating B is a mistake. “The real question,” he claimed, is the following: “Should A be allowed to harm B or should B be allowed to harm A? The problem is to avoid the more serious harm” (Coase 1960, 2). The more serious harm is the allocation of rights that does more damage to total output.
- 6 On the importance Coase ascribed to the Institutional Premise, which reflects the “central theme” i (...)
15How should the economy be organized to avoid the more serious harm? If transaction costs are zero, this question does not arise. Spillover problems can be efficiently resolved by the actors themselves. They may negotiate a mutually satisfactory result. Or one party may decide to buy all the property rights at stake and settle spillovers in accordance with its interests—in which case, other parties agree to sell their rights at a mutually satisfactory price. In the standard transactions of a complex economy, these methods are ineffective because transaction costs are unacceptably high (Coase 1960, 17). Coase held that in the real world of commonplace transaction costs, total output is a function of the institutional order: the configuration of institutions in which economic transactions are embedded. In Coaseana, there is only one respect in which variations of institutional configurations or regimes are significant: the extent to which they diminish, restrict, increase, or maximize total output. “The basic problem of economic policy,” he stated in 1970, “is to consider how alternative social institutions would work in practice—and to evaluate their performance. That is to say, we should compare the total effects of adopting these alternative social arrangements” (Coase 1970c, 42). This doctrine, the Principle of Comparative Analysis of Institutional Configurations (in the ensuing, the Institutional Premise), is an axiom of the Coasean economy to which he repeatedly returned. “Without some knowledge of what would be achieved with alternative institutional arrangements, it is impossible to choose sensibly among them. We therefore need a theoretical system capable of analyzing the effects of changes in these arrangements” (Coase 1988b, 30).6
- 7 Two themes of Coase’s thought that have fascinated and distracted economists—the economics of zero (...)
16In the final analysis, therefore, Coasean economics is a comparative investigation of institutions, with a view to identifying the institutional regimes most effective in maximizing total output. It is a project of discovering, designing, building, and repairing these regimes. Institutional regimes qualify as alternatives because they allocate rights over economic resources by employing different criteria. For Coase, the key differences center on the question of how effectively alternative allocations maximize total output. He anticipated no general answer to this question, and thus no possibility of a grand general economic theory. Every allocation of rights must be investigated to discover whether it exhibits properties that may be decisive for regime selection and resource allocation. His methodological conclusions: the necessity for careful, factual investigation and the rejection of abstractions.7
17In the spirit of Coase’s admonition to abjure abstractions, we turn to a very small province of the global economy: the New Jersey shore, an area that has achieved near universal notoriety for desecration of the environment by relatively unfettered economic development. Raritan Bay is the large southern portion of lower New York Bay that has made New York City attractive as a port for centuries. On its southern banks are the Bayshore communities of Monmouth County, New Jersey, for many decades home to small but thriving businesses operated by fishermen who made their living from striped bass, fluke, flounder, bluefish, blue crabs, clams, and mussels that lived in and along the Bay. The Bayshore communities and the Atlantic beaches a few miles to the south are also the site of a summer tourist industry: recreational fishing, beach clubs, seafood restaurants, boardwalk amusements, and many bars that feature bands with the distinctive sounds of Jersey Shore rock music. In 1951, International Flavors and Fragrances (IFF) built a 41-acre chemical plant on a 105-acre property in Union Beach, a Bayshore community. The property bordered both the Bay and shoreline tidal wetlands as well as residential areas and New Jersey Transit rail lines. Between 1951 and 1997—when the plant closed operations—IFF processed, packaged, and stored chemicals, selling its products to food, household goods, and cosmetic industries.
18In 1985, Friends of Clearwater—a New Jersey environmental group later renamed New Jersey Friends of Clearwater—collected samples from Raritan Bay on six different dates. The samples recorded concentrations of several chemicals toxic to fish, shellfish, and humans that had spilled into the Bay through groundwater seepage. They included toluene, chloroform, methylene chloride, 1,2 dichloroethene, and benzene. On October 30, 1986, under threat of litigation, IFF agreed to pay a penalty of $1.25 million to New Jersey, reimburse the state for an investigation of the site estimated to cost $750,000, and clean up the contaminated site (see the Environmental Protection Agency n.d.; Conway 1986a, 1986b).
19The Bayshore spillover is a paradigm case of an externality. The business of IFF was the production of chemicals used in a wide range of consumer goods. One byproduct of manufacturing chemicals is toxic waste, which is why sites that border on water are advantageous to chemical companies. The methods IFF used to dispose of toxic waste reduced its production costs. However, the fishing and tourist industries depend on largely uncontaminated waters from the Bay. In Coaseana, this phenomenon is a conjunction of reciprocal damages. If IFF contaminates the Bay, tourism and fishing on the Jersey Shore are damaged. However, imposing preventative or remedial decontamination measures will damage IFF. Employing Coasean economic logic, how should this problem be managed?
- 8 Coase did not limit the economic function of the state to the definition of property rights. On his (...)
20By global standards, or even from the perspective of the economy of the United States, the economic dimensions of the Bayshore spillover are hardly formidable. Because of the complexity of the Bayshore economy, however—the number of actors involved and the range of their interests—Coasean private-sector institutional options for a solution, the firm and the market, are ruled out. If they are excluded, he left open the possibility, albeit somewhat remote, that intervention by the state—which he invariably called “the government”—might provide the most efficient allocation of rights.8 Does this hold true in the present case? Suppose that the conflict between IFF and Bayshore businesses became a case falling under the jurisdiction of a Coasean court. What result could be expected?
21It might be supposed that the following issue is essential to this question: should the Bay be understood as a right to a site for commercial fishing and tourism or a right to a chemical waste dump? On Coasean premises, this supposition is mistaken. In Coaseana all property rights are created equally, in which case distinctions among them made on grounds independent of the Total Output Premise are irrelevant. The only material issue concerns the most efficient assignment of rights to the Bay. Which allocation of rights contributes most to total output? Which enterprise contributes more—IFF or Bayshore businesses? Or perhaps both are needed to maximize total output. If that is the case, in what volume or proportion? Most important, how can these problems be resolved within the limits of Coasean premises?
22Suppose that judges begin with the intuitively reasonable assumption that maximum total output depends on both IFF chemicals and Bayshore fish. In that case, several questions, all requiring complex and expensive research of indefinite duration, arise. Which IFF chemicals, of what quality, and in what amounts does maximum total output require? What catch of Bayshore fish, in what variety and quality, and according to what timetable are needed? What levels of toxic chemical discharge into the Bay are consistent with an aquatic environment sufficiently robust to nourish the requisite catch? Alternatively, what levels of discharge of chemical waste are necessary to produce the requisite array of IFF chemicals? And how can the most efficient balance be achieved between the minimum toxic discharge needed for chemical production and the maximum permissible for the requisite supply of edible fish? Would the cost of this research diminish total output? If so, by how much? Are the cost and its impact on total output worth the price? These questions are critical, because Coaseana is governed by instrumental rationality. Firms enter transactions as an efficient means to achieve their objectives. The same holds for Coasean judges, who make decisions based on assessments of the most efficient means of maximizing total output. In all cases, a comparative judgment of costs and benefits of projected decisions—the methodological sine qua non of Coasean calculations—is required to arrive at measures best calculated to achieve the objective at stake.
- 9 Twenty years ago, A.W. Brian Simpson anticipated this argument. See his criticisms of Coase’s view (...)
23In the Bayshore spillover, the initial step is a cost-benefit analysis to determine whether the cost of the above research and its impact on total output are worth the price. This is Analysis 1, the performance of which also entails costs. The prospect of its costs raises the question of whether the analysis is worth the candle, calling for another calculation, Analysis 2, to determine the relative value of Analysis 1. The same question arises concerning Analysis 2: do its benefits outweigh its costs? This question can be answered only by means of a further cost-benefit analysis, the presumed value of which generates another cost-benefit analysis, and so on. Attempts to address the spillover on Coasean premises become enmeshed in endless questions of comparative costs and benefits, none of which can be answered without first determining the value of the answer. Because of the Total Output Premise and the Reciprocity Premise, which are responsible for the necessity of deciding on the distribution of property rights, and the exclusively instrumental rationality of Coasean economics, that determination can never be made. The result is a self-stultifying orgy of cost-benefit questions in which a judgment on the original spillover problem cannot be reached, because the fundamental epistemological deadlock cannot be broken. The result: an infinite regress that constitutes the Coasean Paradox. An escape is possible only by breaking the endless chain of cost-benefit questions and terminating the analytical regress.9
24In exploring possible state solutions to spillover effects, Coase argued that the most efficient action may be no action. If the cost of resolving a spillover exceeds the cost of the spillover itself, total output is maximized by following a course of inertia. The state does not intervene and the spillover continues. The reason: under less than ideal circumstances, this may be the best of all economic worlds. However, he did not see that the inertia option, which he regarded as only one of several possibilities, is a consequence of the Coasean Paradox, thereby eliminating all other options. Coasean economic logic operates according to its own premises, the paradoxical consequences of which force the state into a position of inertia. The spillover remains, but not for the reason Coase supposed. On the contrary, it persists because of the consequences of his analysis, which render any resolution logically impossible.
25An irony of Coasean thought is worth noting at this point. Coase was consistently skeptical of concentrating economic power in the state and discouraged routine state intervention in economic affairs. He regarded ‘government failure’ as more worrisome than ‘market failure’. Thus his conclusion: “The ubiquitous nature of ‘externalities’” and the presumed absurdity of attempting to eliminate them by means of economic policy suggest that “there is a prima facie case against intervention” (Coase 1988b, 26). Yet, as the above considerations show, in order to achieve the distribution of property rights best calculated to maximize total output, he delegated to the state an analytical project of staggering complexity. As a result, Coase, tacitly and surely unwittingly, credited the state with a level of competence in economic decision-making and endowed it with powers to manage the economy that no liberal-democratic polity has ever enjoyed.
26As judges struggle over intractable epistemological quandaries, what are the implications for Coasean total output? The most significant implication is a radical and economically destructive state of uncertainty. Suppose we begin with the plight of the Bayshore fishermen and their options. The status quo is a chemically contaminated Raritan Bay. As unpaid mortgages and business loans mount and customers disappear because of toxicity, it seems futile to persevere with commercial fishing on the Bay. Perhaps those who can afford to relocate will consider doing so. However, for instrumentally rational Coasean businesses, this option seems pointless and perhaps financially disastrous. Discovering less troubled waters entails significant search costs. Moreover, there is no reason to assume it will be possible to escape the problems of comparable spillovers elsewhere. Converting fishing boats into small sightseeing vessels is also a dangerous bet. Coasean tourists are instrumentally rational. Why would sightseeing on chemically contaminated waters appeal to them? In light of these contingencies, a decision to sell fishing boats and equipment, retrain, and enter another business might seem attractive. However, it assumes that the insecurity of rights on the Jersey Shore will not be reproduced in other industries and locales. This is a vain assumption. Spillovers are endemic to the Coasean economy, and the Coasean Paradox is not limited by geography, legal jurisdiction, or industry.
27What lesson can economic actors in Coaseana be expected to glean from the Bayshore spillover? Would instrumentally rational businesses enter a transaction on the prospect that their rights might be suspended or lost under conditions they cannot anticipate and in circumstances on which they could form no reasonable judgment? This is a rhetorical question. The implications of its answer for the viability of Coaseana seem quite destructive. The rate at which businesses are formed would drop precipitously since reliable estimates of profit and loss cannot be made. The same holds for the rate of investment. Potential borrowers would not borrow in the absence of confidence that they can repay their loans. Banks would not make loans, because they have no reason to expect repayment. In the extreme case, the economy would be frozen in paralysis.
28There is an obvious response to this picture of the torpor of a Coasean economy. Suppose that the state assigns property rights in an ad hoc fashion to achieve a semblance of confidence on the part of buyers and sellers, borrowers and lenders. On the assumption of an ad hoc policy, any allocation of rights is preferable to economic crisis. Since economic planners would not employ cost-benefit reasoning in arriving at their decisions, the state escapes the infinite regress of the Coasean Paradox.
29This tactic is ruled out by Coasean principles. Coase heaped scorn on ad hoc state regulations: they are generally ill-informed, influenced by special interests, and frequently in conflict. More often than not, they reduce total output. These are the main grounds for his position that there is a prima facie case against state intervention in the economy. Thus the Coasean state does not indulge in ad hoc economic policy. Accurate and comprehensive knowledge of the economy is essential to policy, which means that the necessity of cost-benefit analysis reappears. In his essay of 1970 on public policy and environmental degradation caused by economic enterprise, Coase stated the fundamental problem at issue here: “What we really want is the right amount of pollution—the amount, that is, which maximizes the value of production” (Coase 1970c, 43). To paraphrase: what we really want is the right level of uncertainty—the level that maximizes the value of production. How can that level be determined? Only by an economic analysis that compares costs and benefits of alternative degrees of uncertainty in order to estimate “the right amount”—the amount that maximizes total output. This analysis restarts the logic of the Coasean Paradox.
30Coaseana is an economic order of innumerable, routine, and relatively innocuous externalities that can either be resolved or tolerated without inflicting significant damage on the economy. Coase did not consider externalities of a magnitude that would reduce total output dramatically. Nor did he consider a complex, tightly coupled economic system in which spillovers in one subsystem invade other subsystems, rapidly multiply, increase in magnitude, and threaten to destroy the entire system. Such systems are at high risk for “normal accidents” that can end in disaster (Perrow 1984). Tight coupling has been a property of modern economies for more than a century. Whether by inattention, design, or optimism, Coase’s theoretical apparatus does not include tools to confront crises or impending catastrophes such as the Great Recession or irreversible environmental degradation.
31The above scenario of crisis and catastrophe is, of course, not the view Coase took on the consequences of his change of approach. Once alternative institutional configurations are investigated and compared with a view to discovering the regime that operates most efficiently, economic actors will have the epistemic framework they require to make instrumentally rational decisions. Economic analysis will either show how spillovers can be resolved or demonstrate that they are less costly than their resolution.
- 10 Coase studied the development, operation, and impact of various state monopolies, ranging from Brit (...)
32What are the bases and logic of such an investigation, the ambitions of which seem unprecedented in the empirical social sciences? Who will conduct it? Who will pay for it? How will investigators be selected? On what grounds will their conclusions be assessed? There is also a more troubling philosophical question: what are the epistemological conditions for its possibility? On these matters, Coase was silent. In 1960, he admitted that the few informal remarks he devoted to the problem of “choosing the appropriate social arrangement” for dealing with spillovers, the centerpiece of the new approach, were unsatisfactory (1960, 18). In the following three decades, the extreme laconicism of his remarks on the investigation that would initiate the forthcoming transformation of microeconomics suggests that the terms and conditions of this project are obvious and unproblematic. They are neither. Coase noted that the costs of the project would be formidable. Engaging “many scholars over many years,” it could not be expected to bear fruit “within the foreseeable future” (1964, 196; 1970b, 61). In the interim, all spillovers that could not be contained by markets or firms would continue, new spillovers would occur, and the costs of both would mount. He did not regard these considerations as of great moment. A full accounting of economic data takes precedence over timely resolution of spillovers. Indeed, timeliness in developing policy seems to be largely irrelevant: “Ours is not an enterprise in which everything has to be sacrificed for speed. After all, we are not trying to get to the moon” (Coase 1964, 196). Coase’s casual allusions to choices among “alternative social institutions” (1970c, 42) conceal the fact that a comparative analysis of institutional regimes extends far beyond the question of whether the state or the private sector is more efficient in settling spillover problems.10
33Suppose that Coase’s statements on the crucial role of comparative institutional analysis for economics are taken seriously. Suppose also that he had attempted to answer a question critical to the comparative investigation he proposed: how many different institutional regimes are there? An answer to this question is necessary in order to compare all regimes with a view to identifying those best configured to maximize total output. In surveying institutional forms that come into play in economic policy analysis, Coase considered three: the firm, the market, and the state (“the government”). However, there is no ‘firm,’ ‘market,’ or ‘state.’ Even in the West, there is an impressive range of firm, market, and state structures that may enter into an indefinite variety of combinations in forming institutional regimes. Either Coase was serious about a comparative analysis of institutions, or he was not. If he was not, his program for the transformation of microeconomics collapses. If he was, a systemic accounting of institutional regimes is essential, and it is indispensable to establish that it is complete. Otherwise the regimes that maximize total output may be overlooked. These considerations pose a dilemma. On the one hand, an exhaustive compilation of institutional regimes is necessary. On the other hand, it is impossible to ascertain whether this condition is met, because there is no criterion for determining how many different regimes there are. Thus the Coasean project of a comparative institutional analysis would appear to be blocked before it begins.
34Assume that we bracket this problem: an exhaustive conspectus of institutional regimes is not out of the question, and Coasean investigators have arrived at a complete inventory. Would the result advance a comparative analysis of institutional regimes? It would not. Different regimes are defined by differences in resource endowment, the bargaining power of economic actors, firm structures and technologies, costs of production, prices, and even methods of calculating total output and performing cost-benefit analyses. Because of these differences, comparative analysis of institutional configurations based on calculations of total output across regimes is impossible.
35Return to IFF and the fishermen of Raritan Bay. Consider two institutional configurations for resolving the conflict: a Lilliputian and a Brobdingnagian regime. The former favors small firms on the assumption that they maximize total output; the latter, oligopolies on the contrary assumption that they achieve this result. A Brobdingnagian regime is characterized by principal-agent relations, expensive and sophisticated technologies, and large-scale production. Due to their massive contributions to political campaign funds, Brobdingnagian firms are considered too big to fail, encouraging them to take extravagant risks that seem relatively costless. Lilliputian firms have none of these properties. They are labor-intensive and operate with simpler and less expensive technologies. Economies of scale and large production runs are ruled out on grounds of restricted access to financing. No firm is large enough to influence policy makers or threaten the viability of the economic system. Although the institutional structure of production in both systems produces externalities, their profiles are quite different.
36Given the above conditions, each institutional configuration will generate regime-specific criteria for measuring total output, its constituents, and their weights. In a Brobdingnagian regime, the weights assigned to IFF products are greater than those assigned to the output of Bayshore fishermen. In contrast, a Lilliputian regime assigns more weight to the ecological integrity of the Bay and the catch of its fish. Cost-benefit analyses in the two regimes do not measure the same variables, making their results incommensurable. If, however, inter-regime analyses cannot be performed, Coase’s method for analyzing externalities is invalidated, and his change of approach for economics is nullified due to self-destructive logical flaws.
37Any system grounded in more than one axiom is vulnerable to axiomatic inconsistency: the possibility that fundamental premises may conflict. This possibility can be an occasion for daunting and systemically destructive questions. In the present case, the following problem is evident: what are the grounds for supposing that the Reciprocity and Institutional Premises are mutually consistent? Put another way, what were Coase’s reasons for assuming that the most efficient institutional regimes—the social arrangements that succeed above all others in maximizing total output—rest on the Reciprocity Premise? This is not a conclusion he established, nor did he offer evidence for it. His assumption that reciprocity regimes maximize total output begs the question against all regimes that violate the Reciprocity Premise, suggesting that a comparative analysis of institutional regimes is unnecessary. After all, it seems that he had already decided which set of institutional configurations is preferable.
38Coase did not consider arguments that a reciprocity regime may become extraordinarily inefficient due to failures in regime legitimacy. Suppose that the economic actors of Coaseana take the view that the Reciprocity Premise is intuitively preposterous and ethically reprehensible. They believe it is tantamount to the position that victims of fraud are co-responsible for the machinations of fraudsters, that victims of rape bear responsibility for their fate, and that victims of murder share the guilt of murderers. The question of whether their beliefs are sound is irrelevant. The only matter of importance is the strength of their convictions and the extent to which they are prepared to act on them. If they are persuaded of the irrationality and injustice of a reciprocity regime, productivity and efficiency can be expected to drop, reducing total output. Collective action against the regime based on public perceptions of its illegitimacy will further diminish efficiency and total output. There are numerous instances—occasionally of considerable historical consequence—in which collective action is translated into collective violence (Tilly and Tilly 1981; Tilly 2003). If collective action overturns an institutional order based on the Reciprocity Premise, total output may be reduced to a level at which any attempt to measure it would be senseless—assuming that the agencies needed to arrive at this conclusion were still in place. At the peak of the Terror during the French Revolution, radical Jacobins declared that the Republic had no need of savants. Suppose that the citizens of Coaseana decided—on what motives or grounds is immaterial—that the economy had no need of the Reciprocity Premise and refused to work under such a regime. The consequences for total output would be shattering. In sum: there is no basis, prior to a comparative analysis of institutional configurations, for Coase’s assumption that the regime selected on the basis of the Total Output Premise will adhere to the Reciprocity Premise.
39Several readers of an earlier draft of this essay raised the objection that it is only incidentally a critique of Coaseana. Its real target is cost-benefit analysis as an economic method. On their view, we are arguing that cost-benefit analysis is methodologically useless because it ends in an infinite regress. Thus the objectives of the essay are much more ambitious than we claim and perhaps futile. Here we respond to this objection.
40What is the source of the infinite regress that constitutes the Coasean Paradox? Two Coasean principles: (1) the Total Output Premise requires cost-benefit analysis in order to estimate the most efficient means of producing total output. In Coaseana, maximizing total output is the objective of the economy. Efficiency as determined by cost-benefit analysis is the maximand. (2) The Reciprocity Premise holds that externalities are symmetrical. Because of their axiological neutrality, there is no principle that rules out certain economic transactions, mechanisms, or structures as illegitimate. Such a principle would supersede maximally efficient total output as the sole consideration in policy and forestall an infinite regress of cost-benefit analyses. As a method of economic reasoning, cost-benefit analysis is intrinsically unobjectionable. It becomes problematic by virtue of its tie to the Coasean metaphysics of political economy, which rests on the Total Output Premise and the Reciprocity Premise. As noted, it is possible to escape an infinite regress of cost-benefit analyses by a principle that overrides the Total Output Premise. However the Reciprocity Premise rules out such a principle because it would violate the symmetry of externalities.
41Some theories of policy analysis are vulnerable to self-destruction by the Coasean Paradox. Others are not. Vulnerable theories rest on the Total Output Premise and the Reciprocity Premise, two axioms of the Coasean world. Theories invulnerable to the Paradox reject at least one of these axioms. Thus there are two possibilities to consider. A theory that rejects total output but embraces reciprocity is not necessarily committed to cost-benefit analysis as a method of resolving externality problems. Thus the Coasean Paradox does not arise. A theory that rejects reciprocity but embraces total output is committed to the asymmetry of externalities. This means that some economic transactions, practices, or structures are objectionable independent of economic considerations. Even if the theory employs cost-benefit analysis in addressing problems posed by externalities, an infinite regress is forestalled or terminated by its axiological assumptions. There are various possible assumptions that can produce invulnerability and a variety of arguments on which they may be based. Both the assumptions and their grounds may be open to objections that are independent of Coasean premises. That consideration is irrelevant to the issues that are at stake here: the limits of our critique of Coaseana and the range of theories of policy analysis that is vulnerable to the Coasean Paradox.
42The first systematic theory of policy analysis in the English language was developed by Pigou in Wealth and Welfare (1912), a dense and closely reasoned book of almost 500 pages. He elaborated the theory in even greater detail eight years later in The Economics of Welfare, perhaps the last encyclopedic treatise in economics executed by a single hand and a work of staggering length—almost 1000 pages—and immense complexity. Yet its conceptual architecture and logical structure were remarkably clear, even simple.
43Pigou rejected the Total Output Premise. In the early pages of The Economics of Welfare, he stated three generalizations that he proposed to substantiate. In the literature, they are often called the Pigouvian ‘welfare axioms’ on the mistaken assumption that they have the status of bedrock presuppositions—premises that lie beyond verification rather than hypotheses that call for confirmation. All other conditions remaining the same, economic welfare is likely to increase if (1) the size of the national dividend increases, (2) the dividend is distributed more equally, raising the share allocated to the poor, and (3) fluctuations in the magnitude of aggregate output are reduced (Pigou 1920, 47-67). Although increases in total output are one of the conditions for economic welfare, they are not the summum bonum of the economy. The weight ascribed to the size of the national dividend depends on two other factors: distribution and stability. If an increase in the size of the dividend compromises either factor, Pigou’s welfare criterion may require a decision in favor of lower rates of growth or productivity. Because of Pigou’s rejection of the Total Output Premise, such a decision need not be made on the basis of cost-benefit considerations. It might, for example, be reached on the grounds of pragmatism or executory convenience, tradition, or the balance of power in the economy or the polity. Suppose, however, that it were made by employing cost-benefit analysis. What would stand in the path of the infinite regress of the Coasean Paradox? Pigou’s rejection of the Reciprocity Premise.
44In Pigou’s work, externalities are asymmetrical, because economics is grounded in ethics. In his inaugural lecture as Alfred Marshall’s successor in the chair of political economy at Cambridge, he defended the conception of economics as a “hand-maid” of ethics and argued that the study of ethics is indispensable to economists (Pigou 1908, 14). Because practical results are the raison d’être of economics—its contribution to the lives of “the suffering and degraded, who have been worsened in the industrial struggle”—economics is not an autonomous science (ibid, 12). Its normative principles, which state what should be done to improve human welfare, are derived from ethics. Given these principles, economics investigates economic causes that make improvement in welfare more probable. Pigou’s point in specifying the relationship between economics and ethics was an epistemological thesis. In order to arrive at scientific conclusions concerning economic welfare, economists require knowledge of human welfare generally, which is the objective of moral philosophy.
45What conclusions concerning the resolution of externalities did Pigou draw from these considerations? Human beings, he observed, are both ends in themselves and instruments of production. Conceived as ends, they embody the moral qualities of human life. Conceived as means, they can be used as tools to “produce things whose use yields welfare” (Pigou 1932, 12-13). There are circumstances under which the moral and the economic conceptions of human beings conflict: “efforts devoted to production of people who are good instruments may involve a failure to provide people who are good men” (ibid., 14). The ethical quality of life is damaged if increased output is achieved through excessive hours of work, dangerous working conditions, oppressive employers, or labor that is alienated, bonded, or enslaved (Pigou 1932, 12-17). Because economics is subordinated to ethics, the moral quality of life overrides its instrumental or economic utility. In his conclusion to Wealth and Welfare, Pigou stressed the point that policy analyses are “instruments for the betterment of human life.” Because economic welfare serves the more general ends of total welfare, the policy maker “needs always to be aware lest, in his ardour to promote an economic benefit, he may sacrifice unwittingly some higher and more elusive good” (Pigou 1912, 487-88).
46Conflicts over the resolution of externalities may begin with cost-benefit analyses of which uses of human beings best produce, distribute, or stabilize the national dividend. If these conflicts cannot be resolved by comparing production efficiencies and inefficiencies in the employment of persons as economic resources, economic methods of analysis give way to ethical questions. Which uses of persons are best designed to advance the moral ends of life? And what should these ends be? The result: economic policy analysis does not terminate in an infinite regress of cost-benefit estimates but in an ethical inquiry into alternative ends of life.
47The point of this survey of Pigou’s theory of policy analysis is, of course, not to make a case for the foundation of economics in ethics but to identify the logical source of the Coasean Paradox. It lies not in cost-benefit analysis but in the Coasean premises of Total Output and Reciprocity.
48Consider the following objection to our critique of Coase’s program for a comparative analysis of institutional regimes: “Yes, the constituents of total output and their weights may vary across institutional arrangements; no one denies that. However, it does not follow that total output in different regimes cannot be compared. After all, total output under all regimes can be measured in dollars, pounds, euros, bitcoins, or some regime-neutral currency. Thus it is mistaken to conclude that measures of total output under different regimes are incommensurable. Even if the premises of our incommensurability argument are admitted, its conclusion—the incommensurability of regimes and the impossibility of a comparative institutional analysis—does not follow.”
49This objection conflates monetary and economic values. Coase’s theory of policy analysis assumes that if a Brobdingnagian regime produces $5 trillion of total output, it is superior to a Lilliputian regime that produces $4 trillion. The basis for this assumption seems innocuous and self-evident: five is a greater number than four. This is obviously true, but it is also irrelevant. How can a Coasean economist establish that the economic value of $5 trillion in a Brobdingnagian regime exceeds the economic value of $4 trillion in a Lilliputian regime? Because there are no regime-neutral criteria for arriving at total output, comparative cost-benefit analyses across institutional regimes cannot be performed. In calculating total output, the considerations that decide which facts count and the weights that are ascribed to them are anchored in axiological assumptions, generally tacit and rarely articulated, concerning the impact of institutions on the quality of human life and character. Which is preferable: an economy in which the labor force consists chiefly of employees who are paid handsomely for work they regard as inherently meaningless, increasing the value of corporate stock held by members of a politically influential rentier class? Or an economy in which proprietors of small firms less well-endowed with the goods of this world pursue enterprises in which they also find extra-economic significance? Which institutional regime produces a superior quality of life: a world of large or small organizations? An economy in which consumers have easy access to fresh fish, the catch of Bayshore fishermen? Or a world in which they are liberally supplied with products flavored and scented by IFF chemicals?
50These questions do not lie within economic science, neither in the empirical nor the formal theoretical sense. Following Frank Knight, Coase claimed that the “problems of welfare economics must ultimately dissolve into a study of aesthetics and morals” (Coase 1960, 43). And, he might have added, politics and religion. However, he did not recognize that extra-economic values have economic consequences. Just as there is no free lunch, there are no free ethical, political, aesthetic, or religious principles. They all entail costs, and they may also confer benefits. This is why answers to the above questions and the reasoning underpinning them are among the determinants of the economic value of total output, giving them a crucial place in the Coasean program. Perhaps the most illuminating way to consider these issues is to pose the following question: if the economic value of total output rests on extra-economic values that vary across institutional regimes, in what regime are Coasean policy makers positioned? This question is critical because, in the final analysis, their choices among alternative institutional configurations will effectively settle the issue of how the economic value of total output in a given economy is determined. There seem to be two possibilities. (1) They may be positioned in an institutional regime. In that case, their decisions are circular. Deciding within one regime and employing its assumptions and logic, they presuppose a resolution of the issues at stake, begging the question against alternatives. (2) Or they are situated in an extra-institutional space. But such a space is no-where, the transcendent world of the Aristotelian unmoved mover, a metaphysical sphere of supra-mundane decision-makers beyond institutional orders.
51The same question can be raised about Coase. As an analyst of alternative institutional regimes, what is his epistemological and axiological locus? If he presupposes the institutional status quo, his conclusions are circular. If he is a god-like observer of the economic universe who transcends all institutional regimes—hovering above the frenzied world of externalities like Hermes, the god of traders and entrepreneurs—then his status is obscure and mysterious. In either case, this dilemma is not attractive for the Coasean theory of policy analysis, which ends in either a vicious circle or a metaphysics.
Research on this paper was supported by the Jack T. Kvernland Chair, Monmouth University. Our thanks to two referees, Steve Medema, and Samuel Ferey should not be understood as a conventional pro forma acknowledgment. In following their suggestions, we were pushed to recast some of the fundamentals of the essay, although not perhaps in the fashion they intended.