September 10, 2020
Ronald Coase’s Inimitable Contribution to Evidence Based Policy, By Victor Menaldo
Nobel Prize winning economist Ronald Coase’s 1960 piece, “The Problem of Social Cost”, is one of the most helpful ways to think about cost benefit analysis. It is therefore a pillar of evidence-based policy.
Setting the Table: What is Evidence Based Policy?
What is evidence-based policy? All public policy starts with a theory, whether it admits it or not. As practicing social scientists, we are all too aware that not all theories are created equal. While some are complete, logically consistent, and supported by evidence, the social science graveyard is littered with several bad ones that have caused untold damage. These include dependency theory—the idea that poor countries should seek to become self-sufficient by closing off their borders—and modernization theory, which states that developing countries will transform into liberal democracies once they reach a critical threshold of Per Capita Income. The short of it: after China flew past that threshold it turned to strong man rule. Poland, Hungary, and Turkey have fallen away from democracy, despite their relatively high standards of living, not to mention high urbanization and literacy rates.
Evidence-based policymaking seeks to evaluate policy based on the cogency of the theories that drive policy and, more importantly, on the evidence behind the claims they imply. When conducting evidence-based policy, therefore, there are several questions that should be asked. What is the problem to which the policy is the solution? What are the reasons (and logical organization) underpinning the claims made by the policymaker? What is the evidence in support of the reasoning? What data are used? Is there additional evidence (facts, data, and data analysis) that might be found and analyzed that complements, supplements, or challenges the findings that underpin the policy?
The first order of business with evidence-based policymaking is to be clear about our objectives and how to measure them and reach them. Disciplines such as economics, political science, and sociology can then help us understand the best strategies to deliver these objectives.
Before launching into particulars, however, evidence-based policy demands that we look strictly at the facts that cannot really be altered by policymakers, as those are important constraints that must be addressed regardless of objectives and strategy. It is also honest about the stakes involved: what is the magnitude of the problem that seeks to be assessed and how does it compare to other possible social problems? Finally, evidence-based policy demands that we ask: what would the counterfactual look like if we did nothing at all? Could this be the best possible route? Might alternative, market-based mechanisms emerge to deal with the problem?
The great thing about evidence-based policymaking is that it can save lives and make lives better if we allow it to, but that demands that we seek the truth and eschew motivated reasoning and confirmation bias. Those illogical ways of weighing theories and evidence allow bad policy making based on bad theories to spread virally.
Cost Benefit Analysis is Fundamental to Evidence Based Policy
When singing the praises of certain policies, we cannot only focus on the benefit part of the analysis, but must also think about the costs and net effects. This is where economics comes in and Coase’s “The Problem of Social Cost” in particular. The normative orientation of economics’ science of decision making is unapologetically utilitarian, based on maximizing total social benefits, which it imperfectly measures with outcomes that reveal themselves through market exchange (prices) and total economic activity (GDP). This is not because it is callous or blind to important human needs, wants, and rights beyond the measuring rod of money. Rather, it is because these things are hard to compare intersubjectively or measure and are highly correlated with variables such as consumer and producer surplus and “value added” in any event.
Therefore, whether we use prices or GDP or whatever other quantitative yardstick to evaluate the wisdom of a policy is irrelevant. You can pick whatever quantitative social outcome you’d like; the point remains: unless you choose the wellbeing of some actors over others, or prioritize someone’s right to life over somebody else’s, it is the total social consequences of policies and behaviors that you may want to consider. And, moreover, whatever outcome we choose, we should conduct a cost-benefit analysis to figure these things out. Because resources are scarce and there are tradeoffs, making one decision about how to allocate resources means they are not available for alternative uses. Cost benefit analysis therefore attempts to put a price tag on everything from natural resources to lives in terms of lifespans and future earnings.
Conducting cost benefit analysis also means articulating the right time horizons ex ante and making this discount rate clear. If we value outcomes that materialize in the future less it means we are “discounting” the future. Humans tend to do that because we have finite lives and are not infinitely patient or altruistic towards future generations. But zeroing in on a particular discount rate has nothing to do with science. Science is neutral about time horizons. Just like it cannot determine the values that guide our policymaking or how to prioritize one policy objective over another, it cannot tell us what time horizon we should care about. Disciplines such as ethics are potential sources to draw on when thinking about the right (social) discount rate. So is culture.
Now Let’s Think about why “The Problem of Social Cost” is so foundational
There are three main contributions in “The Problem of Social Cost” that have proven to be blockbuster ideas in political economy and, particularly, the study of environmental harm. At first blush, they may seem painfully simple and perhaps of limited appeal, but have actually made a big impact on sundry problems and puzzles. They have radically changed the study of externalities such as pollution and climate change and have weakened the hegemonic stranglehold that Arthur Pigou once had on these questions before 1960.
The central idea in Coase’s seminal article is that there is no such thing as a “victim” of an externality; across economic activities, harm may occur to parties who are not part of an exchange but that harm is always reciprocal.
For example, if a factory pollutes (implying that it has a property right to produce and pollute) this may create harm to the folks downstream who are exposed to its sounds, smells, effluents, and emissions. But, equally, if a factory is barred from polluting (implying that those affected by its pollution have the property right to breathe clean air), it may mean that it’s not producing products desired by consumers and not generating profits for its owners. The abatement of pollution therefore harms those consumers and producers. And that harm may be greater than the harm suffered by those affected by the factory’s pollution; in other words, in some cases there may be too little pollution – a point we will return to.
The second big idea is that, if three key conditions are satisfied, a system of market exchange may allow those who value reducing a spillover such as pollution to actually do so: to pay those who are polluting to stop. And, alternatively, it may allow those who value producing something that creates pollution (as a byproduct) to go ahead and do so: to compensate those who value a lack of pollution sufficiently, so that they will allow pollution to occur. Importantly, no matter what the original allocation of the property right (residents have the right to breathe fresh air versus a factory has a right to produce and thus pollute), you will end up with the exact same outcome: those who most value a given outcome will obtain it and, moreover, this will be the socially efficient outcome (the one that maximizes social welfare). In short, if pollution abatement is most highly valued by society, then the end result will be less pollution. But if products and their associated pollution are more highly valued, you will get pollution (or at least more of it). This is irrespective of the original allocation of property rights.
To elucidate this last point let us first return to our example before we spell out the three conditions.
First, assume that residents have the property right to breathe clean air, but a factory (and, by extension, its customers) value the products it produces highly (and, by implication, value the pollution sufficiently). If so, the factory owners will simply pay the residents for the property right, thus allowing their factory to produce and pollute. Both parties will be better off after engaging in this voluntary exchange, including the residents, since they will receive a payment above and beyond the monetary value they attach to breathing clean air. And society will be better off as well because resources will have been allocated to their most valued use.
Now assume the opposite is true: that the factory has the right to produce and pollute but the residents highly value clean air. If so, the residents will pay the factory to stop producing and polluting. Again, both parties will be better off after engaging in this voluntary exchange, including the factory owners, since they will receive a payment above and beyond the monetary value they attach to producing and polluting. And society will be better off as well because resources will have been allocated to their most valued use.
When does Coase believe that the magic of the market will allow these spillover problems to self-correct: allow agents to negotiate and reach a resolution on their own? Enter the three key conditions.
Condition 1: some folks must be willing to pay to stop the spillover. In other words, there must be a demand curve. To return to the example: if the factory has the property right to produce and pollute, the residents negatively affected by the pollution must be willing to pay to abate the harm. Alternatively, if the residents have the property right to breathe clean air, the factory owners must be willing to pay them to produce and pollute (compensate them for the damage).
Condition 2: property rights must be secure (assigned and enforced). This is logically required because, in essence, these parties are buying and selling from each other and thus transferring property rights to do certain things.
Condition 3: transaction costs must be low enough to allow the affected parties to negotiate in order to reach this resolution (to allow them to put a price on the harm and to pay to abate it — to buy the property right from each other). Spelling out these transaction costs and how they can be reduced has become a cottage industry in political economy. Among the most important are things such as standardizing weights and measures, units of exchange, language, and contracts. Establishing trust so that promises are credible is also paramount. Reducing incentives to misrepresent your willingness to pay for the property right under negotiation is also important.
The third big idea and the least understood one – but we would submit the most important – is that because in the real-world transaction costs are non-negligible and often quite prohibitive, it is important for the government to get the initial allocation of property rights correct. Assuming that the government’s goal is to maximize social welfare, then it must allocate these property rights to the user that suffers the most harm and that would have paid to have that harm reduced, but may be prohibited from doing so. The reason? Due to high transaction costs, said party may not have the opportunity to buy the property right from the offending party even though it’s willing to pay for it. And, thus, if the property right is originally assigned to the party that values it the least, social welfare will be harmed.
Let’s return to the example used above. Suppose the property right is allocated to the factory. It has the right to produce and pollute. However, there are affected residents who would pay to reduce the harm if they could (there is an “abatement demand curve” that represents the willingness to pay to stop the damage and it potentially intersects with the factory’s opportunity costs). Yet, because they cannot coordinate and negotiate as a unit (the transaction costs of doing so are simply too high), the affected residents are unable to reach a market (negotiated) solution with the factory to stop producing and polluting (or at least reduce the pollution somehow, perhaps by paying for it to install scrubbers).
This leads to a socially suboptimal outcome. In this case, it is better for society if the government redistributes the property right to pollute to the residents, since they value abating pollution more than the factory (its owners and customers) values producing and polluting.
And, moreover, Coase was agnostic about what tools the government should use. It might be, his professed favorite, litigation. But taxation is also possible. Or imminent domain. Or some other type of regulation. That is less important than righting the ship and transferring the property right in a way that makes society better off because it maximizes social welfare.
The implications are profound. First, this is a consequentialist view of markets and the role of government. If it’s about maximizing social welfare (and we believe that Coase’s own objectives were such and they are definitely the normative objectives of scores of political economists who follow him), then the government is responsible for engaging in constant cost benefit analysis. It must seek to understand who values particular outcomes the most and would be willing to pay for them. It must, by extension, help steer social resources to their highest value use by reallocating property rights. It also means trying to reduce transaction costs so that the government does not have to get involved in every single situation where there are social costs (reciprocal harms). Why? Because the government often lacks the information and technology to do the cost benefit analysis correctly and to reallocate the property rights correctly. And there are opportunity costs to the government’s time and resources.
One implication is to allow markets that were once missing to emerge, including markets for capping and trading carbon emissions, for example. This means satisfying the three conditions we mentioned above: making sure there is truly a demand (willingness to pay) to abate harm, assigning and enforcing property rights, and reducing transaction costs so parties can exchange the property right.
The last implication is counterintuitive from the Pigouvian perspective about externalities. In some cases, it means that the government allocates property rights to what seem like perpetrators instead of victims. Coase gives an instructive example at the end of “The Problem of Social Cost.” Residents who live near an airport may suffer damage from the noise or airplane exhaust associated with this facility; yet they might not be willing to pay the “airport beneficiaries” enough to mitigate the harm. The beneficiaries may get so much value from the ability to fly in and out of town that they would not accept whatever compensation is on offer. It would simply be too low vis-à-vis the benefits these fliers derive from using the airport. In this case, Coase would prescribe giving the airport the property right (to fly and pollute), not the residents complaining about its externalities. This brings to bear the fact that, for Coase, harm is always mutual and banning the airport will cause greater social harm to the potential victims of the ban (the travelers who derive the direct benefit and others who derive indirect benefits).
Conversely, for Pigou, it would not have been controversial to state that the residents are the obvious victims in this case. Ergo, the airport should be relocated or at least highly regulated. For Coase, the government would have to do the math first and figure out, after conducting a cost benefit analysis, who should merit the property right in terms of who values things the most (keeping versus relocating the airport) and, thus, what is the most efficient allocation of social resources. Sometimes that means giving airports and factories the right to pollute.
Let us summarize. If maximizing social welfare is our overriding goal, then markets are a possible means and always the best means if they exist. But so is government regulation if markets don’t exist. In those cases, governments have to be mindful about the initial allocation of property rights or their reallocation. This is because transaction costs may make it difficult for folks to achieve the most efficient allocation of resources through voluntary exchange. And in order to determine the initial allocation of property rights, governments should conduct a cost benefit analysis.
