The Rise of Environmental Externalities

At the beginning of the 1950s, the US government was concerned with resource scarcity due to the huge and soaring demand for non-renewable resources (coal, oil, minerals, etc.) in a postwar context of strong economic growth. This issue, raised by the Paley Report,Footnote ⁴ led to the creation of the influential research organization Resources for the Future (RFF) in Washington DC in 1952. RFF is considered to have been an influential supporter of the development of benefit-cost analyses and economic instruments in the management of environmental issues (especially in water management; see Banzhaf Reference Banzhaf2009, Reference Banzhaf2010). While it was initially created to focus on the use of scarce natural resources, in the 1960s it began to study the problem of pollution, following rising concerns about the issue. Kneese, one of the fathers of environmental economics, joined RFF in the early 1960s and became the director of the water resources and environmental quality research programs. In 1966, he organized the sixth RFF forum, entitled “Environmental Quality in a Growing Economy,” which was entirely devoted to this issue. As stressed in the introduction to the conference proceedings:

This new focus on quality of life and pollution was related to the renewal of the theory of externality. Compared with its growth in the 1960s and 1970s, the literature on externalities was relatively scarce during the 1950s: moreover, when it was mentioned at all, the concept of externality could refer to very different phenomena such as increasing returns and indivisibilities—mainly, all kinds of market failures—although a few contributions were aimed at providing a clear definition (for example, Meade Reference Meade1952; Scitovsky Reference Scitovsky1954; Bator Reference Bator1958). It was only in the 1960s that the use of the concept really took off (e.g., Coase Reference Coase1960; Buchanan and Stubblebine Reference Buchanan and Stubblebine1962; Davis and Whinston Reference Davis and Whinston1962; Baumol Reference Baumol1964; Mishan Reference Mishan1965; Arrow Reference Arrow1969, etc.), and that a more consensual definition of “externality” emerged, mainly referring to unpriced individual interactions (see Berta Reference Berta2017 for a history of this emerging definition). In addition, even if, from a theoretical point of view, the suboptimality induced by externalities was broadly recognized in the early 1960s, externalities were still regarded as “exceptional and unimportant” phenomena, pertaining to hypothetical rather than empirical issues (Medema Reference Medema2020). According to Kneese (Reference Kneese1971a, p. 2), “even that subset of economists especially interested in externalities seems to have regarded them as rather freakish anomalies in an otherwise smoothly functioning exchange system.” And, as Peter Bohm and Kneese also noted, “it is not until the last five years or so that the economic literature has begun to recognize explicitly that externalities are a pervasive element in highly developed exchange economies” (Bohm and Kneese Reference Bohm, Kneese, Bohm and Kneese1971, p. xii). The growing interest in the theory of externality in the 1960s owed a lot to the growth of environmental concerns, since the concept found its main field of application in environmental issues, especially pollution.

Economists versus Regulators: Worlds Apart

Although considering environmental externalities as an important source of damage was still a new idea in the early 1960s, economists had already developed a clear and fairly consensual idea of the best way to manage such market failures. According to the new welfare economics, market failures were regarded as deviations from the optimum, requiring State intervention. The Pigovian tax, set so as to be equal to the social marginal damage, was supposed to internalize the externality and achieve optimality. The Pigovian approach dominated the theoretical literature on externality until the 1950s (see Sandmo Reference Sandmo2015 for a history), but this situation also persisted in the 1960s when externalities became a subject of environmental studies (see Oates Reference Oates2000, p. 137). Kneese, for example, was a dominant figure among the early advocates of the use of taxes to manage environmental issues. He wrote two highly influential books on water quality management (Kneese Reference Kneese1964, reprinted and extended four years later in Kneese and Bower Reference Kneese and Bower1968); relying on the externality concept, he stressed the cost savings that chargesFootnote ⁵ could achieve in environmental policies compared with standards and direct controls, and more generally promoted the economic valuation of pollution control (see also several contributions on air and water pollution at the sixth RFF forum in 1966, published later in Jarrett [1966] Reference Jarrett1970; e.g., Gaffney [1966] Reference Gaffney and Jarrett1970; and Turvey [1966] Reference Turvey and Jarrett1970).

Coase (Reference Coase1960) had already proposed another way to internalize externalities in a Pareto optimal way, through bargaining on the externality price and level between the receptor and the emitter. However, his influence remained relatively slight in the environmental literature of the 1960s (see Medema Reference Medema2014 for a history of the reception of the Coase theorem in environmental economics). Coasean bargaining was mentioned by Kneese and Ralph Turvey at the sixth RFF forum, for instance, but was not taken seriously, mainly because it was not considered appropriate for the large number of agents involved in most pollution issues. As argued below, Coase (Reference Coase1960) obviously had an influence on the idea of the permits market,Footnote ⁶ but the Pigovian tradition largely remained dominant in the economics literature on pollution issues.

However, from the policy makers’ point of view, things were different. As noted by Thomas H. Tietenberg (Reference Tietenberg2010, p. 359), “by the late 1950s, both economists and policymakers had formed quite well developed and deeply entrenched visions of how pollution-control policy should be constructed. Unfortunately, these two visions were worlds apart.” While economists promoted the Pigovian solution, policy makers preferred traditional regulation: direct controls such as quotas, norms, or standards. In the 1960s the environmental legislation mainly relied on such instruments, as witnessed in the United States by the 1963 Clean Air Act, for instance, and this remained the case until the 1970s, as illustrated by the 1970 Clean Air Act Amendments and the 1972 Clean Water Act Amendments, which mainly relied on pollution or effluent standards imposed on polluters. As later stressed by Anthony Fisher and Frederick Peterson:

So, the first argument is that Congress and executive agencies were ignoring economists. Yet, even in the 1970s, environmental issues were still not clearly regarded as economic issues, despite economists’ recommendations to use economic criteria and to estimate the costs and benefits of environmental policies. For example, the Council of Environmental Quality, created in 1970 to advise the US president on environmental issues, included only a tiny minority of economists. As a result, for environmental economists promoting the “rationalization” of policies through cost evaluation, economic analysis “has been excessively abbreviated and appears to have had no effect on the discussion of political developments on other parts of Environmental Quality [the annual report of the Council of Environmental quality]” (Mills and Peterson Reference Mills and Peterson1975, p. 267).

The second explanation for the gap between regulators’ and economists’ opinions lies in the “impractical” policies recommended by the latter. Kula (Reference Kula1998) evoked the predilection of economists for unnecessary mathematics and jargon, as well as the lack of available information on the marginal costs and damages functions. And Oates would later argue, regarding the Pigovian approach of externality:

And this is the point on which I want to focus here: what was at stake was the unworkability of the Pigovian solution and more largely of optimal solutions that required perfect knowledge of the marginal cost and damage functions.

The Issue of Damage Measurement

Yet, from a theoretical point of view, the calculation of the optimal Pigovian tax relied on the balance—and thus on the measurement and monetary valuation—of the externality marginal cost and damage. And applying this approach to the growing pollution problems was quite new. Among the most influential works that addressed the practical issue of pricing abatement costs were Kneese (Reference Kneese1964), and Kneese and Bower (Reference Kneese and Bower1968). As already stressed, these authors were strong advocates of the economic rationalization of pollution controls through their cost minimization, and promoted the use of taxes in general. They tried to compare the costs of several policies, mainly relying on empirical studies, e.g., those applied in Germany’s Ruhr Basin or in the Delaware estuary in the United States.Footnote ⁷ And their work was considered as a promising attempt at pricing, though imperfectly, both the abatement costs and the pollution damage (see, e.g., Falkson Reference Falkson1969 for a review of Kneese and Bower Reference Kneese and Bower1968).

It should be noted that although trying to apply the Pigovian approach to pollution issues was quite new in the 1960s, benefit-cost analysis had already been applied to public investment projects in the United States from the 1930s, and especially to water resource projects that tried to manage irrigation, navigation, and flood control (see Prest and Turvey Reference Prest and Turvey1965 for a state-of-the-art of that time; and Castle et al. Reference Castle, Kelso, Stevens, Stoevener and Martin1981 or Hufschmidt Reference Hufschmidt2000 for broader histories). In the 1950s and 1960s, several publications and reportsFootnote ⁸ tried to provide new principles and procedures to estimate the benefits and costs of water resource projects, using both welfare economics and related microeconomic theory. These benefit-cost analyses raised several issues, including the valuation of benefits, especially of non-market goods like intangible goods (mainly outdoor recreation), and the incommensurability of different kinds of benefits (Banzhaf Reference Banzhaf2009, Reference Banzhaf2010). And this issue of the measurement of the benefits of water resource projects obviously echoed the issue of damage measurement in pollution control.Footnote ⁹ So it was not surprising that most of the participants at the sixth RFF forum also stressed, with more or less optimism, the difficulty of pricing abatement costs and the damage associated with pollution, and the lack of available data. “Our skill in estimating future benefits and costs in a changing world is not large and our practice of using money market rates for discounting future streams of uncertain benefits and costs may not be fully appropriate for questions concerning social regulations of environmental quality in a perpetual society” (Barnett [1966] Reference Barnett and Jarrett1970, p. 19).

At the same time, again regarding the balance between cost and benefit, Turvey noted:

Measuring the damage—in other words, measuring the benefits of pollution control—was considered more difficult than measuring the abatement costs. And indeed, at the end of the 1960s, very few empirical studies were devoted to this issue of pollution damage measurement.Footnote ¹⁰

To sum up, in the late 1960s, the growing environmental problems such as pollution were regarded as important externalities to be managed at a very practical level. From a theoretical point of view, following the new welfare economics, many environmental economists considered that such market failures should be managed by Pigovian taxes. They thus started to promote such policies, stressing the cost savings that they could achieve compared with the traditional regulations favored by policy makers. However, from a practical point of view, these policies raised the contentious issue of the estimation of the pollution damage. Therefore, this context was a fertile ground for new policies that would break with traditional regulation—regarded as workable but inefficient—and also with Pigovian taxes—regarded as optimal but hardly workable. In other words, there was a perceived need for compromise.

The Pigovian Tax and Its “Fairly Esoteric Marginal Net Damage”Footnote ¹¹

Baumol and Oates were both at Princeton University in 1971. Baumol had already worked on several topics linked to new welfare or public economics, especially on the theory of State intervention and the related concept of externality.Footnote ¹² Oates was younger; he joined Princeton just after having received his PhD from Stanford in 1965, and had mainly worked on fiscal policy and public economics.Footnote ¹³ Their 1971 paper marked for both the beginning of a series of contributions in environmental economics—including their later seminal book The Theory of Environmental Policy (1975).

Influenced by public economics, they strictly followed the Pigovian approach. First, as usual among new environmental economists, they considered that it would be more efficient than traditional regulation. They did not even discuss the relevance of the latter, and rather sweepingly affirmed instead that “the inefficiencies of a system of direct controls, including the high real enforcement costs that generally accompany it, have been discussed often enough; they require no repetition here” (Baumol and Oates Reference Baumol and Oates1971, p. 44).Footnote ¹⁴ But at the same time they acknowledged the unworkability of the Pigovian approach: “in practice, such an approach has rarely proved feasible because of our inability to measure marginal social damage” (Baumol and Oates Reference Baumol and Oates1971, p. 42). They recognized that Kneese and Bower (Reference Kneese and Bower1968) “report some extremely promising work constituting a first step toward the estimation of the damage caused by pollution of waterways including even some quantitative evaluation of the loss in recreational benefits” (ibid., p. 43). But they remained relatively pessimistic, first, because of the large number of agents involved in most pollution issues—since both considered that most environmental issues require a general equilibrium approach; second, because of the intangible nature of widespread damage (esthetic or health damages, for instance);Footnote ¹⁵ and, third, because in order to find the optimal tax level, we need to know not only the total amount of damage associated with pollution but also the marginal damage. Furthermore:

What they imply here is that we need to know not only the marginal damage associated with a specific pollution level but also the whole marginal damage function. For all these reasons, they considered that this optimal level is impossible to compute—an argument they maintained a few years later, in the same terms, in The Theory of Environmental Policy: “It is not easy to obtain a reasonable estimate of the money value of this marginal damage,” so that “it is hard to be sanguine about the availability, in the foreseeable future, of a comprehensive body of statistics reporting the marginal net damage” (Baumol and Oates Reference Baumol and Oates1975a, p. 135).

The Trade-Off between Practicability and Theoretical Purity

As a consequence, Baumol and Oates considered that we should give up any claim to Pareto optimality. But at the same time, they considered that “the economist’s predilection for the use of the price mechanism makes him reluctant to give up the Pigovian solution without struggle” (Baumol and Oates Reference Baumol and Oates1971, p. 44). Thus, they proposed that the tax system be selected so as to achieve “specific acceptability standards rather than attempting to base them on the unknown value of marginal net damages” (ibid., p. 45). These acceptability standards then represent “the decision-makers’ subjective evaluation of the minimum standards that must be met in order to achieve what may be described in persuasive terms as ‘a reasonable quality of life’” (ibid., pp. 44–45). This approach reflects Baumol and Oates’s pragmatism, their common willingness to provide practical policies for decision makers who had to face growing concerns about pollution issues. In later presentations of their Reference Baumol and Oates1971 standard-and-tax approach, they stressed that it “recommends itself more for its promise of effectiveness than its theoretical nicety” (Baumol Reference Baumol1972, p. 307) and that rejecting it “on the grounds that they will probably violate the requirements of optimality may well be considered a kind of perverse perfectionism” (Baumol and Oates Reference Baumol and Oates1975a, p. 151).Footnote ¹⁶

Baumol later provided comments on his method in general that help us to understand this choice (Baumol Reference Baumol1984; Krueger Reference Krueger2001; see also Noci Reference Noci2019). He stressed that he has always been oriented towards applied theory and never interested in the construction of models for their own sake. “Optimality in model construction must be based on the tradeoff between these two desiderata—accuracy of representation of reality and usability in analysis” (Baumol Reference Baumol1984, p. 7). This is why the standard-and-tax approach can be seen as the result of such a trade-off between pure theory and its related optimality criterion and policy-oriented theory that agrees to set aside optimality for better feasibility. In the introduction of their 1975 textbook, Baumol and Oates clearly announced: “This volume is primarily theoretical and is consequently addressed to our fellow economists. However, it is not meant to be theory for theory’s sake. Here our prime concern is policy; we are interested in the theory as a means of understanding the complexities of environmental programs” (Baumol and Oates Reference Baumol and Oates1975a, p. xi).

Oates, contrary to Baumol, did not write any papers on his own method in economics. But in a later survey of environmental economics, written with Maureen L. Cropper, he would reassert his interest in policy-oriented and applied economics in similar terms: “[T]here is a real opportunity for environmental economists to make valuable contributions in the policy arena—if, as we shall argue, they are willing to move from ‘purist’ solutions to a realistic consideration of the design and implementation of policy measures” (Cropper and Oates Reference Cropper and Oates1992, p. 676). And all his work focused mainly on the economics of environmental policy issues, with a number of journal publications aimed at operationalizing and evaluating pollution-control schemes (McGartland Reference McGartland2015).

Baumol and Oates’s Least-Cost Theorem

Baumol and Oates showed that a uniform tax system would allow the predetermined acceptability standards to be realized at least-cost. As is widely acknowledged today, in response to a tax, a firm is assumed to reduce its pollution until the point of the equalization of its marginal abatement cost to the tax. This would lead to the equalization of the firm’s marginal costs, the condition for total cost minimization. The last requirement is then to adjust the tax level so as to reach, in response, the given pollution standard. However, according to Baumol and Oates, the authorities’ experience allows them to set appropriate tax levels so that less information is needed for an iterative adjustment tax in order to reach the given standards.

Yet, this idea was fairly intuitive and, as the authors acknowledged, it was not totally new: “it or a very similar proposition has been suggested in a number of places” (Baumol and Oates Reference Baumol and Oates1971, p. 46), and they referred to Kneese and Bower (Reference Kneese and Bower1968) and also to a paper from Larry E. Ruff (Reference Ruff1970).Footnote ¹⁷ However, they continued: “we have been unable to find an explicit statement of this result anywhere in the literature” (Baumol and Oates Reference Baumol and Oates1971, p. 46).

Their mathematical proof of cost-efficiency relied on the individual production functions of price-taker firms. First, a program of total cost minimization, set for a given level of output and a given pollution standard, was solved by a Lagrangian function, which gave the first-order conditions. The second step consisted of introducing a uniform tax instead of the imposition of a fixed standard. The first-order conditions for the individual cost minimizations under such a tax are the same as the previous condition for total cost minimization: the Lagrange multiplier of the emissions constraint, the shadow price of emissions, being interpreted as a tax.

Baumol and Oates did not recognize that the same cost-efficiency could be achieved with a permits market. Furthermore, and quite remarkably, they did not refer to the idea of a permits market at all. They did mention Coase (Reference Coase1960) but just in passing, in a footnote, since they did not take the Coase theorem seriously into account. While they acknowledged that the latter “may in some instance” be optimal, they immediately stressed that “one can hardly imagine, for example, a voluntary bargaining process involving all the persons in a metropolitan area” (Baumol and Oates Reference Baumol and Oates1971, p. 47; and also Baumol and Oates Reference Baumol and Oates1975a, Reference Baumol, Oates and Mills1975b). In other words, they considered it was not suitable for the large number of agents usually involved in pollution issues—a traditional criticism of Coase’s theorem at that time; so that the Coasean solution was, like the Pigovian one, rejected for its unfeasibility and not on theoretical grounds.Footnote ¹⁸

The Lack of Property Rights

Dales was a Canadian professor in economics at the University of Toronto, working on the pollution of the Great Lakes. He presented his idea of a permits market in a book entitled Pollution, Property and Prices: An Essay in Policy Making and Economics (Reference Dales1968a), which was summarized the same year in an article published in the Canadian Journal of Economics (Dales Reference Dales1968b). At that time, Dales had already published a few papers on industrial development—especially on the Canadian manufacturing industry, stressing its impacts on energy resources or the effect of protectionism on its development. But he had never before published any contribution on environmental issues—and neither would he do so afterwards. Dales’s book on pollution was mostly an applied study and, contrary to Baumol and Oates (Reference Baumol and Oates1971), was not underpinned by any formal basis. It provided a broad reflection on property rights, applied to environmental issues. In the preface, Dales affirmed that studying social problems as pollution issues had led him to focus on the link between law and economics, and more precisely between prices—“the stuff of economics”—and property rights (Dales Reference Dales1968b, p. xi). And indeed, two chapters were entirely devoted to property rights, which seemed to be one of his main topics. This focus on property rights led him to associate environmental issues, especially water pollution, with lack of ownership, and to stress “the enormous social friction and economic waste that result from not applying an ownership-rental system to water” (Dales Reference Dales1968b, p. 795).

On this issue of property rights, he referred mainly to Coase (Reference Coase1960), but also to H. Scott Gordon’s (Reference Gordon1954) seminal paper.Footnote ²⁰ He considered that Gordon (Reference Gordon1954) had shown “the wastefulness of common-property arrangements” (Dales Reference Dales1968a, p. 111). And claiming that “common property,” which he understood as a free good, simply means “no property,” Dales rejected it. Dales did mention Coase regarding the relationship between law and economics and the definition of property rights,Footnote ²¹ but it is interesting to notice that he never discussed or even referred to Coasean bargaining. The Coase theorem had just been established by George Stigler (Reference Stigler1966), and, as already mentioned, its influence on the environmental economics literature was still limited (Medema Reference Medema2014). Contrary to Baumol and Oates, who would reject it on the grounds that it is not suitable to a large number of agents, Dales did not evoke the idea at all. Thus his idea of a permits market was not presented as a development or an adaptation of the Coase theorem to situations involving a large number of agents. In other words, contrary to what we might have expected retrospectively, the idea of a permits market was not a way to address the unworkability of the Coasean solution.Footnote ²² As stressed below, it was mainly presented as a way to address the unworkability of optimal solutions, especially the optimal taxes and their shadow prices, which were impossible to compute.

Obviously, though, this does not mean that Dales favored the private ownership of natural resources like air or water or more broadly the privatization of nature. Following Coase, he drew a distinction between property rights on an asset and property rights on the use of an asset. And, more importantly, he drew a distinction between private property and property rights, considering that property rights can be vested in individual groups, firms, or governments, and do not necessarily imply private property.

“Simple Problems, Simple Solutions”Footnote ²³

Dales favored incentives-based policies as against traditional regulation. However, he also rejected solutions that attempted to reach optimality. He set out two conflicting uses of water: amenity use and waste disposal. The social optimum involves ensuring that these two uses are allocated in such a way that the marginal benefit of one use is equal to the marginal cost of the other. However, this social optimum cannot be measured because we lack the relevant information. So Dales did not reject such an idealized optimal solution on theoretical grounds; rather, his arguments were pragmatic. In a chapter entitled “Simple Problems, Simple Solutions,” he gives an enlightening example of the pollution of an imaginary lake, which leads to health issues, to reduced recreational uses of water, and to losses of esthetic value. He then summarizes all the benefits and costs of the several available policies in a table and concludes that such a table “provided us with an orderly way of thinking, … a guide to straight thinking” (Dales Reference Dales1968a, p. 29) that helped to solve the problems. But, he added ironically, “the problems are not solved by the benefit-cost machine, it is solved because we took great pains to adopt assumptions that made them solvable” (ibid., p. 36). In particular, such an approach assumed that all the benefits and damages are known and measurable, and furthermore that they are measurable in terms of dollars.

Like Baumol and Oates, Dales acknowledged Kneese’s work and the progress made in pricing damage and the cost of water pollution.Footnote ²⁴ He also referred to a survey of benefit-cost analysis by Alan R. Prest and Ralph Turvey (Reference Prest and Turvey1965), which mainly provided favorable arguments regarding the use of such analysis.Footnote ²⁵ But, again like Baumol and Oates, he remained far more pessimistic and focused mainly on the difficulties associated with measuring pollution damage, adducing very similar arguments. First, he underlined that environmental issues usually involve many agents and require a general equilibrium approach (Dales Reference Dales1968b, p. 799); second, he stressed the intangible nature of pollution damage, especially regarding recreational uses of water, health damage, or esthetic value.

He therefore rejected every attempt to reach an optimal solution. And, regarding the principle of Pigovian taxes, he noted: “the administrative problem of approximating optimum shadow prices by actual user charges promises to be a nightmare” (Dales Reference Dales1968b, p. 791). Or, in other words, “in principle” the values of marginal benefit and damage should be equal, but “in practice” the optimal amount cannot be computed (p. 797). Therefore, the suitable level of pollution has to be a political decision: economics cannot be of any significant help in making this decision (Dales Reference Dales1968b, p. 799). This does not mean that economists have nothing to say on the matter, it just means that “if economics has nothing useful to say about the ownership decision of how water should be used, it has a great deal to say about how the decision, once made, should be implemented” (Dales Reference Dales1968b, p. 799). And hence Dales proposes his permits market solution.

His proposal is well known, and is close to what are today usually named “cap-and-trade” schemes. It does not require a long explanation here. Like the standard-and-tax approach, it involved a clear separation between the predetermined standards and the pricing procedure, which is left to the market. So once the pollution level is politically chosen, an equivalent number of pollution rights are issued, put up for sale, and made tradable.

Dales reviewed several policies—regulations, taxes, subsidies, etc.—that could be implemented to achieve these standards, and then compared their relative economic efficiency and their administrative costs. The trading of emissions rights is expected to equalize the polluters’ marginal abatement costs—the condition for total cost minimization. Dales did not straightforwardly or formally establish this property of cost minimization. A few years later, W. David Montgomery (Reference Montgomery1972) established a rigorous proof of the cost-efficiency of the permits market.

Montgomery’s Proof of Cost-Efficiency

Montgomery’s contribution was part of his PhD submitted at Harvard University in 1971, and supported by the Environmental Protection Agency (EPA). Unlike Dales, Montgomery was not writing against the Pigovian tradition and did not even mention it. He plainly adhered to the Arrow-Debreu competitive equilibrium tradition.Footnote ²⁶

More precisely, his work was deeply inspired by Dales and Kenneth Arrow. Montgomery was Arrow’s student, and he clearly acknowledged the influence of Arrow’s Reference Arrow1969 paper on the market internalization of externalities. At the very end of the 1960s, two different modes of market internalization were suggested, each, under specific assumptions, allowing the achievement of Pareto optimality: Coase’s bilateral bargaining in an Edgeworthian framework with price-maker agents; and Arrow’s (Reference Arrow1969) competitive market in a Walrasian framework with price-taker agents (see Berta and Bertrand Reference Berta and Bertrand2014 for a detailed comparison of the two kinds of internalization). Montgomery clearly followed the Arrovian approach. This is probably why he did not even mention Coase (Reference Coase1960)—whom he probably knew at the time.Footnote ²⁷

Arrow (Reference Arrow1969) proposed an internalization mode that expanded the complete system of markets to include externalities in a general equilibrium framework (see Berta Reference Berta2017). More precisely, Arrow established the conditions for Pareto optimality in a general equilibrium framework with competitive markets for externalities. He provided an example of widespread externalities in consumption. Reaching Pareto optimality required broadening the markets system to all externalities characterized as “missing markets.” New commodities were created, one for each personalized externality, each being characterized by two agents: the producer and the recipient. Arrow then showed that the new competitive equilibrium, with personalized prices, exists and is Pareto optimal. Nevertheless, as underlined by Arrow himself, the price-taking assumption, which is usually justified by “atomistic competition,” is not relevant in such a market that involves only two agents. Given this lack of realism, Arrow did not regard his own theoretical solution as empirically relevant.

Montgomery’s paper started precisely by underscoring the same issue—in other words, the unworkability of Arrovian markets:

Montgomery rightly stressed that, contrary to Arrovian personalized externalities, in many forms of pollution the receptor does not care who precisely the emitter is. Thus, referring then to Dales, he proposed the creation of competitive markets that would “bring together many buyers and sellers” (Reference Montgomery1972, p. 395). While Pareto optimality cannot be reached with such markets, Montgomery proved that the competitive market equilibrium, today called the “least-cost allocation,” can achieve specific standards at least cost for the regulated firms. The structure of mathematical proof was similar to that of Baumol and Oates because the cost-minimization program under specific pollution standards required solving a Lagrangian function whose multiplier, the shadow price for pollution, can be associated with a market price. Tietenberg stressed that “because the mathematics is perfectly equivalent for the two cases, the result derived for a charge system immediately was recognized as relevant for emissions trading” (Tietenberg Reference Tietenberg2010, p. 360). However, Montgomery mentioned Kneese and Bower (Reference Kneese and Bower1968), but not Baumol and Oates (Reference Baumol and Oates1971). He seemed to have followed two different ideas, those of Arrow and Dales, whose influences were clearly acknowledged in the introduction. Hence the similarity of the mathematical proofs is not sufficient evidence that Baumol and Oates’s proof, even if it came slightly earlier, was recognized as relevant for emissions trading.

Dales’s Focus on the Administrative Costs Issue

Dales was aware that the standard-and-tax approach could also minimize the total abatement cost through the equalization of the polluters’ marginal costs. As already stressed, the idea that mixing standard and tax could achieve important cost savings was first suggested by Kneese (Reference Kneese1964)—a source to which Dales often referred. But he stressed that, given the uncertainty on the firms’ marginal abatement costs, a uniform tax would require costly ex post adjustments of the tax level by public authorities to achieve the predetermined standards. This is why he called the standard-and-tax approach the “trial-and-error procedure” (Dales Reference Dales1968a, p. 91). This today is a common argument in favor of a market instead of a tax, since the former obviously guarantees the achievement of the standard. It should be noted that Baumol and Oates (Reference Baumol and Oates1971) acknowledged that their solution required such ex post adjustments, but they did not see it as a problem:

More precisely, Dales’s main argument was that, compared with these tax adjustments, a permits market would reduce administrative costs since “the virtues of the market mechanism are that no person, or agency, has to set the price—it is set by the competition among buyers and sellers of rights” (Dales Reference Dales1968b, p. 801). So, finally, Dales compared not only the abatement costs—on this issue, both solutions were equivalent and considered as superior to regulation—but also their administrative costs, that is, mainly the costs involved by their enforcement and implementation.Footnote ³⁰ His main aim was always to find the less costly solution—comparing both the abatement costs for polluters and the administrative costs for public authorities, in a very Coasean way.Footnote ³¹ This led him to conclude that “the administrative simplicity of the [permits market] scheme is certainly one of its main attractions” (Dales Reference Dales1968a, p. 97).

Montgomery used very similar arguments: even if he did not refer to Baumol and Oates (Reference Baumol and Oates1971), he was also aware that a tax system would lead to exactly the same outcome as the market solution. And he also stressed the administrative cost savings allowed by his solution, compared with the “iterative process by varying the tax …, a cumbersome and politically unattractive procedure,” while “the market makes the necessary calculations independently in the course of reaching equilibrium” (Montgomery Reference Montgomery1972, p. 411).

It should be noted that neither Dales nor Montgomery provided an explanation of how the competitive equilibrium prices would be reached. Specifically, they always implicitly assumed that the competitive equilibrium—that is, the least-cost allocation that is expected to equalize polluters’ marginal costs—would be reached at no cost. Thus both Dales and Montgomery pointed out the costly adjustments that a tax system would require, but at the same time both implicitly assumed the costless achievement of the competitive allocation, the costless market price adjustments—calling upon the “virtues of the price mechanism” or the market’s “course of reaching equilibrium.” Indeed, it is only under such a hefty assumption that a tax system can be regarded as costly compared with a permits market: actually, the standard-and-tax approach could have significant administrative costs in its enforcement by public authorities, due to the ex post administrative adjustments it requires; but the market solution could also bring significant costs, not only in its administrative enforcement, which both Dales and Montgomery considered lower than those costs of a tax, but also in its process of price adjustments.Footnote ³²

Baumol and Oates’s Skepticism Regarding the Permits Market

On the side of the standard-and-tax proponents, things are different. Baumol and Oates did not discuss or evoke the permits market idea in their Reference Baumol and Oates1971 paper, and they did not even mention Dales’s work. It is possible that they did not know about it in 1971, but this was no longer the case after 1972 when Baumol (Reference Baumol1972) briefly referred to it, for the first time, as “an earlier proposal which is very similar in spirit” (p. 313). But this was only in a single phrase, relegated to a footnote. A few years later, in Baumol and Oates’s (Reference Baumol and Oates1975a) textbook, the idea of permits market still was mentioned only in passing, in two footnotes,Footnote ³³ and not set out, despite the publication in the meantime of Montgomery’s (Reference Montgomery1972) contribution (which was also not mentioned). So they knew that Dales’s proposal could, from a theoretical point of view, achieve the same cost-effectiveness as their tax system, but they neither discussed nor criticized it. Given the comprehensiveness of Baumol and Oates’s 1975 book, this omission seems quite strange.

That same year they published a paper entirely devoted to the spectrum of available policies for pollution control (Baumol and Oates Reference Baumol, Oates and Mills1975b). The permits market proposal was again consigned to a footnote, with Dales not even mentioned. But the footnote was more enlightening regarding the reasons why they treated this idea so cavalierly: “The auctioning of pollution rights could be added here. However, considering the major environmental problems before us, the practicality of the proposal seems to us rather limited” (p. 97). And indeed, this is the more plausible explanation of their systematic relegation of Dales: they simply did not consider the idea seriously, regarding it as unworkable. Oates would later, at several times, suggest: “my first reaction to the appearance of Dales’ work was one of skepticism” (Oates Reference Oates2000, p. 142); and in his foreword to the re-edition of Dales’s book in 2002, “I must admit that when I first read the Dales’ essay, although I found the basic economics of the argument quite sound, I simply shook my head when thinking about the political feasibility of the proposal” (Oates Reference Oates and Dales2002, p. viii). It is interesting to notice that Oates was not skeptical regarding its technical workability but rather that his argument was mostly political: the environmentalists were strongly opposed to the use of taxes at that time, considering it immoral, and Dales’s proposal in such a context seemed to him politically unacceptable.Footnote ³⁴

It was only in 1979, more than a decade after its publication, that Dales’s contribution was really discussed by Baumol and Oates. In their second book, Economics, Environmental Policies and the Quality of Life (Reference Baumol and Oates1979), they stressed that in principle, both tax and market solutions are equivalent, and they provided some elements of comparison, all very pragmatic.Footnote ³⁵ But contrary to Dales or Montgomery, they did not engage in a comparison of the administrative or transaction costs involved. Like Martin Weitzman (Reference Weitzman1974), they focused rather on the two different kinds of uncertainty involved (uncertainty on the quantity with a tax, or on the price—that is, on the cost that will finally bear on society—with a market). Finally, for policy makers, the choice “may then be a matter of which risk constitutes the greater danger” (Baumol and Oates Reference Baumol and Oates1979, p. 253).

So compared with 1971 or 1975, they had changed their mindsFootnote ³⁶ but did not have a well-entrenched opinion, and rather sought a balance between the two solutions or, more precisely, between all the available policies. Even if they favored the use of taxes, they always remained pragmatic, since “there is no one approach or one technique that is best suited to deal with all problems in all circumstances. There is no panacea, no one simple method, that is always best (or even always workable)” (Baumol and Oates Reference Baumol and Oates1979, p. 229). It is pertinent to stress that they again raised the issue of the political attractiveness of the market solution, a point that echoes Oates’s later acknowledgment of his early skepticism about the proposal: “In sum, if the public and government officials can ever be weaned from the idea that auctioning pollution permits is a wild, ivory scheme, they may find that the device has a good deal to recommend it” (Baumol and Oates Reference Baumol and Oates1979, p. 251). But they immediately balanced this opinion: “it does not seem to have a high degree of political appeal. Noneconomists usually regard it as an impractical and cumbersome product of the convoluted minds of academia” (ibid., p. 252).

Dales, Baumol, and Oates thus all stressed the theoretical equivalence of their solutions, and all favored their own proposals on pragmatic grounds. But they did not focus on the same issues: Dales focused on the administrative costs involved in each solution, in a very Coasean institutional comparison, even if he singularly did not take into account all the transaction costs involved by his permits market; Baumol and Oates focused on the kind of uncertainty (on price versus quantity) involved in the real world where marginal abatement costs are not perfectly known, an argument that would become a standard element of comparison between the two instruments.