What this covers

Richard Epstein defends classical liberalism against hard-line libertarianism in conversation with Russ Roberts. The argument turns on method: where libertarians adopt a deontological stance—opposing state power as inherently wrongful regardless of outcome—classical liberals embrace consequentialism, asking whether limited government actually solves coordination and welfare problems that markets alone cannot. Epstein contends that real social order requires not just freedom of contract and force enforcement but state solutions to holdouts, monopoly power, and common-pool resources. The U.S. Constitution, structured around enumerated powers, separation of powers, and federalism, models how to regulate the monopoly of force without inviting tyranny.

The discussion ranges across historical cases, economic theory, and constitutional law. Epstein examines railroad rate regulation and the problem of natural monopoly; why strong patent protection, despite the patent thicket objection, sustains innovation by covering specific devices rather than underlying methods (illustrated by Samuel Morse's telegraph); how common-pool resources like oil fields demand state-mandated unitization when voluntary bargaining breaks down; and the economics of employment, wages, and life expectancy in the progressive era—arguing data contradicts exploitation narratives. He revisits Lochner v. New York as a public choice gambit by unions rather than genuine health regulation, and scrutinizes Justice Holmes's celebrated dissent as eloquent but flawed reasoning. Throughout, Epstein emphasizes that classical liberalism avoids defending indefensible positions (zero taxation, private streets, no public education) that progressives exploit to discredit limited government, while hard-line libertarianism offers no coherent theory of taxation or institutional design suited to imperfect real-world conditions.

Sharpest takeaway

Epstein argues that classical liberalism, with its consequentialist willingness to permit limited, well-structured government, is a more defensible and effective political philosophy than hard-line deontological libertarianism, because real social order requires not just enforced contracts but state solutions to coordination, monopoly, and public-goods problems.

  • Classical liberals and hard-line libertarians agree against progressive expansion but split on methodology: consequentialist vs deontological
  • Markets and freedom of contract cannot solve holdout, coordination, and common-pool problems, which require limited state action
  • The U.S. Constitution's structure (enumerated powers, separation, federalism) is the model for disciplining the state's monopoly on force

The argument · threads13 threads · 36 claims
0.66

Classical liberalism opposes monopoly and cartels because they reduce output and consumer welfare, not on deontological grounds.

3 pointscentrality 5/5
  • Whether antitrust law is beneficial depends on who enforces it: focused on cartel-like activity (as under Bork) the error rate is low, but an administration that treats predation as the dominant offense will treat better products at lower prices as ruinous competition, reverting to a Progressive mode that punishes the very competition markets want.

    “The danger, of course, is that you will get some other administration that will believe that predation is the dominant offense, so when people start getting better products at lower prices, this now becomes ruinous competition”

  • On cartels, there are three distinct approaches: the libertarian sees no problem because free entry self-corrects; the classical liberal favors making such restraint-of-trade contracts legally unenforceable (without criminal sanction) to hasten their decay; and the modern antitrust position uses the Sherman Act to punish cartels civilly and criminally because cartels can endure even when agreements are unenforceable.

    “The common law took the position these contracts are contracts in restraint of trade, and its preferred [?] was not to enforce those contracts, even though it did not impose criminal sanctions on the parties who entered into it.”

  • Classical liberals and hard-line libertarians often agree in opposing progressive policies, but they differ fundamentally in method: the classical liberal is consequentialist (opposing monopoly because it reduces output and welfare), while the hard-line libertarian is deontological (opposing the government's use of force regardless of consequences), and these methodological differences produce serious divergences over a wide range of issues.

    “So, one of them tends to be more consequentialist--that's the classical liberal--the other tends to be more deontological--that turns out to be the hard-line libertarian.”

0.66

Strong property rights, including patents and trade secrecy with limited terms, are essential to preserve innovation incentives and prevent stagnation.

3 pointscentrality 5/5
  • The 'patent thicket' theory—that injunctions and many overlapping patents block innovation—is mistaken because the pace of innovation has stayed very high despite dense patent fields, and a single pioneering patent can render many earlier patents irrelevant, acting like a superhighway built over city streets that eliminates day-to-day traffic, which is why the system should encourage such pioneer patents with strong remedies.

    “sometimes you develop a patent which renders so many earlier patents irrelevant, that far from being a thicket, it's like building a superhighway over a series of city streets.”

  • Strong patent protection encourages voluntary contracts precisely because a patent covers only the specific device, not the functionality, so a competitor who finds a different technique to accomplish the same thing cannot be stopped—illustrated by Samuel Morse, who could patent his telegraph device but not the use of the electromagnetic spectrum itself.

    “the patent only covers the device. It doesn't cover a functionality. So... if somebody can figure out a same way to do the same thing with a different technique, you can't stop him.”

  • Trade secrecy can protect a process but not a thing, because once a thing is on the market a competitor can inspect and build it without incurring the fixed cost, paying only variable cost and underselling the original producer until it drops out and innovation stagnates; therefore strong patent protection by statute, with limited terms, is needed to preserve incentives—especially for drugs costing over a billion dollars to develop—since information once released reproduces at zero cost.

    “The problem with a trade secret is you can use it for a process, but you cannot use it for a thing. Because the moment you put the thing out, somebody can look at it and just build it himself.”

0.53

Where competitive markets cannot exist, intelligent regulation approximating competitive outcomes is vastly preferable to either laissez-faire or abolishing government.

3 pointscentrality 4/5
  • Where competitive markets cannot be achieved (network industries, certain natural monopolies), the goal is to use regulation that at lowest cost approximates the competitive outcome, which is a matter of technique that libertarians cannot speak to; understanding the economics reveals that bad regulation is much worse than good regulation, so intelligent contributions can be made within a framework of limited government without resorting to the nuclear option of abolishing it all.

    “What you will discover is that regulation bad is really a lot worse than regulation good.”

  • Bilateral and multilateral employment contracts can be handled by market institutions backed by public force, but many production arrangements—building roads, bankruptcy proceedings, banking, railroad interchanges—require coordination that voluntary means cannot achieve because of holdout problems and monopoly, so the state may need to mandate, for example, interchange between parties on fair terms, since pure freedom of contract does not always reach the right result.

    “sometimes you can't put all the links together in these systems through purely voluntary arrangements because there are holdout issues in monopoly.”

  • Common-pool resources require some regulation to prevent over-fishing, over-hunting, and over-extraction; with oil and gas in the 1890s, once more than about five separate owners share a field, voluntary bargaining breaks down and output is wrecked, so unitization or pooling imposed by the state (and demanded by industry) can yield up to a five-fold increase in output for a roughly 2% increase in cost.

    “the moment you get more than 5 service owners over a field it turns out that everything gets wrecked. Because they can't agree to the voluntary bargain.”

0.49

Voluntary contracts produce mutual gain but progressives wrongly prioritized income parity over total welfare gains and misread worker exploitation.

2 pointscentrality 4/5
  • During the very period progressives claimed exploitation was the norm, market data show wages rose slowly with productivity and hours worked actually declined as workers chose more leisure, and male and female life expectancy rose from about 46-47 years in 1900 to 54 by 1920—gains driven by higher incomes, better food, and discoveries curing vitamin-deficiency diseases like pellagra and beriberi, not by exploitation.

    “In 1900, about 46 years, 47 years, for men and for women, roughly the same. By 1920 it was up to 54. Now you don't get that by exploiting workers.”

  • Voluntary contracts can produce Pareto improvements—both parties gain—yet with unequal gains that widen the gap: if two people start at 10 and one moves to 12 while the other moves to 15, the summed gain is 7 but a parity-focused observer sees a new gap of 3 where there was none; progressives of the era prized parity over total gains and wrongly assumed a strong party could leave a weak party worse off, which is impossible since no one freely enters a contract that makes them worse off.

    “there is a deep mortal tension between Pareto improvements with unequal gains that will often happen in contracts and the notion of parity between parties.”

0.49

Constitutional design must constrain government force through separation of powers, veto gates, and reduced political rents available to special interests.

2 pointscentrality 4/5
  • Implementing the social contract requires a force at the center, but the great risk is that tyranny will displace anarchy; the hard problem for Hobbesians, Lockeans, and anarchist libertarians is to design intermediate institutions that turn the monopoly of force into a well-disciplined, regulated monopoly, which is the purpose of constitutional separation of powers, and most attempts at this fail.

    “if you have a force at the center [?], the great risk is that tyranny will now displace anarchy.”

  • A classical liberal constitution follows a two-part strategy: first shrink what is up for grabs in the public space so fewer unappropriated rents are available to political activists, and second build robust government institutions with veto gates that slow lawmaking, operating on the presumption that a statute is presumptively bad unless shown to be good.

    “the classical liberal Constitution that I talk about in my book has at least a two-part strategy. And one of them is what you try to do is to shrink the things that are up for grabs in the public space.”

0.49

Proportionate taxation justified as exchange for state services minimizes political discretion and avoids the public-choice perils of progressive taxation.

2 pointscentrality 4/5
  • Both libertarians and classical liberals agree on freedom of contract and a state strong enough to prevent interference and enforce contracts, but nothing in the libertarian arsenal—control of force and enforcement of voluntary agreements—yields a tax system; designing taxation and public institutions is where the classical liberal does better.

    “there is nothing in the libertarian arsenal of tricks, which has been the control of force or the enforcement of voluntary agreements, that gets you a tax system.”

  • Classical liberals gravitate toward a proportionate (flat) tax on income or consumption because it minimizes political discretion, raises needed revenue, and prevents private lobbying to dump liabilities on opponents and secure benefits for oneself; taxation under this view is justified as an exchange for a package of state goods and services exceeding the taxes paid.

    “every serious classical libertarian [liberal?--Econlib Ed.] has gravitated towards a proportionate tax on either income or consumption as a way to minimize political discretion”

0.44

Wage and employment regulation, entitlement design, and fiscal structures create perverse incentives that undermine long-term economic and political stability.

8 pointscentrality 2/5
  • The minimum wage is a danger and the living wage (about 50% higher) is more dangerous still; the original mistake was the Fair Labor Standards Act of 1938, and rather than compound that error by raising the wage to an inflation-adjusted level, policy should avoid expanding it.

    “Minimum wage is a danger; living wage, which is 50% higher, is even more dangerous bill.”

  • In the 1890s the due process clause grounded rate regulation of railroads and utilities because firms with very low variable costs and very high front-end fixed costs face a confiscation risk: once capital is committed, regulators could set rates covering only variable costs and the firm would not quit so long as it got a little extra, so 'due process' was read to require 'just compensation.'

    “if you give them only enough to cover their variable costs after they committed the stuff in the ground, they won't quit so long as they get that little extra fraction.”

  • Detroit's bankruptcy resulted from ruinous policies, a large public sector, endless transfer payments, and high union labor costs (about $75/hour all-in versus $45 elsewhere) that could not be offset by efficiencies, while California's state and city pension systems carry roughly $100 billion or more in unfunded liabilities, forcing cuts to police, welfare, and parks to honor pension priorities.

    “a firm which essentially has to pay all-in $75 an hour for a worker as opposed to one that has to pay, all-in, $45 an hour, is losing a lot of costs that they can't make up with efficiencies in other areas.”

  • Younger people see their current tax liabilities as real, fixed, finite and large while future entitlement benefits are speculative, uncertain, and may never materialize, so they experience the system not as a bargain but as an exaction.

    “the liabilities they have today are real, fixed, and finite, but large; and the benefits tomorrow are speculative and uncertain and may never materialize.”

  • If you work past age 70 while earning a decent income, you receive Medicare on one hand but pay much more in Medicare tax on the other, making you a net transferor rather than a net transferee, and the government forces a tie-in preventing opt-out of Medicare Part A without repaying all Social Security benefits.

    “if you start to work past 70 years of age, getting Medicare on one hand, and you are paying much more in on the Medicare tax on the other hand, so what happens is you are a net transfer-or rather than a net transferee.”

  • Warren Harding responded to the 1921 recession by keeping the money supply relatively tight and interest rates fairly high, which 'bled out' the difficulties and allowed recovery to begin in 1922-1923, the opposite of Bernanke's later approach.

    “he kept the money supply relatively tight; the interest rates fairly high [?]; it turned out you bled out all of the difficulties associated with the system and the recovery began in 1922, 1923.”

  • There is an asymmetry in political incentives: a small-government official faces heat when refusing demands for aid, whereas a progressive who already believes in aiding people can easily align ideals with political necessities, making it harder for limited-government principles to survive in office.

    “you face political heat if you are a small government type and somebody demands that you do something to aid them. But if you are a progressive who believes that you ought to aid people anyhow, it's much easier for you to align your ideals with the [?] political necessities.”

  • Government cannot be shrunk to fit in a bathtub as Grover Norquist quipped; even a first-best government does not come close, and to run a serious economy government will probably consume around 20% of resources in one form or another, with the good news being that the richer a society is the smaller that fraction becomes.

    “You want to run a serious economy, you better make sure and understand it's going to probably consume in one form or another 20% of your resources. And the richer you are, the smaller the fraction.”

0.35

Classical liberalism avoids defending libertarian absurdities by admitting some collective action is necessary while restructuring it sensibly.

2 pointscentrality 3/5
  • The political advantage of classical liberalism is that it need not defend absurdities (zero taxes, private ownership of public streets, no public education) that progressives can charge against hard-line libertarians; by admitting the need for some collective action and then sensibly restructuring it, the classical liberal avoids having a single concession (e.g., public streets are sensible) used to discredit the whole limited-government position.

    “the huge political advantage of being a classical liberal is you don't have to fight absurdities which can be charged against you”

  • The success of institutions depends not only on proper formulation but on a culture of cooperation and compliance; if the only thing holding them together is legal constraint, they will fail, and the classical liberal is more sensitive to these institutional and cultural issues than the hard-line libertarian who refuses to rank attitudinal preferences.

    “ultimately success of institutions depends not only on their proper formulation, but on having a culture of cooperation and compliance to make it work.”

0.31

Institutional success requires cultural cooperation and compliance alongside proper legal design; inferior people cannot be saved by good rules.

1 pointcentrality 3/5
  • The ideal institutional design is one that can survive 'choppy seas' with inferior leadership, but in reality you must staff offices with real people drawn at random, so no set of rules can save a system if it consistently gets people at the bottom; therefore you need sound institutions and a culture of achievement that elevate strong people, and you can survive occasional bad actors but not pervasive bad ones.

    “if in fact you consistently get people at the bottom, no set of rules are going to be able to save you from the depredations that they will impose.”

0.29

Historical constitutional achievements must be preserved incrementally; mid-century statutes were exceptionally sound and should be modified, not discarded.

4 pointscentrality 2/5
  • The original U.S. Constitution as augmented by the Bill of Rights does not perfectly implement limited government and contains structural errors and moral travesties (slavery, fugitive slave laws, three-fifths clause), but its surviving structure constitutes one of the greatest intellectual and statecraft achievements of all time given its drafting under political pressure in 1787.

    “for characters in 1787 sitting in a cold room in Philadelphia that figured this out on the fly with all the political pressures on them, turns out to be one of the greatest intellectual and statecraft [?] achievements of all time.”

  • The mid-century statutes—the 1952 Patent Act, the 1938 Federal Rules, the 1962 Commercial Code—were drafted by small numbers of exceptionally able experts (the Patent Act by just Federico and Rich, approved by voice vote) and were so sound in their basic intuitions that they function as quasi-constitutions; the correct approach is incremental modification rather than blowing them up, in contrast to today's rank amateurs.

    “My view is these statues become kind of quasi-Constitution. Because they are so good in terms of their basic intuitions that the correct thing to do is to figure out how to make incremental modifications of them”

  • Justice Holmes's lone Lochner dissent was the most eloquent and influential yet the worst opinion when taken apart: he justified the law on health grounds with a full stop, cited only Jacobson v. Massachusetts (a vaccination case unlike the bakery situation), and famously declared the Constitution does not enact a particular philosophy like Herbert Spencer's Social Statics—but the American Constitution is not merely 'a' constitution; it is 'this' Constitution resting on principles of limited government.

    “The last guy was Holmes, who wrote by far the most eloquent and influential opinion and also by far the worst opinion, if you actually take it apart.”

  • The 1873 Slaughterhouse case narrowed the 14th Amendment's privileges or immunities clause—meant as a general charter of liberty including employment liberty—so that only trivial privileges (use of navigable waters, petitioning the federal government) were protected, withdrawing labor markets from its scope and forcing contract-liberty defenders like Peckham to rely instead on the due process clause.

    “the Slaughterhouse case essentially made it very narrow so that the only privileges that were protected were those to use a navigable servitude on public waters and [?] petition the national government and withdrew from the labor markets.”

0.07

The social contract is a hypothetical device justifying state monopoly on force because achieving universal real consent is impossible.

2 pointscentrality 1/5
  • A decent America depends largely on the now-rare centrist Democrats combining with moderate Republicans to support a market economy and keep the welfare state under control through free trade, relatively flat taxes, few positive rights, and deregulated labor markets with a transfer program on top, though this is not stable; the only stable solution—strong pro-market capitalism with public spending limited to infrastructure and public goods—is unattainable.

    “to give you my hope for a decent America, strangely enough it relies largely on centrist Democrats, of whom there are very few today, combining with sort of moderate Republicans.”

  • Social order requires a mutual renunciation of the use of force in ordinary social arrangements; while everyone would agree to this voluntarily because giving up one's right to kill is worth little compared to being protected from being killed, the cost of achieving universal agreement is impossibly high, so political philosophers invented a hypothetical social contract—'social' because there is no real consent, 'contract' because everyone is better off in uncertain proportions.

    “you have to have a mutual renunciation of the use of force as a tool in ordinary social arrangement.”

0.04

Rate regulation of railroads with monopoly characteristics developed through due process doctrine protecting firms from confiscatory takings.

1 pointcentrality 1/5
  • Railroads exploited monopoly on short routes by charging more for a short journey (Omaha to Kansas City) than a longer one containing it (Chicago to San Francisco), which Ramsey pricing later rationalized by putting joint fixed costs on the inelastic portion of demand; 1887 rate regulation cleverly forbade charging more for a fraction of a route than for the whole, rather than setting rates directly, which raised end rates but was an imperfect partial fix.

    “Ramsey Pricing is basically a proposition... that when you have to figure out how you allocate joint costs, put it on the inelastic portion of the demand curve.”

0.04

Progressive era maximum-hours laws were not health measures but public-choice ploys by unions to disable competitors using different production methods.

1 pointcentrality 1/5
  • The Lochner maximum-hours law was not genuinely a health and safety measure but a public-choice gambit: union bakers, using a different mode of production, backed neutral-sounding maximum-hour statutes that didn't hurt themselves but wrecked the production methods of non-union bakers (whose workers slept on the job) supplying cheaper labor, so the New York Times called the decision striking it down a blow against union dominance.

    “you have one mode of production which is differentially impacted by neutral, [?] neutral set of laws that are passed to drive them out of business.”