What this covers

Richard Epstein of NYU and Stanford's Hoover Institution discusses with Russ Roberts how constitutional interpretation has shifted since the founding and what institutional arrangements best protect against concentrated power. The conversation moves through the relationship between federalism and federal authority, the rise of the administrative state, and how courts should calibrate their role across different doctrinal areas. Epstein argues that the 1787 Constitution—structured around enumerated powers, federalism, and property rights—more effectively constrained factional capture than the post-1937 regime of expert administration and broad delegations. The core claim is not that courts should simply retreat, but that intervention should be evaluated case by case rather than reduced to a single posture.

The discussion ranges across the historical irony that slavery, by making federalism necessary to hold the union together, accidentally constrained the natural drift toward centralization—a restraint abandoned after emancipation. It covers how loose statutory mandates, exemplified by the FCC's "public interest" standard, embed regulatory capture at the constitutional level rather than merely at the agency. Epstein contests the framing of constitutional choices as activism versus restraint, using examples like Kelo (which he reads as judicial passivity despite its unpopularity) and child-rape sentencing to show that disagreement over outcomes differs from disagreement over the proper scope of review. The conversation also addresses motivated reasoning among justices, the ideological drift of conservative appointees, and the tension between originalism and long-standing doctrinal settlements, with Epstein proposing a doctrine of prescription modeled on property law to explain which erroneous moves become constitutionally entrenched.

Sharpest takeaway

Epstein argues that the original 1787 Constitution—built on enumerated powers, structural federalism, and protection of property and contract—was a more sophisticated guard against factional capture than the post-1937 administrative state, and that the proper question is not 'more or less judicial activism' but a retail-level determination of where courts should intervene and where they should defer.

  • Federalism's restraint on federal power was an ironic byproduct of the slavery compromise, and once slavery was gone the natural drift toward centralized Leviathan resumed after 1937.
  • Loose delegations like the FCC's 'public interest' charter build regulatory capture into statutes at the constitutional/congressional level, not just at the agency level.
  • Constitutional interpretation should be disaggregated case-by-case (e.g., flat-tax rate structure vs. tax-rate level) rather than reduced to a single posture of activism or restraint.

The argument · threads10 threads · 43 claims
0.85

The 1787 Constitution deliberately constrained federal power to preserve state sovereignty and inter-state competitive federalism.

8 pointscentrality 5/5
  • The original constitutional design limited federal power to regulate activities within states largely because of the slavery question: to bring the union together, the great compromise let southern states keep slavery and northern states keep abolition, which required powerfully limiting federal authority over intrastate activity—giving the federal government only powers like regulating foreign and interstate commerce and Indian relations while leaving the rest to state police power.

    “The central feature of the great compromise that brought the United States together was the decision to allow the southern states to keep with their slave traditions and the northern states to keep with their abolitionist traditions... In order to make this work, what you had to do is to limit, very powerfully, the ability of the Federal government to regulate activities that go within the states.”

  • Federalism allows states to function as laboratories for policy experimentation in a trial-and-error mode—successful experiments get imitated and failures abandoned, giving a way to compare and contrast that constrains overreaching government—a metaphor famously from Brandeis's dissent in New State Ice v. Liebmann (1932); but the limitation is that some experiments, like an ice cartel justified only by protectionism, are known failures before they are tried.

    “the states are laboratories in which we can have experimentation going on. And it's kind of like the trial-and-error mode of scientific inquiry: this state seems to make it work, that state doesn't”

  • If you ask which constitutional provisions have durability outlasting their creation, a very large fraction do—especially the structure of enumerated powers and inter-state competition policed by federal anti-blockade power, which works as well in the age of the internet and airplane as in the horse-and-buggy era—while the ones that failed (the three-fifths clause and fugitive slave provisions) are all gone.

    “if you look at the Constitution and you sort of ask yourself which of its provisions seem to have durability that outlast the time of their creation, I would say a very large fraction of them do”

  • Under the 'dormant' commerce clause, the Supreme Court inferred that in the absence of congressional legislation it has a duty to preserve a competitive economic union—preventing states from imposing differential barriers on out-of-state merchants and thereby creating a nationwide common market rather than a Balkanized one—a great intellectual achievement that effectively invented a classical liberal doctrine, though it still permits states to block genuine harms like poisons or invasive species.

    “under the so-called 'dormant' commerce clause, it has basically inferred that in the absence of Congressional legislation, it is its duty to essentially make sure that we preserve the basic outlines of a competitive economic union”

  • The Constitution was drafted to cure the defects of the Articles of Confederation—adding a single President, a taxing Congress, and a federal judiciary the Articles lacked—seeking enough centralization for the nation to operate coherently but not so much as to squash individual state variations; it was never designed to create the modern welfare or New Deal state.

    “the Constitution was drafted as a way to get rid of the defects of the Articles of the Confederation, rather than to create the modern welfare or New Deal state”

  • Federalism means almost the opposite of centralized state power: a federal (as opposed to unitary) system is best understood as loosely affiliated states with separate governances but some comity, exemplified by the Full Faith and Credit clause whereby each state must respect the judicial and legislative acts of others rather than relitigating them—with the U.S. complicated further by a federal government layered on top.

    “A Federal system, as opposed to a unitary system, is best understood by going back to the Articles of Confederation, dropping the first syllable. And 'federated' means loosely affiliated, one through another”

  • After the Civil War the 14th Amendment's privileges-or-immunities clause ('no state shall make or enforce any law') gave the federal government a powerful veto power over bad state acts, enforced mainly by Congress, while still leaving states the initiative; the federal power to affirmatively initiate the rules states must follow did not arrive until 1937—so the 1868 design cleverly vetoed harmful state acts while letting good state acts (like tax competition) proceed.

    “what was thought was that you'd have state initiative only, but you have Federal veto power. But what you did not have under the 1868 situation, where it was that the Federal government could initiate those kinds of rules that the state had to follow. That came only in 1937.”

  • It is deeply ironic but clear that the cause of limited government was advanced by the institution of slavery, because slavery made federalism a critically important issue and thereby restrained the natural drift toward centralized Leviathan—much as the Quebec/English divide makes federalism important in Canada—a restraint that was largely abandoned once the slavery issue was gone.

    “it's quite clear that the cause of limited government was advanced by the institution of slavery, because it made Federalism a very important issue. Just the way in Canada it's an important issue because of the differences between Quebec and the English-speaking provinces”

0.74

The 1937 constitutional revolution abandoned enumerated powers and structural limits, enabling an administrative fourth branch with vast discretionary authority.

5 pointscentrality 5/5
  • By 1937 everything changed: separation of powers was viewed with suspicion, the administrative state rose, and the enumerated-powers limit (especially the Commerce Clause) was scrapped for a system of concurrent jurisdiction in which the federal government can act on any important economic issue and the state can too—a huge shift from a lean classical state of small government and strong property rights to a law state with heavy administrative law and broad discretion at all levels.

    “it's a huge difference between a relatively lean classical state--small government, strong property rights--to a law state with heavy administrative law, concurrent jurisdiction. And lots of discretion that's being given to agents at all levels of government”

  • 1937 was not primarily a response to the Great Depression but the culmination of a 40-year movement building since around 1900: progressives steadily expanded both federal commerce power (eventually allowing regulation of intrastate manufacturing, mining, and agriculture and propping up labor and agricultural cartels) and faith in expert administration, so the New Deal expansion reflected underlying forces rather than just crisis response.

    “1937 is in fact an extremely important date. Because it represents the final culmination of a movement that has been building for 40 years, on two dimensions.”

  • By the 1930s the necessary-and-proper clause was reinterpreted to legitimize the administrative state, so that instead of three branches of government there is now a 'fourth branch'—the agencies—and combined with reading the Commerce Clause to reach things that merely 'affect' interstate commerce, these seemingly verbal shifts expanded federal power by at least an order of magnitude.

    “when you get to the 1930s, all of a sudden people are saying necessary and proper is the way in which you understand the legitimacy of the administrative states. So, instead of having three branches of government, you can now have what is called the 'fourth branch of government'”

  • The single largest constitutional issue we face today is whether the judicial system will assert control over federal regulation, which Epstein believes is ruinous to the organization of a competitive economy and contrary to its economic ideals.

    “the single largest issue in the Constitution that we have to face today is whether or not the judicial system will assert its control over Federal regulation, which I think is ruinous to the organization and against its competitive economic ideals”

  • The 'administrative state' is not merely a government that administers tax rolls and highways; it is the conscious effort to subject industrial, manufacturing, and commercial functions to strong oversight by professionals presumed to have the scientific expertise to manage an economy, with the binding constraint being political oversight via popular election rather than freedom of contract and private property.

    “The 'administrative state' refers to the conscious effort on the part of government, to believe that essential industrial and academic... essentially industrial and manufacturing, commercial functions, or to be subject to a very strong degree of oversight by professionals. And these professionals are supposed to have the kind of scientific expertise that would allow them to manage an economy.”

0.60

The Court's post-1937 expansion contradicted its earlier restraint, shifting from deference on economic matters to activism on civil rights with inconsistent principle.

5 pointscentrality 4/5
  • After 1937 the Court reversed its restraint: having decided it was 'absolutely terrible' to strike down minimum wage or maximum hour laws, by Brown v. Board of Education in 1954 it acted as the 'super-est of super-legislatures' in striking down the entire state system of segregation—showing the strange inconsistencies in constitutional law as priorities shifted from economic regulation to race.

    “eventually in Brown v. Board of Education the Court acted as the super-est of super-legislatures when it struck down the whole state system of segregation, after having decided that it's just absolutely terrible to strike down a minimum wage or a maximum hour law”

  • Both progressive and classical-liberal justices reach decisions that are perfectly coherent with their prior political principles, so the realistic view—that people have biases and ideologies, do what they want, and cook up the reasons later—is largely correct; the test against this is whether your interpretive theory always yields a 'world's perfect constitution' that matches your own beliefs, versus being able to find something written there that disagrees with what you favor.

    “the descriptive that you've given, that is that people have their political preferences and then what they do is they organize their judicial theory around it, is in fact correct”

  • There is no single-direction answer to whether there is too much or too little court; constitutional review should be evaluated at a retail level case by case—the Court has gone too far in using the cruel-and-unusual-punishments clause to bar executing child rapists with no basis in 'we the people,' yet too far the other way in letting confiscatory zoning laws pass unchallenged—so labeling decisions one disagrees with as 'judicial activism' is often wrong, since some are actually cases of judicial abnegation.

    “I just don't see this one-directional situation on any relevant issues, which allows me to say there's too much court or too little court. I take this at a retail level, not at a wholesale level.”

  • Kelo, widely condemned as outrageous judicial activism, was actually the opposite—judicial passivity: the Court knew the taking was not a genuine public-use transfer (private party to private party) yet deferred to the state, and because every transfer has some indirect public benefit, the public-use limitation was de facto read out of the Constitution; people wrongly call decisions they dislike 'activism' even when they are cases of complete judicial passivity.

    “I just don't see this one-directional situation on any relevant issues, which allows me to say there's too much court or too little court. I take this at a retail level, not at a wholesale level.”

  • The 'age of miscalculation,' in which presidents appointed justices who later defied their expectations, is now over: every one of the current nine justices is reliably aligned with the appointing party's expectations except possibly Kennedy (appointed 25 years earlier), who nonetheless transformed the Commerce Clause argument by asking whether Congress has the power to create commerce in order to regulate it.

    “the age of miscalculation, as one might call it, I think is over now. If you look at the current Supreme Court, every one of the 9 of them is reforming in the [?]. Except possibly for Kennedy.”

0.56

Regulatory regimes systematically drift from correcting specific abuses toward protecting incumbent industries and concentrating power.

4 pointscentrality 4/5
  • The regulatory capture model does not work unambiguously, because the system gives veto power to many parties: industry groups lobby for relaxed controls while well-organized environmental groups constantly sue agencies for stricter rules, agencies themselves are conflicted and often not pro-industry, so outcomes are indeterminate and impossible to predict in advance—making it hard to generalize across or even within administrative schemes.

    “it would be a mistake to think that the regulatory capture model works unambiguously in all cases”

  • Railroad long-haul/short-haul price inversions—charging more to go from Omaha to Kansas City than from San Francisco to Chicago—were not arbitrary gouging but a form of Ramsey pricing that loads the fixed network costs onto the inelastic portions of demand because otherwise the network cannot be funded; the 1887 Interstate Commerce Act's requirement that short legs cost no more than long hauls misunderstood this and was the first federal rate regulation.

    “this was a form of Ramsey pricing, which means that in effect you put the fixed cost of the system on the inelastic portions of the overall operation. Because otherwise you can't fund it.”

  • Regulatory capture is not merely an inevitable byproduct of a well-designed scheme that gets derailed at the agency level; it is built into the statutes themselves—exemplified by the FCC's charter to advance 'the public interest, convenience, and necessity,' which courts (per Frankfurter's 1943 NBC decision) read to let the agency determine not just the rules of the road but the composition of the traffic, instead of the sensible alternative of assigning and auctioning frequencies.

    “the issue about capture is not sort of an inevitable by-product of a well-organized scheme that somehow gets derailed. It is actually built into the statutes themselves.”

  • The administrative system achieves non-transparency paradoxically by publishing so much information before proceedings that no one can assimilate it, so the serious work gets done behind closed doors—a nominal nod to transparency that conceals how the sausage is actually made.

    “The way in which we've achieved non-transparency is to publish so much information before any one of these particular proceedings take place, and nobody can figure out what it is that's going on.”

0.49

Federalism works well for tax and policy competition but fails when states use zoning and entry restrictions to entrench local monopolies.

2 pointscentrality 4/5
  • Federalism is good because it creates tax competition among states, but bad because it lets states create and protect land-use monopolies through zoning and entry restrictions; so the right constitutional approach is to permit good state acts like tax-rate experimentation while striking down bad ones that entrench local monopolies.

    “Federalism is good because it creates tax competition; Federalism is bad because what it does is it creates land-use monopolies that the state can easily control through zoning regulations or entry restrictions”

  • Railroad rate regulation expanded from controlling price inversions in 1887 to a comprehensive system of rate regulation by the 1920s that effectively produced a cartelization of the industry courtesy of government, illustrating how regulatory regimes drift from correcting specific abuses to protecting incumbents.

    “you now had a cartelization of the industry, courtesy of government”

0.46

Marbury v. Madison and judicial supremacy misread the Constitution's original separation of powers.

5 pointscentrality 3/5
  • The original institutional design intended to marginalize the Supreme Court: there was to be one Court reviewing decisions, but it was checked by a Congress that could limit which cases it took, and it was unclear whether the federal center could override state court decisions on federal constitutional grounds—leaving primary power in federal and state legislatures and executives rather than in the judiciary.

    “you had an institutional design which was intended to marginalize the Court in its original foundation, and to leave primary power in the federal legislatures and the President and the state legislatures and the state governors”

  • Judicial supremacy—the power of courts to veto or override laws—is not in the Constitution's text; Montesquieu and Locke described the judicial role as protecting individual rights under laws passed by the legislature and enforced by the executive, never as vetoing them, so Marbury v. Madison and Martin v. Hunter's Lessee read the separation of powers in exactly the opposite of its original meaning.

    “you read closely what the Constitution says about the creation of the Federal Court system, and it's just not in there. You go back and you check the way in which Montesquieu and Locke describe the separation of powers, and the judicial role was to protect individual rights”

  • The proposal to 'ignore the Constitution' (associated with Seidman, Tushnet, and Kramer on the left) is incoherent because no one truly means we should disregard Supreme Court judgments or let 40% votes become law; even its proponents want First Amendment protections enforced, so the real question is which structural constraints legislatures cannot avoid while still giving government enough power to meet modern challenges.

    “I don't even know what it is to ignore the Constitution. Do we start to say, in effect, that people can disregard the judgments of the Supreme Court?”

  • A judicially enforceable constitutional rule should fix tax-rate structure but not the level: it would be crazy to cap average tax rates or total budgets, but it is sensible to require a broad tax base with flat rates, which would eliminate the political discretion that 'eats you alive,' end the top-1% debate, and transform American politics—so on rate level you want complete congressional control but on rate structure you want very strong judicial control.

    “under our Constitution if you want to eliminate the kinds of political discretion that can eat you alive, what you have to do is to have a broad tax base and the tax rates have to be flat. And if you did that it would be a complete transformation of modern American politics”

  • We all have simple theories that explain some portion of constitutional behavior, but the closer you look at any particular issue the more complicated the cross-currents become—so descriptively it is very hard to figure out how capture theory works under the current Constitution, and normatively it is very difficult to predict which way the justices will come down on the really big cases.

    “we all have simple theories which explain some portion, but the closer you look at any of the particular issues, the more complicated the cross-currents turn out to be”

0.39

Long-standing constitutional errors become entrenched through acquiescence, but selective overruling shows only some errors gain legitimacy.

3 pointscentrality 3/5
  • Constitutional law has two tools to handle doctrines that began as incorrect but became entrenched: originalism (text and modes of construction) and what Epstein calls the 'prescriptive Constitution'—by analogy to property prescription, where a 20-year trespasser becomes the new owner, a gutsy and probably incorrect constitutional move that people acquiesce in over a long period becomes the new Constitution (e.g., Marbury), but only with selective judgment, since some long-standing decisions like Plessy (58 years) were never widely accepted as legitimate and were rightly overruled.

    “the other is what I call in my new book the 'prescriptive Constitution'--i.e., if somebody trespasses on your land, he's a wrong-doer, right, Russ? If he does it for 20 years, he's a new owner. That's the doctrine of prescription.”

  • The Second Amendment is largely a structural provision intended to protect state militias from federal interference rather than a free-standing individual right; today the militia portion is dropped out and it is read as an individual right, which Epstein argues is not originalism—so Scalia (writing for the majority) was wrong on originalist grounds while Stevens's dissent was probably correct, even though it is a poorly drafted amendment.

    “I think it's largely a structural position which is intended to protect the Federal government from regulating the way arms are used in the states so as to allow the states to organize their militias... But today it's read as a free-standing right and the militia portion of it is just dropped out.”

  • The Court will likely move substantially toward creating constitutional equal-protection rights for gay marriage, but although Epstein is politically sympathetic as a libertarian, he is highly doubtful this can be squeezed into the Equal Protection Clause as understood in 1868—making it a judicial fabrication of a new social right, originalism on one hand against a nascent libertarianism on the other.

    “I'm highly doubtful that you can squeeze this into the Equal Protection clause as it was understood in 1868. The Supreme Court, I predict, will move fairly substantially in the direction of creating constitutional equal protection rights, with respect to gay marriage.”

0.29

Tax instability and political interest-group targeting of concentrated federal power undermine economic planning and wealth creation.

4 pointscentrality 2/5
  • No major U.S. tax rate currently in place has a half-life of more than two years—they all get changed in intensity, progressivity, and type—as shown by the estate tax compromise of early 2013 being immediately reopened by the President proposing to drop the exemption from $5 million back to $3.5 million; this erratic instability makes it impossible for people to plan.

    “there is no major tax rate currently in place which has a half-life of more than two years. They all get changed.”

  • Most people want a more active federal government to advance the particular cause they champion and a smaller federal government for everything that harms them—straddling Gerald Ford's insight that a government big enough to give you everything is big enough to take everything away—so academic debates matter enormously because there is no automatic, divinely-ordained guardian of the public welfare; you must change the climate of opinion to change the judges.

    “I think most people want a more active Federal government to advance the particular cause that they champion and a smaller Federal government with respect to all those things which harm them so greatly.”

  • Most of the interventions taken in the 1930s prolonged rather than lessened the Great Depression, which lasted well into the middle of World War II, and this was due to Roosevelt's regulatory policies rather than his spending policies—and excessive economic regulation today amounts to active self-strangulation of the nation, inconsistent with the original protections of property and contract.

    “we also know that most of the interactions that were taken in the 1930s prolonged rather than lessened the Great Depression, which lasted well into the middle of WWII. And it was because of most of the Roosevelt policies, not the spending policies, but the regulatory policies.”

  • Considering how to design institutions against the recurrent dangers of political order and factions, the 1787 Constitution is actually more sophisticated than the 1937 Constitution, and part of the current malaise is that the concentration of power in Washington makes it a huge target for interest groups—because you cannot create wealth if all you do is transfer it from one party to another.

    “the 1787 Constitution is actually more sophisticated than the 1937 Constitution. And part of the reason why we have such a malaise now is there is such a concentration of power in Washington that it's a huge target for every interest group in town”

0.21

Constitutional provisions protecting enumerated powers and state competition prove durable across technological and social change.

2 pointscentrality 2/5
  • Applying his self-test, Epstein concedes the 1787 Constitution unambiguously contained a protectionist commitment against foreign trade—Hamilton was a mercantilist who argued in Federalist No. 11 that the federal foreign-commerce power existed to present a united front against foreign states rather than be at the mercy of market forces—which Epstein considers a bad idea but cannot deny was intended.

    “the question is whether or not in 1787 there was a protectionist constitution against foreign trade, the answer to that question is unambiguously, yes. Hamilton was a mercantilist”

  • In McCulloch v. Maryland John Marshall interpreted 'necessary and proper' to mean merely 'appropriate,' lowering the level of scrutiny on congressional action—'necessary' implies you can't do without it while 'appropriate' is a much looser connection—which let him uphold the national bank; Epstein thinks Marshall misread the clause, which traditionally would only supply ancillary powers (like housing stipends for D.C. residents) rather than serve as an elastic grant.

    “When John Marshall interpreted this provision in a case called McCulloch and Maryland, he said that the words 'necessary and proper' when taken together mean 'appropriate.' And that clearly, as a linguistic matter, lowers the level of scrutiny”

0.21

Constitutional design requires both structural separation of powers and enumerated rights to guard against power concentration.

2 pointscentrality 2/5
  • Hamilton was wrong that constitutional structures alone are sufficient protection for civil rights, and the proponents of a Bill of Rights were correct that a degree of redundancy is needed; when worried about excessive concentrations of power you cannot put all your faith in one kind of remedy, so structure and an enumerated rights document should reinforce each other.

    “in a very deep sense, Hamilton was wrong when he said that the structures are the protections for civil rights. And the guys who wanted the Bill of Rights were correct when they said that a certain degree of redundancy is needed.”

  • It is very common for Supreme Court justices appointed by conservative presidents to develop deeply liberal sentiments once on the Court—Epstein estimates that between roughly 1956 and 2005 about three sitting justices on average each year were conservative-appointed but liberal-leaning—with examples including Blackmun, Brennan, Warren, Stevens, Souter, O'Connor, and even Rehnquist, while the reverse drift from left to right essentially does not occur.

    “each year on average there were three justices appointed to the Supreme Court by conservative presidents who turned out to have deeply liberal sentiments”