Externality: Why Free Societies Also Need Coercion

Throughout this series, we have been discussing why society needs institutions. We talked about li (ritual propriety) as a social interaction protocol, “incentive compatibility” that aligns individual interests with institutional goals, and “procedural justice” that makes quality verifiable. Have you noticed one commonality among all these institutions?

They are all soft.

There is no coercion involved. Observing li can help you avoid many pointless conflicts, but you are free not to. An incentive-compatible system makes you voluntarily do what benefits it. Even a good market allows you to sell bad goods, as long as you choose to bear the consequences.

Soft institutions are beautiful institutions. Some libertarian intellectuals believe that the world should run solely on soft institutions, ideally without any coercion whatsoever.

Back when I was in college, I found it highly undignified when professors took attendance during class. If you lecture well, students will naturally come to listen. If they do not, it means your lecture is subpar. Forcing them to sit there by taking attendance—what kind of practice is that?

If all problems could be solved through markets and soft institutions, that would be wonderful.

However, in this post, we must state: free societies also need coercion. This involves a clash of minds among several genius economists and political scientists, and the origin of it all is called “externality.”

Pigou and the Ledger Mismatch #

The concept of externality was first systematized by British economist Arthur C. Pigou. His 1920 book, The Economics of Welfare, is the foundational work of the field [1].

Simply put, an externality occurs when the person taking an action does not fully bear the consequences of that action.

For example, imagine you live in a four-person university dorm. One of your roommates, Xiao Zhang, dislikes the canteen food, so he sets up his own stove in the dorm to cook.

Cooking benefits Xiao Zhang, but it also has a downside: oil smoke. The benefits are enjoyed entirely by Xiao Zhang, but the costs are not fully borne by him. Instead, a portion of the cost is offloaded to the rest of you, who must tolerate the smell of oil smoke. This offloaded portion is the “negative externality” Xiao Zhang’s cooking imposes on the dorm.

Externalities can also be positive. For instance, a child receives an education primarily for their own benefit—credentials lead to better jobs and a higher standard of living. However, objectively, a more educated population benefits the entire society: crime rates drop, higher incomes generate more tax revenue, and social welfare increases. Thus, even if you do not have children, you still want your neighbors’ kids to receive a good education. This means education has a “positive externality.”

A negative externality means I benefit while others pay; a positive externality means I pay while others benefit. Naturally, we want fewer actions that generate negative externalities and more actions that generate positive ones.

However, you did not charge Xiao Zhang a cleaning fee, and you did not pay your neighbors’ kids a thank-you fee for getting educated. As a result:

Bad things will be overproduced, because you do not bear the full cost of doing bad things.

Good things will be underproduced, because you cannot capture the full benefit of doing good things.

No villain is intentionally creating these problems. The fundamental cause of externalities is a ledger mismatch in the structure of interests: the private ledger does not match the social ledger.

Public Goods, Free Riding, and Collective Action #

Intuitively, the most rational solution to externalities is to align the ledgers: whoever generates a negative externality should pay a tax to compensate, and whoever enjoys a positive externality should pay a fee to sponsor it, right?

But reality is not that simple. Some ledgers are structurally impossible to align.

In 1954, Paul Samuelson mathematically defined a concept called “public goods” [2], which can be understood as a form of deliberately created positive externality. Examples include streetlights at night, basic scientific research, and national defense. Samuelson proposed that public goods possess two characteristics:

First, non-excludability: it is difficult to exclude someone from using it just because they did not pay. You might argue that everyone has a duty to sponsor science, but someone else will say, “Your Mars probe has nothing to do with me.” The only operable way is to make basic scientific achievements shared by all.

Second, non-rivalry: one person’s use does not diminish its availability to others. After all, the military cannot protect only a portion of the population; if there is peace, the entire nation enjoys it.

Therefore, Samuelson argued: the benefits of public goods are too large and diffuse to be charged selectively.

Another complication arises from Maryland economist Mancur Olson’s 1965 classic, The Logic of Collective Action [3], called “free riding.” It means: since I can enjoy the positive externality even if I do not pay, why should I pay?

Suppose you live in an old residential compound, and the homeowners want to install an elevator, requiring collections from door to door. In theory, if you want to use the elevator, you must pay. This seems straightforward.

However, a residential elevator cannot be charged per ride; once built, everyone can use it freely. Thus, some residents will think: even if I refuse to pay, won’t you build the elevator anyway? Once it is built, can you really forbid me from using it?

So, my rational choice is not to pay.

If everyone is this rational, the elevator will never be installed.

Consequently, although the elevator benefits everyone, it never gets built.

This is Olson’s counter-intuitive insight: common interest is not a sufficient condition for collective action.

In fact, the larger the group, the easier free riding becomes, and the less likely they are to act for the common interest.

Both externalities and information asymmetry are market failures. When bad things are overproduced and good things are underproduced, and everyone agrees something should be done but no one does it—can such a society thrive?

It seems the only way is for a strong entity to step forward and say, “I have made the call; this must be done. Everyone pays their share, in money or labor. Whoever refuses will be penalized.”

In practice, this means the government forces everyone to pay taxes, and then the government cleans up pollution and provides public goods. This was Pigou’s recommendation. In the eyes of economists, correcting externalities is the most legitimate reason for government taxation, known as a “Pigouvian tax.”

The Coase Theorem and the Price of Property Rights #

As you can imagine, economists generally dislike coercion. Coercion indicates that the system is not fully incentive-compatible. Stumbling into force means the scholar’s reasoning has failed. Coercion is inelegant.

But you might ask: isn’t there a “revelation principle”? Didn’t we say that any problem could potentially be resolved through incentive compatibility? Actually, the revelation principle states that a mechanism can induce everyone to report private information truthfully, but it does not guarantee that a costless solution exists.

Yet, economists did find an ingenious solution, courtesy of the great Ronald Coase. Coase’s 1960 paper, “The Problem of Social Cost,” is one of the most cited papers in all of economics [4], for which he was awarded the 1991 Nobel Prize in Economic Sciences.

Coase’s solution was property rights.

He imagined a scenario: a cattle rancher and a wheat farmer are neighbors. The cattle frequently wander into the wheat field and damage the crops. This is a classic negative externality. What should be done?

Pigou would say: the government must intervene, either by banning cattle ranching or imposing a tax.

Coase asked: why can’t they negotiate between themselves?

If the cattle bring the rancher $100 in profit but damage $150 worth of wheat, the farmer can pay the rancher $120 to stop ranching. If the cattle bring the rancher $200 in profit while damaging only $50 worth of wheat, the rancher can pay the farmer $80 to compensate for the damage.

As long as they can negotiate and trade, both parties will end up better off than before, without any coercion. Isn’t that elegant? Externalities do not require government intervention; private property rights plus free negotiation solve the problem automatically.

But in reality, things are rarely so simple. Note that the Coase Theorem relies on two assumptions: clear property rights and zero transaction costs.

In the real world, who owns the air and the water? There are millions of people downstream; how do you negotiate with each one? Transaction costs are often so high that negotiation is impossible.

More importantly, where do property rights come from? Who defines them? Who registers them? Who enforces the rulings? Who prevents the strong from directly seizing the property of the weak?

You will find a paradox: property rights themselves are a public good. A social order capable of protecting property rights must be built, maintained, and enforced by coercion—which drags the problem right back to coercion.

This does not mean the Coase Theorem is a failure. Coase made it clear in his 1991 Nobel lecture: the Coase Theorem was a thought experiment meant to make economists realize that in the real world, transaction costs are always positive, making the choice of institutions critical [5].

Coase’s true legacy is not “property rights are all-powerful,” but: before choosing an institution, calculate the transaction costs.

The Tragedy of the Commons and Community Governance #

Simply put, Coase left a back door: as long as transaction costs are low enough, negotiation can solve the problem, and coercion can be avoided.

However, economists had grown almost despairing of reality. In 1968, Garrett Hardin published his famous paper, “The Tragedy of the Commons” [6], presenting a parable: a public pasture open to all herdsmen. Adding one more cow brings all the profit to you, while the cost of overgrazing is shared by everyone. This calculation always favors adding more cattle, leading herdsmen to continually add stock until the pasture is ruined.

The tragedy of the commons is the mirror image of Samuelson’s public goods: public goods are “wanting a positive externality but failing to get it,” while the tragedy of the commons is “wanting to remove a negative externality but failing to do so.” Coase’s solution is almost bound to fail here: there are not two parties, but hundreds or thousands, and everyone has an incentive to free-ride. How do they negotiate?

Consequently, mainstream economics accepted a dichotomy: when dealing with public resources, either the government takes over, or they are privatized. There was no third way.

It was then that a female political scientist, Elinor Ostrom, stepped forward.

Ostrom did not deduce theories in a study; instead, she led her students to travel the world to see how local communities actually managed common resources.

They discovered that countless small communities worldwide successfully managed common resources for generations, or even thousands of years, without government coercion or privatization. How did they do it?

For example, Törbel is a small mountain village in Switzerland where villagers share a high mountain pasture—a classic commons. Yet since the fifteenth century, the villagers have maintained a collective contract regarding water channels and pastures. The most ingenious rule in the contract states:

“No citizen could send more cows to the alp than he could support during the winter.”

What does this mean? Alpine cattle farming is split into summer and winter. In the summer, cows graze on the public pasture for free, requiring almost no individual cost. In the winter, there is no grass on the mountain; cows must return to individual barns and eat hay harvested or purchased by the owner, a cost borne entirely by the individual. Without this rule, some villagers who could only support five cows in winter might buy five extra cows in summer to graze for free, selling them before winter. This rule effectively limits summer public grazing to the owner’s winter capacity.

Its beauty lies in turning a hard-to-monitor metric (“how much grass does Xiao Zhang’s cow eat in summer”) into an easy-to-monitor metric (“count how many cows Xiao Zhang has in winter”). Thus, simple peer monitoring among neighbors suffices, without government intervention.

As a result, Törbel avoided the tragedy of the commons for centuries.

There are many similar stories. Ostrom detailed these findings in her 1990 book, Governing the Commons [7], for which she received the 2009 Nobel Prize in Economic Sciences.

It turns out community self-governance can resolve externalities to a large extent! Ostrom summarized 8 design principles for self-governing institutions:

1. Clear boundaries (without “us,” there is no community of responsibility);

2. Rules must fit local conditions from the bottom up;

3. Those affected by the rules can participate in modifying them;

4. Monitoring must be internal, ideally by the users themselves;

5. Sanctions must be graduated; immediate heavy penalties destroy cooperation;

6. Conflict resolution must be cheap and accessible;

7. Higher-level authorities must at least recognize the community’s right to self-govern;

8. Complex problems must be addressed through nested enterprises.

Crucially, not a single principle suggests running to the government when things fail. This aligns closely with traditional rural China. As sociologist Fei Xiaotong wrote in From the Soil, disputes in rural communities were resolved locally through negotiation rather than going to court. This was called wu song (no litigation).

Ostrom’s school of thought is known as “polycentric governance.”

Nested Governance: Engineering Elevator Installation #

These principles represent an engineering solution that is highly actionable. Returning to the example of installing a building elevator, we can apply them as follows:

First, the homeowners’ committee should not treat the elevator installation as a compound-wide project. Instead, adopt a “one unit installs, one unit resolves” model. This corresponds to Ostrom’s Principle 1: defining the unit community.

Second, do not share costs equally. First-floor residents who do not use the elevator pay little or nothing; residents on higher floors pay more. This matches Principle 2: rules must fit local conditions.

Next, allow all affected parties to jointly draft the rules (Principle 3); and the project accounts must be monitored by the users (Principle 4). According to Principle 5, implement graduated consequences for free riders: do not immediately cut ties; instead, remind, then notify, and temporarily suspend elevator access only as a last resort, restoring it once payments are caught up.

Principle 6 means disputes can be mediated by the homeowners’ committee. Finally, Principle 7 requires recognition from higher authorities, meaning obtaining approval from the subdistrict office and neighborhood committee beforehand.

Ostrom’s Principle 8, “nested enterprises,” means each layer manages what it is best suited to manage. Unit residents decide unit elevator details; the homeowners’ committee coordinates procedures and compares vendors; and the neighborhood committee/subdistrict handles legality and safety approvals.

The core idea is that because local neighbors interact daily, they form a community of mutual monitoring. Therefore, we can negotiate solutions ourselves without invoking government coercion.

The Place for Ultimate Coercion #

Ostrom showed that small communities can self-govern, but this does not mean society no longer needs coercion. When groups become too large, members anonymous, externalities trans-regional, and conflicts threaten to escalate into violence, we ultimately need an organization with final coercive power to step in.

That organization is the government. We will discuss this in the next lecture.

References: #

[1] Pigou, Arthur C. The Economics of Welfare. London: Macmillan, 1920.

[2] Samuelson, Paul A. “The Pure Theory of Public Expenditure.” The Review of Economics and Statistics 36, no. 4 (1954): 387–89.

[3] Olson, Mancur. The Logic of Collective Action: Public Goods and the Theory of Groups. Cambridge, MA: Harvard University Press, 1965.

[4] Coase, Ronald H. “The Problem of Social Cost.” Journal of Law and Economics 3 (1960): 1–44.

[5] Coase, Ronald H. “The Institutional Structure of Production.” Nobel Prize Lecture, December 9, 1991.

[6] Hardin, Garrett. “The Tragedy of the Commons.” Science 162, no. 3859 (1968): 1243–48.

[7] Ostrom, Elinor. Governing the Commons: The Evolution of Institutions for Collective Action. Cambridge: Cambridge University Press, 1990.