In one sentence
A society need not be governed by a territorial monopoly of law. If protection, arbitration, and legal rules were supplied competitively, people could purchase the institutions they value; the resulting system might reduce coercion and align the costs of bad policies more closely with their beneficiaries. Friedman presents this primarily as a consequentialist and economic argument, not as a proof from absolute rights.
Overview
The book begins by challenging the assumption that government is required to provide public goods. It then applies market reasoning to familiar state functions—roads, schooling, welfare, immigration, drugs, and foreign policy—before concentrating on the difficult case: private defense agencies, courts, and law. The central model is a competitive legal order in which agencies insure clients, negotiate with one another, and use arbitration to settle disputes. Friedman also discusses historical evidence, including medieval Iceland, and the problem of whether market-generated law would remain genuinely libertarian. The work first appeared in 1973; the researched edition is the substantially revised third edition published in 2014.
Core ideas
Start with incentives, not ideal citizens
Friedman’s recurring test is whether institutions work with ordinary, self-interested, imperfect people. He contrasts this with systems that require unusually virtuous officials or citizens. The relevant question is not whether markets are flawless, but whether their failures are less dangerous and easier to correct than political coercion.
Government’s monopoly is the problem to be tested
The book treats taxation, regulation, and state lawmaking as coercive arrangements that should not receive a presumption of necessity. Each government function should instead be examined separately: can voluntary organizations provide it, and what distinctive problem prevents them from doing so?
Private protection and competing law
Individuals would buy protection from agencies that investigate threats, represent clients, and insure against legal costs. Agencies would need agreements with other agencies, making arbitration firms and reputationally credible legal rules commercially valuable. Law would therefore be produced through contracts, competition, and bargaining rather than a single legislature.
Why law might become more libertarian
Friedman argues that under private law, people who demand costly or intrusive rules would bear more of the costs of enforcing them. By contrast, under government, voters and organized interests can impose policies whose costs are dispersed among taxpayers or strangers. This is his public-goods argument for why bad laws may be harder to sustain in anarchy.
Freedom as an empirical tendency
Rather than claiming that markets automatically produce morally correct rules, Friedman argues that people usually value controlling their own lives more than controlling others’ lives. Competitive institutions therefore tend to supply more room for choice—provided violence, fraud, and external harms can be contained.
The difficult cases are the real test
The book does not stop at ordinary commerce. It considers externalities such as pollution, public goods, national defense, redistribution, schooling, and legal protection. Its strongest challenge to the reader is methodological: instead of asking whether government is imperfect, ask whether its replacement would perform better under actual incentives.
Historical analogy has limited reach
Medieval Iceland is used as evidence that relatively decentralized legal institutions can exist without a modern state. The example is suggestive, not decisive: a small, socially interdependent society may not scale straightforwardly to a populous, technologically complex country.
A major unresolved issue: market demand is not the same as justice
Critic Eric Mack’s contemporary review identifies a tension in Friedman’s argument: if legal rules are determined by bargaining among agencies and clients, market demand could support nonlibertarian rules as well as libertarian ones. The machinery may maximize negotiated satisfaction without guaranteeing rights-respecting law.
Practical takeaways
- Use institutional comparison, not moral labels: for any proposed government program, identify incentives, who pays, who benefits, and who can escape or correct failure.
- Distinguish ‘can a market provide this?’ from ‘would the market provide it equitably?’ These are separate questions throughout the book.
- When evaluating privatization, focus on coordination and conflict-resolution mechanisms, not merely on whether individual services can be sold.
- Treat concentrated political power as a risk factor: even a well-intentioned state can make harmful policies cheap for beneficiaries and expensive for everyone else.
- Remember that Friedman’s argument is strongest as a challenge to assumed necessity, not as a complete institutional blueprint.
- Pair the book with a serious critique of private-law systems, especially on unequal bargaining power, collective defense, externalities, and the danger that dominant agencies could become governments in all but name.
Caveats and counterpoints
- The book is an advocacy work for anarcho-capitalism, not a neutral survey of political institutions.
- Its thought experiments often move quickly over transition problems: how a present state would be dismantled, how contracts would be enforced during the transition, and how entrenched protection agencies would be prevented from monopolizing force.
- The argument that market law would tend toward libertarian law depends on controversial assumptions about bargaining power, client choice, agency incentives, and the costs of enforcing unpopular rules.
- Historical Iceland demonstrates institutional possibility more clearly than modern scalability.
- Critics argue that competitive bargaining can produce whatever rules powerful groups are willing and able to purchase; competition alone does not establish a theory of rights.
- The third edition includes revisions and new material, so chapter emphases and examples may differ from the 1973 and 1989 editions catalogued in older discussions.
Questions worth revisiting
- Which alleged government functions are genuinely difficult to supply voluntarily, and why?
- Would private protection agencies remain competitive, or would economies of scale produce a coercive monopoly?
- Who protects people who cannot afford insurance, arbitration, or effective legal representation?
- Can a market for law distinguish legitimate demand from demand for victimless-crime prohibitions, censorship, or privilege?
- Are externalities and national defense merely coordination problems, or do they require compulsory provision?
- What evidence would show that private legal competition is safer and more liberty-preserving than constitutional government?
- Does Friedman’s consequentialist case survive if market outcomes increase inequality or leave some people with formally voluntary but practically constrained choices?
Return to this when…
Return to the sections on private police, courts, and laws; the stability problem; the public-good argument; and the discussion of externalities and national defense. Revisit the book when assessing claims that a government function is ‘obviously necessary,’ but read it alongside critiques that separate market efficiency from justice and political legitimacy.
Highlights
In more and more cases… politics and politicians not only contribute to the problem. They are the problem.
One common objection to private property is that it is an immoral system because it relies on selfishness. This is wrong. Most people define selfishness as an attitude of caring only for oneself and considering other people’s welfare of no importance. The argument for private property does not depend on people having such an attitude; it depends only on different people having different ends and pursuing them. Each person is selfish only in the sense of accepting and following his own perception of reality, his own vision of the good.
There is no way to give a politician power that can be used only to do good. If he gives food to someone, he must take it from someone else; food does not appear from thin air. I know of only one occasion in modern peacetime history when large numbers of people starved although food was available. It occurred under an economic system in which the decision of who needed food was made by the government. Joseph Stalin decided how much food was needed by the inhabitants of the Ukraine. What they did not ‘need’ was seized by the Soviet government and shipped elsewhere. During the years 1932 and 1933, some millions of Ukrainians died of starvation. During each of those years, according to Soviet figures, the Soviet Union exported about 1.8 million tons of grain. If we accept a high figure for the number who starved—say, eight million—that grain would have provided about two thousand calories a day to each of them.
Interlude I have talked in the abstract about "private property" and "public property" and have argued for the superiority of the former. But in existing societies, ‘capitalist’ as well as ‘communist’, there is a mixture of public- and private-property institutions. I may own my car, but the government owns the streets. How far can the idea of private property be pushed? Are there some tasks that must be done but that, by their nature, cannot conceivably be done privately and must therefore continue to be done by government? I believe not. I believe that although there are certain important tasks which for special reasons are difficult to do under institutions of total private property, these difficulties are in principle, and may be in practice, soluble. I hold that there are no proper functions of government. In that sense I am an anarchist. All things that governments do can be divided into two categories—those we could do away with today and those we hope to be able to do away with tomorrow. Most of the things our government does are in the first category.