Coase, firms, and governments: two important takeaways


I’ve been thinking about a couple classic papers by Ronald Coase (Nobel Prize in Economics, 1991) lately. Here are some ideas that I think are important.


There are transaction costs associated with participation in markets. Firms (companies) provide a way to reduce those transaction costs, but at the cost of higher inefficiencies that come with non-market solutions. Lowering transaction costs and creating more efficient markets reduces the need for firms, and will lead to changes in how people exchange goods and services.

At one time there were very high transaction costs associated with connecting people who needed rides with people who could drive them around. If I needed a ride, I could call a bunch of people I knew, find out who was closest to me, who was able to pick me up, and give them my address. The transaction costs (knowing people, calling around) were high.

Taxi companies came along to reduce those transaction costs. In a world with taxi companies, I call one taxi company and that call is broadcast simultaneously by a dispatcher to all of the company’s taxi drivers. The closest available driver will respond and pick me up.

Today, thanks to the internet and GPS, marketplaces have been created (e.g., Uber) that drastically lower the transaction costs of participating in the passenger/driver market. Anyone can easily and quickly find a ride or a passenger without the need for the dispatching company.

This is a very powerful trend: as transaction costs are lowered, the exchanges of goods and services that were once controlled end-to-end by large companies can now be facilitated more efficiently by smaller, market-making companies.

Coase discussed the relationship between transaction costs and firms in his 1937 paper ‘The Nature of the Firm’.


Governments are firms. One of their primary roles is allocating goods and services with high related transaction costs (e.g., infrastructure, welfare, schooling). Many people think that only very large firms with unique powers (governments) are suitable for that task. However, as transaction costs fall, more efficient markets could change how these high-transaction-cost goods and services are exchanged as well.

At one time there were very high transaction costs associated with connecting people who drove cars with roads on which they could drive them. If I wanted to get somewhere, I would have to find out which roads I would be allowed to drive on, and I would have to stop and pay a fee to the owner of each road on my route. The transaction costs (knowing which roads were open for business, stopping, paying) were high.

Many people were driving cars and needed roads to drive them on, so we got governments involved to reduce those transactions costs. In a world of government-controlled roads, I can drive on almost any road without worrying about who to pay.

Today, online maps, transponders, and licence plate cameras have drastically lowered the transaction costs of participating in a private driver/road market. Anyone with an internet connection and a transponder (or even just a license plate) can identify private roads on their route and efficiently pay for the use of those roads without going through their government.

This is a very powerful trend: when transaction costs are lowered, the exchange of goods and services that we once required governments to oversee can be done more efficiently between private corporations and individuals.

Coase briefly discussed the idea of governments-as-firms in his 1960 paper ‘The Problem of Social Cost’.


Hopefully these examples get you thinking about what is possible in a world of low transaction costs. There are many advantages to allocating goods and services through markets as opposed to firms, and we should be looking for opportunities to do just that: not only with corporations, but with governments as well.

8 thoughts on “Coase, firms, and governments: two important takeaways

  1. Nice article Carter! Do you think that lowering transaction costs could eventually lead to your libertarian paradise? I could see this sort of working for education. But I feel like taking the government out of these types of things would make the poor (and especially people with poor money management skills) worse off. Like if a person working a low paying job has to take a private road to work, but can’t afford it (either overall or for a particular day), what would happen? And what would stop people who own these roads from really driving up prices and thus drastically decreasing opportunity for those who can’t afford it?

    1. Ha, yes, I was partly trying to suggest that we could reduce the size of governments given that many of their roles as firms can be performed by smaller private firms thanks to lower transaction costs. And smaller governments are definitely part of any classical liberal’s ideal society.

      Regarding road access for the poor. Right now governments are in charge of roads, and they are in charge of wealth redistribution. Because governments pool together the taxes they receive (taking more money from some people than others) and then pay for things (like roads), the two roles kind of get intertwined. If a government builds a road that both me and you are allowed to use, there is some implicit act of wealth redistribution there (from the one of us who pays more taxes to the one of us who pays less).

      But! It is possible (and I would argue, preferable) to separate those two roles of the government. Let there be private roads. Let the price required to drive on those roads be determined by the market. Then, let us provide poor people enough money to cover that price (through taxes or whatever). That arrangement will be more efficient than the one described in the previous paragraph. We don’t have to choose between efficient markets and welfare: we can have them both.

      Regarding your last question: the free market would be the only thing stopping road owners from driving up prices. It sounds scary, but it’s not.

  2. While I think all roads should be privatized for the reasons you explain, there are two reason why they won’t: 1)The high switching costs of the consumer and 2) The large upfront capital investment of the private business owner.

    What are the switching costs of going from taxis to Uber? You might think, “not feeling comfortable with a stranger driving me around.” But fortunately for Uber, I think people have traditionally had bad experiences with Taxi drivers anyway, so people aren’t actually too apprehensive about switching to an unfamiliar system especially when they can read a short bio and few customer reviews before getting the car with an Uber driver. Plus taxi rates are more expensive. Much more expensive. Switching from Taxis to Uber works in large part because, from the customer’s point of view, it’s easy to leave taxis.
    The capital investment of an Uber driver is zero. They already have a car, and signing up to be an Uber driver is free. Building the Uber software and network wasn’t that expensive either, especially when the company starting generating revenue right away.

    Now lets look at roads.
    Let’s first assume that all government roads should be privatized so that transaction costs are lower, making road use cheaper and more efficient. I think this should happen. The reason why this is unlikely to happen is first because of extremely high switching costs. Let’s suppose we would pay $10 less in taxes if the government no longer built roads and sold all it’s current roads to private companies. And let’s suppose it only costs us $5 to use the road in a purely private-run road system. Let’s further assume that people would experience far less gridlock traffic during rush hours. And if you wanted to experience absolutely no traffic you could take the extra expensive road that will ensure you get home in time for dinner at 5:15pm. In this case let’s suppose you’re still only paying $8, instead of $10. Furthermore, private road companies would develop new transportation technology in order to stay competitive, encouraging people to find even more efficient and low cost ways to travel than we can think of currently.

    But here’s the problem: Nobody is going to care that they pay $5 instead of $10, because either way you’re still going to have to pay taxes. Paying marginally less taxes so that you can pay every time you use a road will end up feeling like (whether it’s true or not) paying the same amount of taxes while also having to pay for roads. It would only be worth the switch if grid-lock traffic during rush hour actually ended. The only way that is going to happen is if there are new roads built. In this scenario, the new road would be built by a private company. A large risky upfront capital investment that can only make a return if all roads are private so it’s only competing with other privately run roads who are charging a market equilibrium price (as opposed to competing with a free government road). Let’s assume that it will in fact make a return in the long run. No more gridlock for the rich and cheaper road prices for everyone. But getting there is not like writing a few lines of code to connect drivers with passenger.

    I’m not saying it’s a bad idea or that it wouldn’t be a good investment. I’m just saying it’s different enough from something like Uber, that it’s likely never to happen.

    One last thought. Auri mentioned that if everyone had to pay for roads, then the poor would be screwed. Because, lets remember, they don’t really pay taxes and they drive on the roads for free. But think about our current system where even the poor have to pay for their car, their insurance and their gas. And there are in fact people who are too poor to afford one of all of those things so they just can’t get around by car. Do people care? No. Nobody will every say that the government should control the car, gas or insurance industries so that the poor can drive cars for free. Nobody. Will Ever. Say that.
    Also, I do think that the risky upfront capital investment of a private company building new roads is probably similar to a gas company building gas lines around the world. And gas companies were incredibly good investments.

    Which makes me feel the need to throw in one more point. The utility grid (basically all the utilities that flow into your home through some pipe or wire) were built by a private company, but new private companies can’t just go install a new power grid right beside the existing grid in order to create competition and keep prices down. Well, actually, back when utility grids were first being built, multiple utility companies tried to build grids next to each other and it in many cases, it turned into a mess. You can literally imagine having half a dozen telephone poles all running phone lines to your home so that you can choose one over the another. Will it be any different with a road next to an existing road next to an existing road and so on. In many cases this just wouldn’t make any sense. So in some cases one company is likely to own the only road from point A to point B. It becomes a similar situation to the utility grids. Well what did we do with the utility grids? We regulated them. In some cases, the government says, “nobody else can build a grid here and the company that is already here must charge this price.” Some are regulated differently than other, and these different grid systems might provide great case studies (check this out!: ) for how to think about roads. The point is that privatizing roads has some serious challenges unique from something like Uber.

    In the end I think it’s the perceived switching costs of consumers combined with the large capital investment that make privatizing roads a long shot. Plus there is the challenge of good competition for competitive pricing when several competing road companies might be impractical.

    After all is said and done we can only be sure of one thing: Poor people would probably be worse off because they’d have to pay for one more utility and the money the government would earn from selling off all the roads probably wouldn’t go to welfare. It would go to carving The Trump’s head into Mount Rushmore.

    1. Regarding switching costs. Let’s not forget that some people will have very negative switching costs! In a world of privatized roads, those who don’t use roads get to pocket the money they used to be paying in road taxes. So the aggregate switching costs might not be as high as you think.

      Nice little rundown of the benefits of privatized roads!

      Regarding high barriers of entry to the privatized road industry. Agreed, and I also agree that that doesn’t necessarily make it a bad investment. People take on projects with much higher capital investments than a road (gas companies, as you suggest, also anything Elon Musk does). Also, there actually are plenty of privatized (at least privatized-ish) roads in the world. There’s a huge one in TO (

      I’ve addressed the roads for poor people idea in my comment on Auri’s post.

      There are indeed some unique challenges to large scale infrastructure projects. Your comparison to electric utilities is a good one. But! Having half a dozen electrical poles all running to my house doesn’t sound like a terrible thing. Realistically, once there are two electrical poles coming to my house the price of electricity will already be so low (because of the competition between the two electric companies) that it won’t be profitable for another electric company to invest in another pole to my house. Even more realistically, the first electric company to build a pole will build in a little extra capacity on their pole to rent out to competitors anticipating that if they don’t, then a competitor will build their own pole, increase competition, and bring down the cost of electricity anyway!

      The power of free markets can almost always overcome the types of fears you’ve expressed about the road and electric utility industries.

      To address your last paragraph. Yes, poor people would likely be worse off unless the implicit wealth redistribution inherent to government-controlled industries were provided through more direct means. Which of course could be done. It’s worth mentioning that absent that, a privatized road system would still certainly benefit the average person (who is neither a net benefactor nor beneficiary of welfare), and perhaps even the poor person (if the efficiencies gained by the free-market solution were great enough).

  3. This is a great conversation.

    My first inclination is to be sceptical of TRF.

    Like lots of things, roads should mostly remain government controlled for many of the reasons talked about above. That is, until technology progresses in a way that makes road building cheap, fast, and safe.

    I’ll list a few more problems and I’d love some rebuttles.

    We depends on roads for more than just transportation, which is only upfront cost people care about; getting from A to B. You may want to invest heavily in the road you take to work to get there faster and safer. However, you can’t factor in when you might need another road, say the quickest route to a hospital when you are in an ambulance. How would you incentivize private companies to provide effective public transit to publicly funded locations? Cops and forefighters and garbage men use the roads as well.

    Private companies don’t own the land of a city so they don’t have the building rights to construct over or under other pieces of land. This opens the door to tons of legislation, which means government bureaucracy is increased further. It could be a libertarian nightmare; open the door to privatization while bolstering policies to keep those companies under control.

    Efficient roadways have huge implications for the economy in general, such as transporting resources, trucking, and even mail. You can’t really leave this in the hands of a private company as an experiment.

    What happens when a private company abandons a road?

    What happens when a private company makes payments for a road so high because so few people use it that it becomes a detriment to use it, and reduces economic growth in that area?

    What happens when a private company lobbies to bulldoze a park or ocean view to put a highway in? Are they going to listen to the residents of the city or their most popular users?

    I think as it stands the only way it would work is if one large company lobbied a government to take over all the roads in that region. I think you would still be left with the same result.

    I think the barrier to entry is too high and the cost of failure is too much for any modern government to ever give up control. The only way this will chance is if the face of transportation and roads in general changes, like robots or solar powered roads, when we transition to drone traffic for mail and packages, and when autonomous vehicles make the existing gridwork more efficient. Until then, I think road privatization will be reserved for niche scenarios.

    The problem I see with a libertarian dream city is that it forgoes good design for cheapest alternative, which isn’t always the best product, and doesn’t always lead to the best outcome. The design of a city is responsible for the efficiency of the city, the balance and happiness of the residents, and distributing resources fairly throughout the city. This is necessary to maintain a balance of people wanting to live there, and supporting blue collar folk enough that they want to stay and do rich people’s shitty work. A privatized company doesn’t take all these factors into consideration, whether it be roads or buildings or anything really. The only metric for a privatized company is profit, which isn’t the role of infrastructure.

    Here’s a final anecdote. I think Uber and Air BNB are better alternative NOT because they are cheaper, but because they are more elegant and more useful technology. They attract buyers because the interaction is instantaneous and satisfying and tangible. But roads are boring, roads haven’t really changed, and roads won’t change for the foreseeable future. For me, a company trying to get into the road building business is about as forward thinking as the Holiday Inn making an app to try and compete with Air BnB.

    1. How would you incentivize private companies to provide public transit to publicly funded locations?

      As a homeowner, why would I ever buy a house on a road that doesn’t connect to other roads that lead to a hospital? As a road builder, why would I ever build a road that doesn’t connect to other roads that lead to a hospital? No one would use that road. No one would pay for that road. No one would pay to live on that road. Individuals looking out for their own interests leads to emergent phenomena like roads being connected to each other. No top down planning required.

      Private companies don’t the land of a city and don’t have the building rights to build over or under other pieces of land, this opens the door to tons of legislation, which means government bureaucracy is increased even further.

      Most land in cities is privately owned. The parts that aren’t can be sold off. We can also do away with government-enforced zoning regulations.

      Efficient roadways have huge implications for the economy in general, such as transporting resources, trucking, and even mail. You can’t really leave this in the hands of a private company as an experiment without expecting it to get costly.

      If efficient roadways have huge implications for the economy in general, then the economy in general will pay for them. I would pay an explicit fee (likely through my courier who is charged a fee to drive on a road) for a package to be delivered to me. Remember, we already pay this fee indirectly through taxes.

      What happens when a private company abandons a road?

      It sits there abandoned until the land owner decides to do something else with it. If the best use of that land (as determined by the market) is a road, someone will build a road there.

      What happens when a private company makes payments for a road so high because so few people use it that it becomes a detriment to use it, and reduces economic growth in that are, not to mention farming.

      Then no one will use it, the road owner will lose money, and he’ll lower the fee so that people do use it. The people in the area will find alternative routes to use in the meantime, and the ones who really need the road will pay the high fee.

      What happens when a private company lobbies to bulldoze a park or ocean view to put a highway in?

      Then the owner of the park (could be a group of people) sells the land and lets the company build the road. Or if the park owner really likes the park, he doesn’t sell the land.

      The only way this will chance is if the face of transportation and roads in general changes, like robots that make solar powered roads, when we transition to drone traffic for mail and packages, and when autonomous vehicles make the existing grid work a lot more efficient

      The fastest way to get these types of changes is to privatize roads. Competition drives all sorts of innovation.

      The problem I see with a libertarian city is that it forgoes good design for cheapest alternative, which isn’t always the best product, and doesn’t always lead to the best outcome

      There are plenty of examples of well-designed private infrastructure. When a company develops land for a suburban neighborhood it is responsible for designing where roads go as well as building and maintaining those roads. If the design is good (roads going to all the houses, roads connecting with other major roads), the developer will be able to charge a higher price for homes in the neighborhood (so he is incentivized to employ good design, not just the cheapest alternative). Planning and good design doesn’t go out the window with privatized roads, it’s just done in a more competitive environment by people other than the government.

      1. I think the only answer you gave that I wasn’t really satisfied with was the one about the ocean view. Many would argue that we don’t want that decision to be made by an individual owner of the land. Many would say that that decision is best made by the community that all enjoy the ocean view. There are lots of real life examples of “ocean views” (national forests and parks, literal ocean views, wildlife reserves, etc) that would have been torn down and replaced by a private enterprise (possibly a road) if the government didn’t protect them. This issue might be a little off the core argument of privatizing roads, but I think it’s a problem of the libertarian view of these types of things.

  4. Public goods and negative externalities pose difficult problems for libertarians and non-libertarians alike. Coase’s paper “The Problem of Social Cost” does a great job of framing the issue.

    Might be good to go through the ocean view example.

    The scenario: A farmer owns a stretch of land along the coast of an ocean. There are 100 houses up on a hill that look over that land and have a beautiful ocean view. The farmer is going to build a freeway on his land that will ruin the beautiful ocean view.

    Libertarians like to look at things through the lens of property rights. If you own land you have property rights to that land. You don’t have property rights to lines of sight or views. So the farmer has a right to build the freeway. But of course, if the homeowners really like their view they could get together and pay the farmer to not build the freeway. If the price they offer is close enough to the economic benefit of the freeway, he might accept. Either way, the full economic benefit of the land will be realized.

    But let’s say we live in a world where you can claim property rights to lines of sight or views. In that case, the property rights of the homeowners and the property rights of the farmer are in conflict. An arbitrator would be needed to determine which party is being harmed more by the enforcement of the other party’s rights (a judge would likely make a ruling on the issue). (Actually Coase’s famous insight in “The Problem of Social Cost” is that with low enough transaction costs, a judge may not be needed at all in this scenario).

    Now let’s go back to the world we live in where there are property rights to land but not lines of sight. Let’s say the homeowners all get together and vote for a mayor of their town who promises to enact legislation that prohibits unsightly structures from being built along the coast. If the mayor gets elected, the rights of the farmer will be trumped by the desires of a majority. The full economic benefit of the land will not be realized: we won’t know what value the homeowners would have placed on their ocean view, so we cannot determine if maintaining the ocean view is the most economically beneficial use of the land. In addition, it is not a morally desirable outcome to allow special interests groups (like the homeowners here, or oil companies, or car companies, or seniors, or unions) to act through a government to determine what non-consenting people should do with their property (which includes their money).

    Following libertarian principles would lead to either one of the first two solutions, but definitely not the last.

    There are other far more difficult scenarios than this one when it comes to public goods and negative externalities: does someone have property rights to water flowing through their yard? How about to the air they breathe? These are tough issues and there is no economic or political magic bullet. But I think that if we follow classical liberal principles (which largely boil down to “don’t coerce people”) when faced with these types of situations, we arrive at the most economically and morally desirable outcomes.

Leave a Reply

Your email address will not be published.