Libertarianism Under Scrutiny

Responding to Objections to Libertarianism and Offering My Own

The Increments podcast recently critiqued libertarianism in a series of episodes. Here, I respond to their criticisms and offer my own.

Libertarianism 101

The introductory version of libertarianism asserts that markets are efficient and, hence, effective at solving most economic problems. Want to provide people with food? Have a free market for food so that private suppliers, like supermarkets, can provide consumers with groceries. This works well because supermarkets are competitive, incentivising them to reduce prices and serve their customers better.

This logic fails whenever a market is not competitive, such as in a monopoly market, or when there is an externality, in which case costs to a third party are not considered by producers. These are instances of so-called 'market failures' (a term that is slightly misleading, as it applies not just to markets but to group decision-making more generally, as we’ll see below).

Thus, critics of libertarianism can retort that there is a place for government: the government should regulate markets to ensure they are competitive and that there are no externalities. According to these critics, this is what legitimises government intervention. There goes the naive libertarian account…

Or does it? One issue with these anti-libertarian criticisms is that they work well in theory, but only if one assumes that a government has all the relevant knowledge about the markets it seeks to regulate. For instance, consider a monopoly market. A monopoly is inefficient because it can charge more for a particular good than is Pareto optimal. So, the government might force the monopoly to charge a lower price to solve this inefficiency.

That seems reasonable enough in theory, except if the government sets the price too low, it could bankrupt the monopoly, leading to greater inefficiency than the monopoly itself does. Hence, the government should gather information about the monopoly market, specifically about the monopoly's cost of production, to set appropriate prices. The problem is that the government can only obtain such information by asking the monopoly for it, and the monopoly may be reluctant to provide it. For example, the monopoly might temporarily lower production to increase its average production costs per unit, allowing it to argue that it is already charging a reasonable price and that further reducing its price would bankrupt it. As Gruber notes in his lecture on monopolies, we do not live in a world where one can obtain a market's real-world supply-demand curve at the press of a button. Such information has to be gathered at a cost.

These interventions also presume the government has rightly determined that a particular sector of the economy is a monopoly market, which is not so easily done. If a small town has only one cinema, is it a monopoly, and should its prices be subject to price ceilings? One argument against considering it a monopoly is that this slicing up of the town's economy might be artificial. A more natural market might be 'entertainment options in the town', which all compete with one another, even if only indirectly.¹ A real-world example of this is the antitrust case against Microsoft in 1998, where Microsoft was regulated by the United States Department of Justice (DOJ) for bundling Internet Explorer with Windows, thereby reducing competition for other browsers. Critics of this antitrust case argued that the DOJ focused too narrowly on the market for Intel-compatible PC operating systems, which made Microsoft seem more dominant than it really was.

A related issue is creating good antitrust laws that apply in a way that doesn't disincentivise companies from operating in a particular market for fear of unduly becoming subject to such laws.

This is not to say that the government cannot regulate markets to positive effect, but such regulations are not straightforward, easy to implement, or cost-free. Hence, the more important point is that it is easy for the government to introduce more substantial inefficiencies in trying to solve inefficiencies in the private sector. This is why the more sophisticated version of libertarianism is that although markets are not perfectly efficient, the government is even less so.

Government Inefficiencies

Government is inherently a clumsy device that is not well-designed to perform many of the myriad tasks people want it to do, let alone all of them jointly. One of the main inefficiencies of government is due to voting systems. For a government to implement good policies, voters should have incentives to vote for good policies. In an ideal world where information about a policy and its results is easily obtainable, voters would have such incentives. However, if obtaining that information is costly (e.g. if it requires reading a 300-page explanation of the relevant policy), then being informed might be more costly than it is worth. In this case, rational voters will rightly save themselves time and forgo both the cost and the benefit of informing themselves.

In practice, this means bad laws will get passed because it is too costly for voters to ascertain that particular laws are bad. This is one reason public choice theory provides for why modern governments have hard-to-get-rid-of inefficiencies. Moreover, it is an example of a 'market failure' in government, as it is a case where individuals behaving rationally (by deciding that certain costs are not worth it) produce irrational outcomes at the level of the group (i.e. bad laws).

These inefficiencies become more pronounced in specific voting systems. For instance, proportional representation makes it harder to decide which party to vote for because it makes it hard to predict how one's vote will influence the formation of a (coalition) government. Hence, it is hard to predict what policies will be put into practice as a consequence of how one votes, which is another instance in which relevant information is hard to obtain.

One way in which these problems could be resolved is if the cost of moving to another country was close to zero: in that case, one could move to whatever country had the best policies with zero effort, leaving countries with bad policies without taxpayers. However, the cost of leaving one's country of residence is relatively high, so if one wants to avoid a particularly bad policy, one will only do this if the cost of moving is less than the cost of that policy. This inertia is another reason government policy is worse than it would otherwise be. If people could easily move to another country, governments would be more incentivised to implement better policies to retain their tax-paying citizens.

So, the government is inherently clumsy. It is like a sledgehammer. Sometimes, sledgehammers are useful, but it is not the kind of tool one would use to do anything as delicate as, say, assembling a computer with. Put differently, the government is not well-equipped to create sophisticated laws; it should stick to simple-to-implement ones, like property rights.

Is Taxation Theft?

Libertarians also have moral criticisms of the government: they argue that taxation is a form of theft because there is no voluntary agreement between a government and its subjects that the government may take some money in exchange for the public services it provides. That there is no such agreement, contractual or otherwise, seems obvious, at least to me, yet often people deny that it is so.

For example, in the podcast series I mentioned at the start of this blog, Bruce Nielson argues that governments are entirely voluntary for the following reason: consider a homeowners association in a stateless region. The homeowners' association produces laws for the people who live under it, so in that sense, it is like a government. Yet, it is also clearly voluntary because there is an original contract between the homeowners association and those living under it. Bruce then argues that this is a plausible origin story for modern governments: if a particular homeowners association keeps growing in size, and the contract no longer explicitly states that by buying property in this territory, you become part of the homeowners’ association, it has effectively become a state.

This story is a version of social contract theory, which presupposes that there might have been (or might as well be) a social contract between the government and its citizens, which legitimises the state. However, there is no such contract, and even if there was originally one, it is hard to see how we, the descendants of the people who originally signed it, could still be bound by it. Bruce tries to make sense of that by saying the agreement is implicitly signed by anyone who buys a property within the homeowners’ association, but that would be information that must be available to a buyer. If the homeowners’ association and its functioning aren't specified in any contract, it is hard to see how contract law could be applicable here – in that case, we have moved from a situation in which there was a contract to one in which there isn't.

A similar argument is that the government gets to levy taxes because people voluntarily agree to them when they become residents of a country, much like how business owners charge people for buying items from their shops. This argument fails because the business owners own the shop and the items for sale therein, so it presupposes that the government owns all the land in its territory, which it doesn't.

Hence, without something like an explicit contractual arrangement, the voluntariness of taxation is murky at best. People have put forward other arguments over the years that try to prove the same thing Bruce did, all of which seem, at least to me, equally unsuccessful; for those interested, this is a good piece about some of them.

Does that mean that taxation is theft? Here, I depart from conventional libertarianism and its critics because I think there is a straightforward solution: the government is coercive, but its coercion is legitimised.

How is government coercion legitimised? It is legitimised because, at least in modern democracies, the particular policies that a government enacts are open to revision. The government can be improved; errors can be corrected. So, although the government may sometimes use its monopoly on force in harmful ways, there are mechanisms to correct such mistakes when they arise. This unique ability that defines modern democracies is worth some temporary hardship and coercion. It does not mean such coercion does not exist, but it is tolerable in light of what is gained.

Anarcho-Capitalism

A version of libertarianism that takes voluntary association to its logical limits is anarcho-capitalism, which envisions a society where all services – including law, national defence, and police – are privatised. Various clever arguments have been put forward in favour of this theory, and I find these arguments persuasive in the following sense: my view of anarcho-capitalism is that all theories of the form 'the state is fundamentally necessary' are wrong. In pure economic theory, the state seems to have no fundamental necessity. It is a bit like Laplace's response to Napoleon's question of why God was not mentioned in Laplace's work on celestial mechanics: 'I had no need of that hypothesis.' Perhaps one day, we will discover a fundamental theoretical need for the state, but I doubt this.

This way of viewing anarcho-capitalism makes it less utopian than it is sometimes presented. For instance, my assertion that theories about the fundamental necessity of the state are simply false does not imply that government is fundamentally evil, that we should hate the state, or that we should get rid of the state as soon as possible. Most democracies are far from the anarcho-capitalist ideal, but they function reasonably well. There are many inefficiencies in the modern world, many of which could be resolved by privatising various parts of the economy, but this does not mean that creating a version of anarcho-capitalism is currently practicable.

Currently, we need governments to provide foreign policy, law, and police services. Although these publicly provided services could, in principle, be performed by the private sector, it is hard to see how to transition to a world in which they are, both due to a lack of practical policy proposals to that effect and a lack of political will for such policies. This doesn't mean it is impossible to transition; it just means it is hard at the moment. It is more practical to focus on incremental improvements, as is almost always the case. However, the theory of anarcho-capitalism might well have practical benefits in that it can be used to criticise practical policy proposals.

Criticisms of libertarianism

I have presented various arguments in defence of libertarianism, but that’s not to suggest that libertarian ideology is without its flaws. One such flaw is that libertarianism, as I have described it, is a purely negative theory. It consists of criticisms of other theories, like theories about policy proposals, but it does not generally offer a positive account of what kinds of policies to create (beyond privatising parts of the economy). One reason that is a problem is that in transitioning from a publicly run economy to a private one, there will be inefficiencies and friction that make such transitions costly or even prevent them altogether.

Another flaw with the conventional libertarian account is that it is an overly rationalistic theory. For instance, libertarians often hold up the ‘non-aggression principle’ as an absolute moral rule—this principle states that aggression is inherently wrong except for aggression used in self-defence. That is true whenever aggression is clearly defined, but it isn’t always. Consider a noisy neighbour. At what point is your neighbour aggressing against you by being too noisy? If your neighbour is making noise exceeding 140 dB, a reasonable person might consider that a type of aggression, whereas his whispering at 20 dB is likely to go entirely unnoticed. Somewhere between those extremes is a grey area, which sometimes requires the machinery of law to clarify what is and what is not aggression.

Finally, libertarian criticisms do not always provide a knock-down argument against particular policies. There are other trade-offs to consider, too, the main one being national defence, which might require the government to intervene heavily with certain sectors of the economy. For example, in 1602, Oldenbarnevelt, the Grand Pensionary of the Dutch Republic, merged the various trading companies that existed at the time into a state monopoly known as the Dutch East India Company (VOC). This merger provided a significant revenue source for the young Dutch Republic, part of which was used to maintain a standing army the country was in dire need of. According to libertarian principles, this might have been considered an immoral decision, but it was a practical solution to the real problem of national defence. Perhaps that problem could have been solved without the need for a state monopoly; for instance, taxing the private trading companies might have produced enough or more revenue for national defence. However, it might also have been the best solution available at the time – it is hard to tell, partly because libertarian and economic principles were not the only considerations.

These kinds of considerations seem to me to be far more severe criticisms of libertarianism than the more conventional ones.

1. Also, as David Friedman points out in his book The Machinery of Freedom, modern economies are not very centralised, so the need for such government intervention is minimal.