The Oldest Question in the Field
Is international law law? In 2009, the American Society of International Law organized a panel to discuss this question.[1] The panelists were appalled it was still being asked. Andrew Guzman called it “futile.” Thomas Franck was “surprised” such a panel had even been convened. José Alvarez was “appalled that we are still discussing this 1960's chestnut.”
Hathaway and Shapiro disagree with their colleagues' dismissiveness. The question matters — morally, because legal systems are morally valuable institutions; and practically, because engaging the critics forces us to discover all the ways law can affect behavior, not just the ways we're accustomed to seeing in domestic systems.
“The principal objection made by critics of international law is that international law cannot be real law because it cannot matter in the way that real law must matter. In particular, they argue that international law cannot matter in the way it must to be law because it lacks mechanisms of coercive enforcement.”
— Hathaway & Shapiro, 121 Yale L.J. at 255 (2011)
The core move: answering whether international law is law also answers whether and how it matters. If you can show it's capable of affecting behavior in the way law must, you've simultaneously shown that it matters. The two questions collapse into one.
Figure 3
Two Centuries of Skepticism About International Law
Two Centuries of Skeptics
The critique has three overlapping parts. Austin's Objection: no sovereign issues commands, so there's no law. The Internality Objection: international law doesn't enforce its own rules through its own bureaucracies. The Brute Force Objection: no physical coercion backs the rules. Together, these form the Modern State Conception — the idea that a regime counts as legal only if it enforces rules internally through physical force.
H.L.A. Hart demolished Austin's command theory in The Concept of Law (1961)[2] — domestic law isn't commands from a sovereign either. But the Internality and Brute Force objections survived Hart's critique. Hathaway and Shapiro attack those directly.
The Enforcement Matrix
The paper's central contribution is a simple but powerful framework.[3] Law enforcement varies along two dimensions: who enforces (internal to the regime, or external) and how (through physical force, or nonphysically). This produces four quadrants. The Modern State Conception occupies exactly one of them — internal + physical. The other three are invisible to critics, but they are where most of international law enforcement actually lives.
Figure 1
The Four Modes of Law Enforcement
Outcasting is the term Hathaway and Shapiro coin for nonviolent enforcement through exclusion. Unlike physical coercion, outcasting denies the disobedient the benefits of social cooperation and membership. It does not rely on police or militia. It relies on the fact that cooperation is valuable — and losing it hurts.
Externalized outcasting is the specific form that dominates international law: enforcement carried out by actors outside the legal regime (usually states) through nonviolent exclusion from benefits. The WTO doesn't enforce its own rulings — the aggrieved state retaliates by withdrawing trade benefits.[5] The UN Security Council doesn't impose sanctions itself — it directs member states to do so.
Law Without Police
To prove the Modern State Conception is wrong, Hathaway and Shapiro reach back in history to find legal systems that worked without any of the institutions the critics say are required.[4] Two cases, separated by 2,000 miles and 200 years, use the same mechanism.
Medieval Iceland (870–1264)
Context
Iceland was settled between 870 and 930 primarily by Norwegians fleeing King Harald Finehair’s rule. The society they built was remarkably egalitarian: no king, no feudal lords, no aristocracy. Regional leaders (godi) had little executive power and farmers could switch allegiance annually.
Legal System
Iceland established a national assembly — the Allthing — in 930, which met annually as both legislature and court system. The Lawspeaker recited one-third of the nation’s laws by memory each year at the Law Rock. The Logretta (legislative council) created new laws and the Quarter Courts and Fifth Court resolved disputes.
The Enforcement Problem
Here is the critical fact: Iceland had no executive institutions. No army, no fire department, no tax collectors, no police, no prosecutors, no prisons. Victims had to prosecute their own cases. If successful, courts imposed fines or outlawry.
The Solution: Outcasting
Outlawry was Iceland’s primary enforcement mechanism. "Lesser" outlawry meant three years of exile with property confiscated. "Full" outlawry meant permanent exile, denial of legal personality, and — critically — anyone could kill the outlaw with impunity. The prosecutor was obliged to do so. Harboring an outlaw was itself punishable by outlawry. In a harsh environment where survival depended on community, exclusion was a death sentence.
Figure 4
Medieval Iceland vs. the Modern State: What's Actually Required for Law?
The defenders of the Modern State Conception might concede that premodern systems used outcasting, but argue they were deficient — crude precursors to the modern state. Hathaway and Shapiro have two responses. First, externalization and outcasting are not relics. In American cooperative federalism, the federal government routinely outsources enforcement to states.[7] Federal agencies that fail to execute congressional mandates get their budgets slashed. States that don't enforce federal law lose federal dollars. These are forms of outcasting operating today.
Second, the absence of physical coercion was a feature, not a bug. Icelanders considered their outlawry regime a mark of their egalitarianism — they had no king precisely because they didn't want one. The Church took pride in keeping its distance from Caesar. To call these systems deficient is to impose a modern prejudice that brute power is always preferable to social exclusion.
Designing Outcasting Regimes
Having established that external outcasting exists and works, the paper's final analytical move is to show that it's not monolithic. Outcasting regimes vary systematically along five dimensions, and the variation is predictable — it responds to the characteristics of the legal system being enforced.
Figure 2
Five Dimensions of External Outcasting
Every international legal regime that uses external outcasting can be classified along these five dimensions. The differences are not random — they are predictable responses to the underlying characteristics of the legal systems in which they operate. Permissive regimes work for private benefits (trade); mandatory regimes are needed when outcasting is costly. In-kind works for symmetric obligations; cross-countermeasures are needed for public goods (human rights, environment).
Figure 5
Treaty Spectrum Explorer: Mapping Regimes to Dimensions
Each dot represents where a treaty or legal regime falls along Hathaway & Shapiro's five dimensions of external outcasting. Click a treaty to isolate it and see its full profile. The WTO Dispute Settlement Body (gold) represents the most rule-bound end of the spectrum — permissive, adjudicated, in-kind, proportional, and first-party only. UN Chapter VII sanctions (red) sit at the opposite extreme — mandatory, non-in-kind, and enforced by third parties. The variation is not random: it tracks the underlying structure of each regime's cooperative benefits and enforcement costs. Positions are illustrative approximations based on the paper's qualitative analysis.
These five binary choices combine to produce 32 possible configurations, but Hathaway and Shapiro identify eight actually existing regimes[6], from “simple outcasting” (permissive, nonadjudicated, in-kind, proportional, first-party only — like bilateral air service agreements) to the maximally complex Chapter VII sanctions regime (mandatory, adjudicated by the Security Council, non-in-kind, nonproportional, third parties included).
The design logic: when outcasting is costly for the outcasting state, make it mandatory (Ch. VII). When the regime creates public rather than private benefits, use cross-countermeasures instead of in-kind retaliation (you can't respond to torture by torturing). When outcasting is too attractive — states might outcast for strategic rather than legal reasons — require adjudication and proportionality (WTO DSB).
Three Limits of Outcasting
Hathaway and Shapiro are careful not to oversell. Outcasting is not a panacea. It has three structural limits:
Limit 1: Requires Cooperative Benefits
External outcasting relies on the existence of cooperative benefits that can be withdrawn. If a state has nothing to gain from membership in a regime, threatening to exclude it does nothing. North Korea is barely touched by UN sanctions because it has already cut itself off from most international cooperation.
Limit 2: Favors the Powerful
Powerful states are harder to outcast because other states depend on them for cooperation. Outcasting the United States from the WTO would hurt the outcasting states more than it would hurt the US. This is the enforcement asymmetry that realists have always pointed to — and it's real.
Limit 3: Nonviolent
Outcasting cannot stop a genocide in progress. When immediate physical intervention is needed, outcasting is insufficient. The paper does not claim to have resolved all enforcement problems in international law — only to have shown that the traditional critique vastly underestimates the enforcement tools that exist.
Why It Matters
The paper's contribution is not just analytical but practical. If you only recognize enforcement through police and prisons, international law looks toothless. But if you see outcasting as a legitimate form of law enforcement — the WTO revoking trade benefits, the Security Council directing economic sanctions, the Council of Europe threatening expulsion — then international law appears in an entirely new light.
“Disobedience need not be met with the law's iron fist — enforcement may simply involve denying the disobedient the benefits of social cooperation and membership. Once we broaden our understanding of law enforcement to include externalization and outcasting, rather than limiting it to internalization and violence, we will see that international law matters in the way that legal systems must matter.”
— Hathaway & Shapiro, 121 Yale L.J. at 258 (2011)
This connects directly to the Goldsmith & Levinson paper.[8] Where G&L showed that international law and constitutional law face the same structural problems, Hathaway and Shapiro show they use the same enforcement mechanism. Outcasting operates in domestic law too: impeachment, defunding, judicial review, elections. The President is not arrested for violating the Constitution. He is cast out through impeachment or electoral defeat. The mechanism is identical.
Together, these two papers dismantle the deepest objection to international law from both directions: its structure is not unique (Goldsmith & Levinson), and its enforcement is not absent (Hathaway & Shapiro).
Sources
Oona Hathaway & Scott J. Shapiro, Outcasting: Enforcement in Domestic and International Law, 121 Yale L.J. 252 (2011).
John Austin, The Province of Jurisprudence Determined (Univ. of London 1832).
H.L.A. Hart, The Concept of Law (2d ed. 1994).
Hans Kelsen, General Theory of Law and State (Anders Wedberg trans., 1945).
Robert M. Cover, Violence and the Word, 95 Yale L.J. 1601 (1986).
Jesse L. Byock, Medieval Iceland: Society, Sagas, and Power (1988).
William Ian Miller, Bloodtaking and Peacemaking: Feud, Law, and Society in Saga Iceland (1990).
R. H. Helmholz, The Spirit of Classical Canon Law (1996).
Jack Goldsmith & Daryl Levinson, Law for States: International Law, Constitutional Law, Public Law, 122 Harv. L. Rev. 1791 (2009).
Related Reading