In the analysis of market processes, the concept of externalities has long invoked strong public policy implications among economists regarding the role of government in addressing their alleged presence, or lack thereof.

It is for this reason that analyses of externalities have preoccupied economists, at least since A.C. Pigou. The more important question, however, is how do they matter for economics?


In an excellent and thought-provoking article recently written by Don Boudreaux and Roger Meiners (2019), they provide a comprehensive overview of the origins of the concept of externality, its evolution, and its classifications. Boudreaux and Meiners put forth a subtle though sophisticated argument that transcends how economists generally arrive at particular public policy implications regarding externalities. My goal here is not simply to summarize their point, but also to provide my own interpretation of their argument.

To summarize the concept briefly, an externality refers to a spillover cost borne by third parties to an exchange. Externalities arise when the market price at which a good is exchanged fails to account for the full cost (in the case of a negative externality) or benefits (in the case of a positive externality) of producing a good. Such costs or benefits are, therefore, not only involuntary, but more importantly, unexpected by third parties to the exchange. Externalities are indicative of a deviation from the ideal of perfectly competitive equilibrium, in which all potential gains from trade have been exhausted. Thus, the resulting “market failure” associated with externalities arises from the fact that “there exists a reallocation of resources, such as a change in the structure of market activities that will enrich society” (Boudreaux and Meiners, 2019, p. 21). This restructuring not only includes an adjustment in market prices to more fully concentrate the costs and benefits of exchange upon the relevant trading partners, but also, more importantly, an adjustment in the assignment of property rights, the exchange of which gives rise to exchange ratios, or market prices, in the first place. I will return to this last point later.


The fundamental aspect on which Boudreaux and Meiners focus is not the involuntary nature of externalities, but the degree to which they are expected or not. As they put it, “Externalities exist only when another party’s actions create unexpected spillover effects” (emphasis in original, 2019, p. 29).

Building on Alchian (1965) and Demsetz (1967), Boudreaux and Meiners make clear that property rights “are a bundle of expectations” (emphasis in original, Boudreaux and Meiners 2019, p. 31) about how individuals can choose to use, exclude, and exchange resources. The expectation in a market economy is that property rights allow “harm” to the exchange value of a good or service, but not to its physical characteristics.

Consider a barber who sets up shop in midtown Manhattan, assuming well-defined and enforced property rights, physical impediments to his expected ability to utilize his physical and human capital for providing haircuts constitutes an externality. For example, theft or other physical damage to the barber’s property rights are assumed to be prohibited. But, under these assumptions, he also has – or, as a reasonable person, should have – the expectation that a competing barber may “steal” away his clientele by providing a better haircut, thus reducing his income and the resale value of the capital he employs.

Here is where the concept of externality gets a bit tricky, and why “the term has become nearly meaningless due to its ubiquity,” according to Boudreaux and Meiners (2019, p. 3). Returning to the previous example, suppose that building construction is taking place on 5th Avenue, requiring the use of jackhammers. These jackhammers, no doubt, are a nuisance to the barber, and therefore might impede his ability to provide haircuts in a safe and productive manner. A bad haircut or the slip of a razor blade while shaving a client will result in a misallocation of resources in terms of “too much” building construction and “too few” haircuts and shaves than would otherwise be optimal. Is this indicative of a negative externality? Does this example prove that building construction should be taxed in order for contractors to take into account the noise pollution resulting from construction?

Not necessarily.

The key here is the role of expectations. No doubt, incurring the noise pollution from jackhammering was involuntary if the building contractors did not get the barber’s consent beforehand. However, we can reasonably conclude that when the barber chose to locate his shop in midtown Manhattan, one of the most densely populated islands on earth, he would have expected and anticipated such occurrences. The fact that he nevertheless located there implies that he expects the cost to him of this particular “externality” to be sufficiently low as to not overwhelm the prospect of greater expected monetary income derived from serving a larger and wealthier clientele than if he located in a less populated area outside the city.


What does this way of looking at externalities reveal about the welfare implications of the market process?

The fundamental point that Boudreaux and Meiners raise, as I understand it, is twofold. First, market processes are imperfect, meaning always in disequilibrium, and therefore imply that the expectations of individuals will never fully mutually coincide. If economists start in a world of disequilibrium as their analytic point of departure, then expectations about the costs and benefits of individual decision-making are never perfect (see Hayek 1937), in which case the concept of externalities, in an abstract sense, means everything and nothing. This implies, I would argue, that if the concept of externalities is to have any meaning and tractability, it must be grounded in an analysis of the particular expectations that individuals have in time and place. In doing so, it will provide the economist with a richer understanding of the public policy implications that follow from his or her analysis.

Secondly, to admit or to deny the presence of externalities is not analogous to admitting or denying the presence of market imperfection. Admitting the presence of externalities does not necessarily imply the necessity of government intervention. But, the absence of externalities does not necessarily imply Pareto efficiency in the allocation of resources either. It therefore does not follow that embracing one analytical point of departure or the other implies the dismissal of or appeal to government intervention as a corrective.


As I’ve written elsewhere, imperfect markets do not imply suboptimality or an inherent flaw as compared to the ideal of equilibrium. Rather, imperfection implies “incompleteness” and therefore that markets are processes incessantly moving towards completion. That completion process is facilitated by greater mutual coordination of expectations, requiring corrections in expectations, which makes market processes necessary to addresses misallocations of resources in the first place! As Boudreaux and Meiners make clear, “nothing said here suggests that the absence of spillovers implies a Pareto-optimal allocation of resources” (2019, p. 30). It simply implies the failure of the conditions of the market process to exist, not the existence of market failure (see Candela and Geloso 2020). “The problem, if one asserts there is a problem, is the structure of property rights” (Boudreaux and Meiners 2019, p. 30).

If I have correctly interpreted Boudreaux and Meiners, the question is not whether or not externalities matter for economists, but when they matter for economics, and how they matter for our analysis.


As an example to illustrate and conclude this point, let’s take the example of the solution devised by Julian Simon to the problem of airline overbooking (see Simon 1968). Generally speaking, airlines tend to overbook flights on the expectation that there will be a certain number of cancellations. Airline overbooking can then be reframed as problem of assigning property rights, since it creates a situation in which more than one individual has a claim on an assigned seat. When an airline involuntarily “bumps” an individual to another flight, can we conclude that represents an externality? Again, we must take into the context of time and place.

Indeed, the airline has generated a misallocation of resources through its decision-making. It exchanges a claim to a seat for money with a customer, but by assigning more than one customer to the same seat, there is a potential spillover cost on an individual bumped to a future flight, the full cost of which is not borne by the airline. However, we must conclude in this case that though bumping individuals to a future flight may be involuntary, it is not completely unexpected. Prior to the introduction of Simon’s auction proposal, an individual booking a flight could not rule out the possibility that a flight is overbooked. The anticipation of this possibility by individuals implies that this is not an example of an externality. However, to conclude this is not an externality does not imply this is a Pareto-optimal situation. There is indeed a misallocation of resources, since there are too many claimants to the available seats on a flight. Therefore, there is a profit opportunity to devise an institutional innovation to realize such potential Pareto improvements.

The introduction of the auction solution to the problem of airline overbooking can be understood as private property right solution, and therefore introduces new expectations between the airline and its customers about allocation of property rights. Given the transaction costs associated with correctly estimating the number of cancellations by customers, the introduction of the auction system (whether through a voucher or cash payment) grants all existing claimants to airline seats the ability to exchange, in effect creating private property rights in seats (Alchian 1965). Customers, in effect, not only become buyers of seats, but the auction system allows them to become potential sellers back to the airline in the case of overbooking. This exemplifies the point made by Phillip Wicksteed, namely that the supply curve for a good or service (in this airline seats) is part of the total demand curve for a good or service (see Wicksteed 1914, p. 13). More importantly, the exchange process generated through the auction system not only reduces the transaction costs associated with discovering the individuals with the lowest opportunity cost of moving to another flight (at a particular price that reflects such opportunity cost), it also reduces the transaction cost of economically calculating the minimum price necessary to pay an individual to be moved to a future flight. Such knowledge only arises in the context of the exchange of property rights (see Mises 1920 [1975]), which creates more consistent dovetailing of expectations between individuals.

Thus, the ability to assign private property rights in seats with the introduction of the auction system thereafter creates an expectation that individuals will be compensated if an airline mistakenly overbooks a flight. This brings me to the case of the United Airlines 3411 incident that took place on April 9, 2017, in which a passenger, Dr. Dao Duy Anh, was involuntarily dragged off the flight for refusal to give up his seat. This would seem to be a case of an externality, since the situation represents not only an involuntary cost borne unfortunately by the individual, but also because it was unexpected. Given the expectation that, through the auction system, an individual more willing to give up his or seat could likely have been discovered and paid a lower price than what Dr. Anh would have probably demanded as compensation for the airline’s error in overbooking.


Connecting this example back to Boudreaux and Meiners, the point here is that however one approaches this analysis, the existence or non-existence of externalities does not eliminate the fact that airline overbooking was representative of a misallocation of resources. And, the fact that this imperfection in the market process facilitated an institutional innovation to erode an existing inefficiency in the allocation of airline seats did not depend upon whether or not there existed an externality. And yet, the presence of an externality does not automatically presume a market failure, requiring government intervention, but a failure to secure the conditions of the market process, namely the voluntary exchange of property rights, due to government intervention. We can therefore conclude that the unfortunate circumstances that transpired during the United Airlines 3411 incident indicated the presence of a negative externality, but that this was a result of government failing to provide clear expectations about the security and enforcement property rights in airline seats, not a market failure.


The focus of analyses for economists, therefore, should not be to look backward at a presumed consistency or inconsistency in expectations between individuals, and then to pass normative judgment on it in terms of its conformity with Pareto-optimality or an efficient allocation of resources. Rather, the economist must always approach each analysis of any state of affairs as “nothing but a seething mass of unexploited maladjustments waiting to be corrected” (Kirzner 1979, p. 119), and focus on the constant adjustments that market processes facilitate in an open-ended world of uncertainty.

Rosolino Candela is a Senior Fellow in the F.A. Hayek Program for Advanced Study in Philosophy, Politics, and Economics, and Associate Director of Academic and Student Programs  at the Mercatus Center at George Mason University


Further References

Candela, Rosolino A., and Vincent J. Geloso. 2020. “The Lighthouse Debate and the Dynamics of Interventionism.” The Review of Austrian Economics 33(3): 289–314.

Hayek, F.A. 1937. “Economics and Knowledge.” Economica 4(13): 33–54.

Kirzner, Israel M. 1979. Perception, Opportunity, and Profit. Chicago, IL: University of Chicago Press.

Mises, Ludwig von. 1920 [1975]. “Economic Calculation in the Socialist Commonwealth.” In F.A. Hayek, ed. Collectivist Economic Planning (pp. 87–130). Clifton, NJ: August M. Kelley.