The Question
When the United States decides to use force, what actually governs the decision? The Constitution does, on the surface. But a formalist reading of Article I leaves most of the calculation unaccounted for. Reisman's work from the 1960s through the early 1990s is one long effort to specify the US-specific operational code behind that decision: who can initiate, through what channels, at what level of coercion, with what political and institutional constraints.
The framing question generalizes, which is part of why it travels. What does a legal system do when the formal rule and the working practice diverge? Constitutional war powers, market-manipulation doctrine, tipping liability, fiduciary duty, merger review, benchmark integrity — the mechanism is the same in each. Reisman's answer is the strongest position in the debate: there is no gap. The working practice is the rule, once the actors who matter treat it as authoritative. The formal text that disagrees with the practice is not a rule; it is an outdated description.[1]
An operational code is “a set of norms that operate in a certain sector and that actors deem to be authoritative even though the norms may be inconsistent with formal legal codes.”
— Reisman, War Powers: The Operational Code of Competence, 83 AJIL 83 (1989)
Why This Matters
The operational-code frame travels far beyond war powers. Five cases where the code does more work than the written rule:
Libor and Forex
The scandals were not rule-breaking in a conventional sense. They were operational-code manipulation. Before 2012, no text said rate submissions had to reflect actual funding costs, because no one thought they had to say it. The code carried the rule. When the code broke, prosecutors had to reach for anti-fraud statutes written for a different purpose. Reisman predicts this sequence: the operational code runs ahead of the formal rule until a crisis forces the formal rule to catch up.
Insider-trading tipping liability
Salman and the Newman line are fights about the operational code of what a "personal benefit" means when a tip changes hands. The text of Rule 10b-5 does not define it. The code does, through a practice the courts are trying to catch up with.
SEC enforcement competence
The agency's actual allocation of enforcement authority between commissioners, division heads, and line staff is an operational code. It is inferred from patterns of case selection, deferred-prosecution agreements, and declinations, not from the text of the Securities Exchange Act.
Board and fiduciary practice
Delaware corporate law has been running on operational code since Smith v. Van Gorkom. The written doctrine keeps trying to codify what boards are already doing. Reisman would say: stop trying. The code is the law. The text is the commentary.
War powers (the paper's own case)
Who gets to deploy a carrier, place forces on alert, or conduct freedom-of-navigation operations? No text says. The practice says. Reisman shows the code is stable across administrations and sensitive to congressional signals even when the Constitution is silent.
Figure 1
The operational-code gap: where formal rules place a power vs where the practice does
The Claim
Reisman's central argument runs in three moves.
First, behind every working legal regime there is an operational code that allocates competence among the relevant actors. In war powers, the code distributes authority to initiate, direct, and terminate coercion among the President, Congress, and the executive agencies. The code is not written down; it is inferred from what the actors actually do and from what they treat as within or outside authority.
Second, the operational code is normative, not merely descriptive. The actors treat departures from the code as illegitimate and correct them through political, not judicial, means. When a President claims authority the code does not grant, the reaction is a hearing, a budget cut, a leak, or a headline, not a court order. The sanction is political, but the norm being enforced is legal.
Third, the formal constitutional text is not the law. It is a description that may or may not reflect the current operational code. When the two diverge, the scholar's job is to describe the code honestly, evaluate it against policy goals, and propose reforms where the code is defective. The New Haven School's policy-oriented jurisprudence is that discipline.[2]
The Mechanism
How does an operational code form and stay stable? Written as a because-chain:
- A sector has recurring decisions that the formal text does not fully govern.
- Relevant actors (agencies, legislators, courts, professionals, firms) develop expectations about how those decisions should be made.
- Expectations harden into norms when departures are punished and conformity is rewarded, even without a court order.
- The norms are inferred from pattern: past practice, professional conventions, political costs of departure, institutional memory.
- The actors themselves treat the norms as authoritative, invoking them in argument, policing others by reference to them, and adjusting behavior in their light.
- At that point the norms are law in the working sense, whether or not a court has ever pronounced them.
The same chain fits the Libor rate-setting market pre-2012, the insider-trading tipping norm, corporate board practice on going-concern decisions, and the war-powers allocation of authority to issue threats. The mechanism is sector-neutral. What varies is the speed and visibility of code formation, not the form.
Figure 2
Reisman's predicted sequence: formation → steady state → disruption → codification
The timeline makes the regularity legible. Each case runs through the same four phases. Formation happens when a community of actors settles on an unwritten expectation. Steady state is when the code does the enforcement work silently. Disruption is typically a visible scandal or external shock. Codification is when the formal text catches up, partially, often preserving the operational logic the code had already established.
Why the New Haven School is load-bearing here
Reisman inherits a method from Harold Lasswell and Myres McDougal that treats law as a process of authoritative decision, not a set of commands from a sovereign.[3] That method is the foundation of the operational-code move: if law is what authoritative actors actually decide when they act, then the operational code is law by definition once you can identify it.
The method comes with a policy lens. Reisman does not only describe the code; he evaluates it against a stated set of values (human dignity, world public order). A contemporary securities-law reader will recognize the move. Legal realism in its modern market-abuse form does exactly this: document the practice, evaluate it against the policies the securities laws are supposed to serve, then argue for reform where the code is defective.
Evidence & Method
Reisman is not running a causal-identification design. Like Waxman in Part I, he is building a conceptual argument, but the identification problem is sharper because Reisman makes a stronger claim. Honest labeling of the method matters.
Identification strategy
Reisman reads post-war US practice on military coercion short of declared war and infers the allocation of competence among the branches. The inference is historical and interpretive, not statistical. Its strength is the richness of the cases; its weakness is that multiple honest readers could read the same cases differently.
Key assumptions
- The relevant actors form a stable community with shared expectations. When the community is contested, the code has not formed.
- Departures are reliably sanctioned through non-judicial channels. If they are not, the norm is rhetoric, not code.
- The observer can identify “the actors who matter” without introducing circular reasoning. This is the hardest assumption.
Threats to the argument
- The operational code risks motivated reading: the observer privileges the practice that supports their preferred policy. Hart pressed this objection in a different register. It lands here.
- When codification erases the code (post-Dodd-Frank benchmark rules, for example), is that evidence the code was pre-law, or that the code migrated into the text? Reisman's framework doesn't cleanly resolve it.
- Operational codes in markets form through different mechanisms than in constitutional law. Reisman treats the mechanism as general. A careful cross-domain comparative could narrow that claim.
External validity
Travels cleanly to most working legal regimes with a professional community and repeat-play decisions: securities enforcement, corporate fiduciary duty, antitrust merger review, international trade dispute settlement. Travels less cleanly to constitutional criminal procedure or contract law, where the formal text does more of the heavy lifting.
Beyond the 1989 paper · Reisman's US-force corpus
The Operational Code of Competence is the spine, but it is not the only Reisman piece on US force. Three adjacent papers sharpen the argument; the dissection above reads through all three without always naming them by title.
1966 · The Legal Use of Force Short of War
Published in the US Naval Institute's Proceedings. Reisman's earliest US-specific treatment of force-that-is-not-war. The spectrum from diplomatic pressure to covert action to limited strikes to declared war was already the organizing frame. Three decades before the 1989 paper, the operational-code habit of thought was in place.
1989 · War Powers: The Operational Code of Competence
The spine of this dissection. Codifies the allocation-of-competence move and names the operational code as law. The paper is narrow in scope (US constitutional war powers) and wide in method (New Haven School policy-oriented jurisprudence). Its claim is that the unwritten code is more law than the written Article I.
1992 · Regulating Covert Action (with James E. Baker)
The book-length companion on US covert action. If the 1989 paper argues the operational code exists, this book shows what it looks like in the hardest case: a domain where practice is deliberately unrecorded. Reisman and Baker infer the code from leaks, declassifications, and executive-branch doctrine, then evaluate it against policy. A contemporary securities-law reader will recognize the method: this is the intellectual ancestor of modern securities enforcement empirics, applied to a sector that does not want to be studied.
The three pieces together describe the US as a country with a calculated, stable code for when and how to move to force. The written Constitution is a constraint on that code but not the code itself. A contemporary reader picking up any one of the three will hear the other two in the background.
Doctrinal Voices
Four positions on operational code, mapped from the strongest version of the claim (Reisman himself) to the most cautious. The securities-law parallel is not a fifth position in this debate; it is a source of concrete cases that let us see Reisman's mechanism at work, treated in its own panel below. Click a dot in the map to jump to the matching card.
Figure 3
Where the four voices sit: method (text vs practice) × framing (descriptive vs normative)
What Would Falsify This?
Four candidates, in order of how much they would move the posterior if they landed.
The operational code does not actually bind
If actors describe the code one way and consistently behave another way, the code is fiction, not norm. No binding force means no law under any theory.
Evidence that would move the posterior:Systematic evidence that presidential administrations treat the war-powers operational code as pure rhetoric, or that benchmark-setting firms treat integrity norms as optional, would collapse the claim. The 2008 Libor revelations come close in the markets case.
Codification reliably erases the code
If every time a formal rule codifies part of the operational code, the code dissolves and the text takes over, then the code is pre-law, not law. It is the scaffold that gets torn down.
Evidence that would move the posterior:Post-Dodd-Frank benchmark regulation is the test case. If the operational code around Libor fully migrates into the formal text, Reisman's claim that code is law independent of text weakens.
Cross-domain generalization fails
If operational codes form in war powers but NOT in securities, or vice versa, the framework is domain-specific, not a general theory of law-practice interaction.
Evidence that would move the posterior:Comparative work showing that the code-forming mechanism differs fundamentally between state-centric law and market-centric law would force the theory to narrow.
The policy lens is motivated reasoning
The New Haven School defines the operational code by asking whether the actors treat it as authoritative. But the observer picks the actors. If the framework is manipulable — we will recognize the code that supports our policy goal and reject the one that does not — it is not a neutral diagnostic.
Evidence that would move the posterior:Critics from analytical positivism (Hart, later positivists) have pressed this. A convincing demonstration that two honest New Haven School analysts reach opposite code readings on the same facts would undermine the methodology.
So What? (Bridge to Part I)
Read the two pieces together and the difference between them becomes the point. Waxman and Reisman both see that the formal war-powers text under-describes what actually happens. They disagree about what to do with that observation.
| Dimension | Waxman (Part I) | Reisman (Part II) |
|---|---|---|
| Diagnosis | Informal channels constrain executive threats; scholarship has underweighted them. | Informal channels are the law of war powers; the written text is a description. |
| Epistemic move | Observe constraint; describe mechanism; stop short of legal status. | Observe practice; identify code; declare it law when actors treat it as authoritative. |
| Where they meet | Both reject the pure-formalist reading of Article I. Both accept that Congress shapes executive threat credibility through informal means. Both would say the War Powers Resolution is not the whole story. | |
| Where they diverge | Waxman keeps the formal / informal line visible; Reisman collapses it. This is the load-bearing disagreement. | |
A careful reader can accept Waxman's observation without accepting Reisman's conclusion. The operational code exists; whether it IS law depends on your theory of law. That is a productive disagreement, not a failure of one position.
What the markets research reveals · Operational codes in securities law
Reisman's argument is easier to see when you watch the mechanism form in a sector that is not constitutional law. Contemporary market-abuse scholarship has been documenting operational codes in securities regulation for a decade, often without calling them that. Four research lines in that literature show the same mechanism Reisman names in war powers. They illuminate the argument, they do not contest it.
Benchmark integrity
The pre-2012 Libor regime is a textbook operational code. No text required submissions to reflect actual funding costs; the practice did, until the practice broke. Contemporary work on benchmark manipulation traces the collapse in exactly Reisman's terms, without naming him.
Salman and tipping liability
The personal-benefit standard is an operational-code question. Rule 10b-5 does not name it. The practice names it, and the courts try to stabilize the practice. Every circuit split on Newman-Salman is a fight about whose reading of the code counts.
Spoofing
The law on spoofing is still catching up to the code traders have been using. Dodd-Frank named the behavior; the operational norms about what counts as spoofing and what counts as aggressive-but-legal trading still live in the practice.
Peer-to-peer lending
Early work on peer-to-peer lending is a case study in code formation under regulatory ambiguity. Before Prosper and Lending Club fit under the Securities Act, the lenders had a working code for how the market should operate.
The corollary: Reisman's insight generalizes beyond war powers. War powers is his case study. The insight is that any working legal regime has an operational code, and the scholar's job is to surface it. The securities-law cases are the best available evidence for that generalization because the codes there are so visible, so contested, and so recently disrupted.
Editor's extension · Jenn's reading, not Reisman's argument
If Reisman is right that operational codes are law, the implication runs further than he carries it. It means most formal-doctrine scholarship is a second-order activity: documenting, interpreting, or critiquing a working system the scholar did not set up. That is not a bad thing. But it changes what scholarship is for.
The New Haven School's answer is to add a policy lens: describe the code, then evaluate it against stated values, then push for reform where the code is defective. That is a working answer for a scholar. It is less clear what the answer is for a judge, who is supposed to apply rules, not evaluate policy; or for a regulator, who is supposed to enforce text, not codify practice.
The open question is whether operational-code theory needs a theory of its own institutional discipline — one that says who has authority to invoke it, under what conditions, and to what end. Reisman gestures at this. The full theory is not in the 1989 paper. It is still open.
The Clean Takeaway
Reisman's 1989 paper makes the strongest available case for the proposition that practice is the rule. Waxman in 2014 accepts the observation and stops short of the conclusion. The useful stance for a modern reader is neither surrender to the code nor fealty to the text, but the reformist one: document the code, price the divergence, and press the text to catch up where it matters. The securities-law cases on Libor, benchmarks, and tipping liability are where that document-and-price work is most advanced today, which is why they are the best concrete illustration of what Reisman was describing in 1989.
Sources
W. Michael Reisman, War Powers: The Operational Code of Competence, 83 Am. J. Int'l L. 83 (1989).
Myres S. McDougal & W. Michael Reisman, International Law in Contemporary Perspective: Public Order of the World Community (1981).
Harold D. Lasswell & Myres S. McDougal, Jurisprudence for a Free Society: Studies in Law, Science and Policy (1992).
H.L.A. Hart, The Concept of Law (2d ed. 1994).
Hans Kelsen, General Theory of Law and State (Anders Wedberg trans., 1945).
Matthew C. Waxman, The Power to Threaten War, 123 Yale L.J. 1626 (2014).
Salman v. United States, 580 U.S. 39 (2016).
United States v. Newman, 773 F.3d 438 (2d Cir. 2014).
Contemporary benchmark-manipulation scholarship in the law-and-finance literature traces the Libor and Forex cases through the same operational-code mechanism Reisman names in war powers.
Return · Part I of the War Powers Diptych
The Power to Threaten ←Matthew Waxman's narrower, more cautious version of the same move: the informal channels exist and constrain executive threats, but they are not law until someone says they are.
Related Reading