Note: This blog has moved to https://substack.com/@governology. See my new posts there.
In my rather long previous post, I talked about the two fundamental market failures: externalities and globally sub-optimal Pareto optima. To understand this post, you really have to go back and read part 1.
Solutions to Market Failure
Because market failures are the only cases where the market can be improved upon, it stands to reason that the job of government is only to find solutions to those market failures, where feasible.
Solutions to Externality
Previously, I identified four necessary and sufficient conditions for an economy to tend toward Pareto optimality:
- Well defined property rights
- A sufficiently large number of potential market actors of a given type
- Actors motivated and able to act in their own self interest
- Actors that never reach their optimal utility (who could always be happier)
Item 1: Externalities are caused when property rights aren’t defined specifically enough, usually because of high costs involved in doing that. But most externalities that occur naturally (without the influence of government) can be solved by a government defining property rights. So government solution #1 is defining (or redefining) and upholding property rights – determining who owns what. This is the best solution because it augments the free market, providing individuals ways of protecting their own interest. Because its basically a decentralized solution, it’s very powerful. Individuals can bring up disputes with the government organically when they arise, rather than relying on a central agency to determine where problems exist.
However in some situations, it can be costly to define property rights for individuals. For example, in the pollution case, a polluter may be affecting millions of people only a tiny bit each. In such a case, it wouldn’t be worth the cost of each of those people enforcing their property rights. And so they let the externality happen. The costs aren’t “internalized” in the behavior of the polluter, and so more polluting happens than is socially optimal. This is where solution #2 comes in: taxes and subsidies.
While, for a given polluter, it may be infeasible to know which specific individuals were affected by their pollution, it is often a lot less costly to know how much total pollution that polluter produced. By using a tax on the polluter equal to the externality imposed per unit pollution, the cost can be “internalized” (the cost can be put on the producer of that pollution). Ideally, the money collected by those taxes should go back to the people harmed by the associated negative externality, and the money used to subsidize positive externalities should come from those that benefit from that externality, tho its almost never feasible to do that very precisely.
I should mention that reversing the effects of an externality might cost more than the externality itself. That’s why it’s not appropriate to set a pollution tax at the cost of cleaning up that pollution. The appropriate thing to do is find out the cost of keeping a unit of pollution and set the tax at that rate, rather than setting a tax at the cost of removing the pollution.
Item 2: When the motivations of market actors restrict entrants into a market as I talked about with the (usually misunderstood) case of monopolies and anti-competitive markets, this imposes an opportunity cost on current and potential competitors and customers. For a tax to solve a problem, it would have to be possible to determine how much opportunity cost the monopoly is imposing on others. Since the nature of innovation is that we don’t know what improvements to make or how much value they might have, there’s no way of accurately determining what such a tax should be. And even if potential value *could* be found, its unlikely a monopoly would find its way to the optimal solution on its own (ie without the power of a Darwinian competitive market).
So we have to resort to yet another form of government intervention: anti-trust market restructuring. If a government could identify a way to restructure the market such that an anti-competitive market was turned competitive, its possible that this government action would fix this opportunity externality. I should clarify that this is a very very specific type of regulation, and 99% of anti-trust government regulations are not of this type.
An example of the kind of market restructuring I’m talking about is electricity buy-back, also known as net-metering. Before net-metering, electric companies that owned their local power distribution network kept out any electric generators except large power plants. Net-metering has allowed anyone to participate in generating electricity which has opened up a huge amount of opportunity that has yet to be fully realized.
Item 3: As I mentioned last post, everyone is motivated to act in their own self interest. The only way for a person to act in a way not in their own best interests is if they have a lack of information or if they didn’t fully evaluate the information at hand. But beyond that, someone has to be *able* to act in their own self interest. For example, in the case of children, its clear they aren’t able to act in their own self interest, no matter how much they scream and cry about it. And so rules of guardianship are needed to protect their interests.
Rules of guardianship are very similar to property rights, in that they define how guardians are chosen for wards and what a guardian can and can’t do with their wards – ideally protecting wards from externalities imposed by their guardians. Parents are responsible for almost all actions of their kid, just as they’d be in the case of a horse or a computer they own. That guardian is also responsible for treating that child humanely, as is also the case for a horse they would own (and maybe someday, an intelligent computer).
Item 4 is always true in real markets and so no solution is necessary.
One more solution has a merit in an incredibly narrow set of circumstances: nationalization, where the government takes over the operations for some kind of product or service. I’ll argue in a later post that almost all cases of nationalization are likely to be worse than solutions that focus on creating correct incentives for private competition. But nationalization has real solid merit in the areas of policing and especially national defense for 2 reasons.
1. Police and armies are a fundamental part of how a government enforces its policies. Delegating those things to private companies carries a huge risk of those companies enforcing their own private policies rather than government policies – in the most extreme case, with a coup. History has shown that the risk of violent subversion is very real and I think provides a strong case that eating the inefficiency of government programs is well worth the peace of mind of having a nationalized police and military.
2. Also, for a private military company, no war means no money. Not necessarily so for a public military, and that’s a good thing. Private armies have an infamous history, from the White Company who regularly fought on both sides of a war and raided towns in times of peace during the 1300s and the Catalan Company which annexed and ruled over large sections of Greece for 75 years because of a payment dispute, to the more modern shady doings of Executive Outcomes in South Africa in the 1990s and Blackwater (aka Xe, aka Academi) in Iraq.
Even our public police and military have the problem of itchy trigger fingers, but the problem is bound to be far worse for private companies selling violence as a service. That isn’t to say that national armies and state police have a stellar history either tho.. I might write more about that in a later post.
Solutions to Sub-optimal Pareto Optimums
This last piece that needs a solution is the problem of how to get to the right Pareto optimum. While the market has no mechanism of moving from one Pareto optimal state to another on its own (other than random chance), the Second Fundamental Theorem of welfare economics proves that Pareto optimum in a given economic environment can be moved to with a “lump-sum” redistribution of wealth – meaning a redistribution that takes place all at once. And whenever the economic environment significantly changes, another lump-sum redistribution can correct things once again. This is where welfare programs come in. I should note here that the ‘welfare’ in “welfare economics” is different from the one in “welfare programs”. Welfare economics is essentially about maximizing the total happiness (utility) of the people, whereas welfare programs is about helping the poor monetarily, hopefully increasing their utility in the process.
The progressive tax system, where those who earn more pay a higher percentage in taxes than those who earn less, is one of the major ways the government implements a redistribution of wealth. Direct welfare programs, like social security, medicare, or food stamps are another common method which fill the need for wealth redistribution not provided by a progressive tax system.
However, there is a cost to wealth redistribution. As plenty of government policy makers have said, taxes work as a deterrent (a methodology for taxes not justified by economics). The same principle applies to taxing income, and so wealth redistribution will have a negative effect on productivity.
Redistribution programs are usually designed to help the poor, but its worth mentioning when a redistribution would be beneficial more generally. If your goal is to maximize the total happiness of everyone, there are unsettling theoretical scenarios that are possible. Like lets say the economy consisted of two people and their Pareto states looked like this:
In this scenario, its clear that state with the most total utility is C, where Person 1 gets pretty much all the happiness. Because Person 2 is so bad at being happy, if your goal is to maximize total happiness (utility), the solution here is to give Person 1 as much as possible, and consequently Person 2 as little as possible. If reality were at all close to this, this would put us all in a really awkward situation. Thankfully, there is strong evidence that this isn’t the way humans work and so we can avoid this kind of mess.[1][2][3][4]
I’ll have more to say about common government actions that aren’t justified in a later post, but this and the last post should make it clear how I’ve determined that my list of solution categories is complete.
The Appropriate Role of Government
Why don’t I pop the stack again and finally get back to the original goal of defining the role of government.
To summarize, another way of saying that the government’s role is to make things better for its people is to say that the role of government is to correct for chronic structural market inefficiencies, which I described in detail as market failures in part 1. One obvious and relatively well-understood market failure is externality, and the other is an inefficient initial distribution of wealth (which can be improved through wealth redistribution).
Free markets can’t uphold property rights, only governments can. And as I’ve mentioned in an earlier post, the dispute resolution process that is required to secure defined property rights was the genesis of government in the first place. Externalities cause market failures and the market has no mechanism of going beyond Pareto optimality.
Finally I can define the appropriate role of government clearly and specifically:
The role of government is only to:
- define and uphold property rights,
- define and uphold guardianship rules for those that aren’t able to act in their own self-interest,
- in situations where #1 and #2 are infeasible, correct for externalities with taxes or subsidies,
- in situations where a market can be restructured to allow more competition, restructure the market,
- provide policing and national defense services, and
- to reallocate wealth
when there’s a strong case that the costs of a government solution is significantly less than the value created or costs saved.
The requirement that the government solution be better than the market solution should be obvious, but is so often and tragically neglected that I think it needs to be in that definition. “Correcting” a market failure with a worse government solution is an error, not a correction. The mere existence of a market failure shouldn’t be reason to introduce government involvement. There needs to also be a plan that considers how much more efficient a government solution would be, including analysis of the costs of that government involvement.
Some Intellectual History in Support of this Definition
While I came to this specific definition via my own logic, reviewing iconic economist Milton Friedman’s thoughts on the subject seems to indicate that he would have agreed with most of my definition and logic. Friedman said that the classical liberal view, and also his view was that “the justification for government action is to prevent coercion and to promote voluntary cooperation among responsible individuals”.
He went further to say that this “leads to a very short list of basic functions which governments should undertake”:
- “First of all: to prevent one man from coercing another.”
- “Second: providing for external defense. These two are really part of the same. To prevent coercion from within. To prevent coercion from without.”
- “And beyond this, to promote voluntary cooperation among people, by defining the terms under which we are going to cooperate together, and by adjudicating disputes.”
- “To provide a substitute for voluntary cooperation, when such cooperation is.. not feasible. There are two classical cases which come under this heading. One is the case of technical monopoly. When for reasons of physical circumstance, it’s not possible to have competition. The essence of avoiding coercion is the availability of alternatives. … The more serious problem.. is the problem of what is called neighborhood effects, or if you want the jargon of economics, externalities.”
- “Another [function for government] is to protect irresponsible people. We can only really believe in freedom for responsible individuals. But a society includes irresponsible individuals, of whom there are two major classes; children, and the insane.”
Items 1 and 2 fall under the clear category of externalities and cover E in my definition. Item 3 is the definition of property rights (A). Item 4 falls under externality including the monopoly externality I mentioned last post (C and D). And item 5 is the definition of rules of guardianship (B).
He didn’t mention wealth redistribution (F) in this list, and I’ve discovered that he doesn’t believe there is an appropriate role for government in income redistribution. While he has always supported a negative income tax as a replacement for current welfare programs, his preferred method of helping the poor is through private charities. Certainly everyone would agree that it would be ideal if private individuals did voluntarily provided everything we need from government, free of charge. Certainly if everyone just decided to not steal or fight, and if there were enough people that would volunteer to put out fires, we wouldn’t need police or firemen. But its clear to me that the incentives lead to those kinds of areas being under-served without government involvement. This is an area it seems I disagree with Friedman.
Anyway, Friedman furthermore was careful careful to note the same qualification that government is only justified in attempting to fix market failure where that government solution is less costly than the market solution. Friedman noted that:
“The approach has been to regard any.. market failure.. as a sufficient excuse for government interventions. The market has failed, therefore the government should step in. But this is a basic error, because it involves a double standard. There is not only such a thing as a market failure, there is also such a thing as a government failure… The cure may be worse than the disease.”
I also believe Adam Smith would mostly agree with my definition. Smith wrote of three duties of government:
- “Protecting the society from the violence and invasion of other independent societies”
- “Protecting, as far as possible, every member of the society from the injustice or oppression of every other member of it, or the duty of establishing an exact administration of justice”
- “Erecting and maintaining those public institutions and those public works, which though they may be in the highest degree advantageous to a great society, are.. of such a nature, that the profit could never repay the expense to any individual, or small number of individuals; and which it, therefore, cannot be expected that any individual, or small number of individuals, should erect or maintain.”
Item 1 is correcting for clear negative externalities in the context of national defense (E), item 2 is correcting for externalities and upholding property rights (A and C), and item 3 is about correcting for positive externalities (C again).
Adam Smith didn’t include anything about monopolies since, in his day, monopolies were seen as always created by government, and the idea of a natural monopoly was only just being developed by John Stuart Mill. So there was no need in Smith’s mind for a government to do anything about something that didn’t exist. Today, natural monopolies are just as rare, but strangely imagined to be everywhere.
Smith also didn’t include the concepts of guardianship, but he lived in an era where he may well have considered children to be more or less the property of their parents. I disagree with Adam Smith’s third point that public works are appropriate for government, and I think he would have been convinced of this by modern theorems of welfare economics, which were rigorously proven only very recently, in the 1950s.
He also doesn’t mention anything about wealth redistribution in this list, but in Wealth of Nations, he says:
“No society can surely be flourishing and happy, of which the far greater part of the members are poor and miserable.”
I think he would have been happy to learn that the second fundamental theorem of welfare economics provides a rigorous theoretical basis for wealth redistribution in order to help those most in need.[4]
Reconnecting with 21st Century Government
After all these definitions and explanations, I should congratulate you for making it through all that. You’re really a trooper! Here, I want to bring us back from a mythical past or possible future to the modern day where we live in a complicated world. I want to show that the critical definition of the appropriate role of government isn’t very far off from what most of us consider good government.
I’ve already mentioned national defense and policing as protecting from the externalities of violence from outside the country and inside the country. Here are other examples of things the government currently does that fit its appropriate role:
Noise ordinances in any town protect people from undue noise at night so people can get some sleep. A noise complaint is justified because someone creating loud noise imposes a cost (an externality) on those around that person.
Providing public health information and public service announcements gives people information that private companies wouldn’t be likely to provide. This corrects for the positive externality that some kinds of information gives people. According to the methodology I’ve discussed, incentives for private companies to provide that information would be a better alternative, but the basic need of some kind of government intervention is there.
The fire department has some justification because fires can cause negative externalities. A house in one building is likely to spread to nearby buildings, and potentially entire towns (which has happened), unless put out quickly. Whether publicly or privately run, it seems warranted to subsidize it with public money.
Driver licensing is widespread as an attempt to ensure a certain level of safety on the road. Without licensing, the risk of injury while driving on the road or walking on the side of the road would be higher. This increased risk is a negative externality on everyone. How well this works is debatable, but it fits within the appropriate role of government if done right. Similarly, drunk driving laws can theoretically reduce the risk of danger on the road, ignoring the inaccuracies of most legal definitions of “drunk”.
Child protection services ensure that guardianship laws regarding children are upheld and that children are treated appropriately. This is also protecting against externalities, since the whole point is to ensure that the guardian isn’t mistreating the child – which is an externality on that child.
Gas taxes are partially used to pay for roads, which isn’t justified by my methodology. But part of that tax also accounts for the cost of the pollution that gasoline produces when used. The externality of the pollution can be corrected for with gas taxes.
Slavery in the US was without a doubt the most egregious thing the country has ever perpetrated. The rhetoric of the time was that black people were somehow inferior and thus “needed” to be a ward of a guardian – the owner. The problem with this rhetoric is that even if it were true (which of course it isn’t), the welfare of the slave was almost never in mind, and so the slave laws at the time allowed the negative externality of slave mistreatment to go nearly unchecked. While there were some laws on the books that theoretically protected slaves in certain states, these were only rarely enforced in cases like the hanging of William Pitman in response to his murder of his own slave in Virginia.
Regardless, like everyone else, black people are perfectly capable of acting in their own self interest, and thus any guardianship laws controlling them perpetrate a negative externality. The government role in defining more appropriate rule of guardianship (namely removing any for black people) certainly fits in the appropriate role of government. I should note that while government action took large steps in solving this problem, the problem began by government action as well – a government failure, not a market failure.
Now some of these things are done in inefficient ways, and I didn’t consider the government costs of any of these things, but these are all situations where government action might have a positive effect if done right. I’ll discuss things governments currently do that don’t fit in the role of government in a future post.
Conclusion
I’m going to repeat the appropriate role of government here because it’s probably the most important thing I’ll write in this entire blog and the basis for everything else I’ll write:
The role of government is only to:
- define and uphold property rights,
- define and uphold guardianship rules for those that aren’t able to act in their own self-interest,
- in situations where #1 and #2 are infeasible, correct for externalities with taxes or subsidies,
- in situations where a market can be restructured to allow more competition, restructure the market,
- provide policing and national defense services, and
- to reallocate wealth
when there’s a strong case that the costs of a government solution is significantly less than the value created or costs saved.
I can’t stress enough that understanding each specific part of this definition is critical for improving the governments of the world and creating new ones. I’ve shown a small slice of the intellectual history as well as my own logic supporting this definition, and I hope I was convincing to you, the most important reader.
In my next post, I’ll be talking about rights and freedoms using the role of government to guide us.
Note: This blog has moved to https://substack.com/@governology. See my new posts there.
Governology 
2 thoughts on “The Role of Government: Part 2”