Memo 05 — Institutional Economics

Series: The Landscape of Normative Systems (LNS) Memo number: 05 of 12 Primary JD questions: Q1 (sources of normative authority), Q2 (six-part analysis per source), Q3 (market/contract norm-generation behaviors), Q9 (Hayek and Ostrom threads). Status of D0: Immutable. This memo neither modifies, criticizes, nor defends D0. The Hayekian knowledge/calculation argument is stated here as setup and is not adjudicated; its adversarial use is reserved for Memo 11.

Scope note. This memo maps the economic sources of normative systems: the mechanisms by which markets, contracts, property regimes, firms, commons, and political-market institutions produce, legitimate, enforce, and mutate ought-statements. The organizing claim under examination is that economic institutions are genuine sources in the LNS sense (§0.6) — they emit norms with modality, subject, condition, and consequence — and that several of them generate norms without any central legislator, via equilibrium, agreement, or self-governance. We cover: markets as decentralized norm/price generators (norm-from-equilibrium); contracts as bilateral/multilateral private legislation (norm-from-agreement); property rights as a bundle of Hohfeldian relations; the New Institutional Economics (the positions associated with Coase, North, and Williamson); Ostrom's design principles for commons (norm-from-polycentric-self-governance); public choice theory (the political process modeled as an exchange system); and game-theoretic accounts of norms as equilibria (conventions, focal points, repeated games, tit-for-tat). We apply the mandatory six-part analysis (JD Q2) in full to two sources — market norms and contracts — and in compressed form to property, the firm, and the commons. All substantive claims carry epistemic tags per §0.4; when uncertain we tag down. We use the shared vocabulary of §0.6 throughout: source, norm, and the sharp separation of legitimacy / validity / efficacy; Hohfeldian jural relations; and deontic operators O/P/F.

05.1 Orientation: why economics is a distinct family of norm-sources

Economic institutions are unusual among the sources surveyed in this series because a large fraction of the norms they emit are not authored. A statute has a legislator (Memo 01); a doctrine has a magisterium or a canon (Memo 04); a precedent has a court (Memo 02). By contrast, a market price, a customary trade usage, or a convention such as "drive on the right" has no identifiable author and yet functions as a binding constraint on behavior. FACT that these institutions constrain behavior; [INFERENCE that "no identifiable author" is the salient structural feature distinguishing them from legislative sources.]

This gives the economic family three recurring structural signatures that the compiler team should track:

1. Emergent authorship. The norm is a byproduct (equilibrium, aggregate, or fixed point) of many local decisions rather than a promulgated text. INFERENCE
2. Incentive-based enforcement. Deviation is punished by loss (of money, of trading partners, of reputation) rather than, or in addition to, by sanction from an authority. FACT for the paradigm cases; the strength of this mechanism varies.
3. Content that is often implicit. The norm may never be written down as an ought-statement at all; it exists as a behavioral regularity plus an expectation that others conform. INFERENCE This is the hardest feature for any D1 representation to capture, because there is no surface text to compile (compare M4: language is presentation — here there may be no linguistic surface at all).

These signatures set up the two hard problems this memo hands to D1/D2: how to represent a norm that has no canonical text, and how to represent enforcement that is distributed and incentive-mediated rather than adjudicated. Both are flagged in §05.10–05.11 and again at Handoff.

A terminological caution for the compiler team. Economists use the word "institution" in a specific sense that differs from the colloquial one. In the tradition associated with Douglass North, an institution is a rule of the game — a humanly devised constraint on interaction — while an organization (a firm, a party, a club) is a player that operates within those rules. FACT that this rule/player distinction is North's stated usage. In LNS vocabulary: institutions in North's sense are sources and norm-sets; organizations are subjects that the norms bind and also secondary sources that emit their own internal norms (treated in Memo 07). We keep this distinction throughout because conflating rule with player is a known category error that would corrupt any dependency graph (JD Q4).

05.2 ESTABLISHED CONSENSUS

This section states what the relevant fields (microeconomics, contract theory, law-and-economics, New Institutional Economics, and non-cooperative game theory) broadly agree on. Disagreements are deferred to §05.3.

05.2.1 Economic institutions are genuine sources of norms

FACT Across the traditions surveyed, it is not seriously contested that markets, contracts, and property regimes shape behavior by generating enforceable expectations. The disagreement (§05.3) is about how and how legitimately, not whether.

INFERENCE Reduced to LNS vocabulary, the consensus is: economic institutions emit objects with all four norm-components — a modality (typically permission or obligation, occasionally prohibition), a subject (a party, a class of traders, an owner), a condition (a triggering circumstance: possession, agreement, membership), and a consequence (payment, forfeiture, exclusion, loss of future dealing). This qualifies them as sources under §0.6.

05.2.2 Property is a bundle of relations, not a thing

FACT The dominant analytical view in law-and-economics and in analytic jurisprudence treats property not as a relation between a person and an object but as a bundle of jural relations among persons with respect to an object. This is standardly rendered in Hohfeldian terms.

INFERENCE The canonical bundle decomposes cleanly into Hohfeld's primitives (§0.6): - the owner's privilege to use the thing (and others' correlative no-right to object); - the owner's right that others not interfere (and others' correlative duty of non-interference); - the owner's power to transfer, license, or bequeath (and a transferee's liability to have their jural position changed); - the owner's immunity against unconsented expropriation (and others' correlative disability).

This is the cleanest mapping of any source in the series onto the Hohfeldian modality set, and it is the strongest single piece of evidence that the Hohfeldian vocabulary is at least locally adequate. INFERENCE — it is evidence, not proof; adequacy is stress-tested in Memos 10–12.

Property as a Hohfeldian bundle (owner O, world W):

"Ownership" = the co-holding of all four. Partial bundles (leases, easements, licenses, security interests, usufruct) = proper subsets, and are exactly representable as subsets.

HYPOTHESIS Because partial property interests (a lease, an easement, a lien, a license) are subsets of the full bundle, property rights are representable as sets of typed Hohfeldian relations with set-algebraic operations (intersection, difference) corresponding to real legal operations (co-ownership, encumbrance, subdivision). If true, this is directly compilable. Stated as hypothesis because the correspondence is clean for textbook cases but frays at equitable interests, future interests, and defeasible conditions (§05.10).

05.2.3 Contracts create binding private norms

FACT It is not contested that a valid contract creates obligations enforceable between the parties. In LNS terms a contract is a source that manufactures norms binding on a closed set of subjects (the parties), as opposed to legislation, which binds an open class.

INFERENCE This is the sense in which contract is "private legislation": the parties exercise a legal power (Hohfeld) to generate new duties and rights binding on themselves. The state supplies the power-conferring rule (roughly, "agreements meeting conditions C are enforceable"); the parties supply the content. This two-layer structure — a public power-conferring norm enabling private content-generation — recurs throughout this memo and is the single most compiler-relevant pattern in it. See §05.5 for the full six-part treatment.

05.2.4 Transaction costs are real and shape institutional form

FACT The claim associated with Ronald Coase — that using the market is itself costly (search, bargaining, and enforcement costs), and that these transaction costs explain why firms exist and where their boundaries fall — is a foundational and broadly accepted result of the New Institutional Economics. FACT that this is the received reading of Coase's "The Nature of the Firm" (1937) and "The Problem of Social Cost" (1960).

INFERENCE The governance-relevant corollary: the choice of enforcement mechanism itself has a cost, and institutions economize on it. A norm-system that ignores the cost of its own enforcement will mis-predict which institutions survive. This is a direct warning to D2 (execution): enforcement is not free and its cost is part of the design space.

05.2.5 The Coase theorem and its standard caveat

FACT The proposition commonly called the Coase theorem states: if property rights are well-defined and transaction costs are zero, parties will bargain to an efficient allocation of resources regardless of the initial assignment of rights. FACT that this is the standard textbook statement; Coase himself framed it largely to show that the zero-transaction-cost world is not ours.

INFERENCE The governance reading: in the frictionless limit the initial normative assignment does not affect the efficient outcome, only the distribution. Because real transaction costs are positive, the initial assignment of rights (i.e., the initial norm-set) does matter for efficiency. Hence "the initial allocation of Hohfeldian entitlements is outcome-relevant" — a claim that matters for any system that assigns rights computationally.

05.2.6 Institutions are constraints that reduce uncertainty

FACT The position associated with Douglass North holds that institutions — formal rules and informal constraints (customs, codes, conventions) plus their enforcement characteristics — structure human interaction and reduce uncertainty by making others' behavior more predictable. FACT as a summary of North's stated thesis.

INFERENCE In LNS terms North supplies a functional definition of a norm-system: its job is to lower the entropy of expectations about others' conduct. This reframes "enforcement" as "expectation stabilization," which is broader than sanction and includes the mere fact of shared knowledge that a rule exists.

05.2.7 Governance structures are chosen to economize on transaction costs

FACT The position associated with Oliver Williamson holds that transactions are aligned with governance structures — market, hybrid (long-term contract, franchise), or hierarchy (firm) — so as to economize on transaction costs, with the alignment driven chiefly by asset specificity, uncertainty, and frequency. The greater the specific investment exposed to hold-up, the more likely the transaction moves from market toward hierarchy. FACT as a summary of Williamson's stated theory; "hold-up" and "asset specificity" are his terms.

INFERENCE Governance-relevant reading: the same underlying obligation can be housed in different enforcement architectures (spot market, relational contract, or command hierarchy), and the choice is driven by measurable transaction attributes. This is direct evidence that norm-content and norm-enforcement-mechanism are separable dimensions — the same substantive "ought" can be enforced by price, by contract, or by fiat. That separability is a candidate invariant for Memo 12.

Williamson's discriminating alignment (schematic):

05.2.8 Commons can be governed without privatization or state control

FACT The empirical finding associated with Elinor Ostrom is that common-pool resources (fisheries, irrigation systems, forests, pastures) are, in many observed cases, sustainably governed by the resource users themselves through self-organized institutions — i.e., neither pure private property nor central state control is necessary for success. This was established against the prior expectation (the "tragedy of the commons" reasoning) that such resources must be either privatized or nationalized. FACT that Ostrom documented such cases empirically and received the 2009 Nobel Memorial Prize in Economic Sciences for this line of work.

FACT Ostrom distilled a set of design principles commonly observed in long-enduring self-governed commons. As commonly enumerated they are: (1) clearly defined boundaries (of resource and of who may use it); (2) congruence between appropriation/provision rules and local conditions; (3) collective-choice arrangements letting most affected users participate in modifying the rules; (4) monitoring by monitors accountable to the users; (5) graduated sanctions for violations; (6) accessible low-cost conflict-resolution mechanisms; (7) minimal recognition of the right to organize (external authorities do not undermine self-governance); and (8) for larger systems, nested enterprises in multiple layers (polycentricity). FACT as the standard enumeration of Ostrom's design principles; the exact wording varies across her texts.

INFERENCE Every one of these principles maps onto a component the compiler team already tracks: boundaries = subject-set definition; congruence = condition-fitting; collective choice = the mutation mechanism; monitoring + graduated sanctions = the enforcement mechanism; conflict resolution = a dispute sub-system; recognition of the right to organize = an inter-source hierarchy relation (the state grants the local source a power/immunity); nesting = layered composition of sources. Ostrom's principles are, in effect, an empirically-derived checklist for a well-formed self-governing norm-source. This is treated at length in §05.7.

05.2.9 Many social norms are equilibria of games

FACT In non-cooperative game theory it is established that certain recurring interactions have multiple equilibria, and that a norm can be understood as a selection among them — a "way we do things" that is self-enforcing because, given that others conform, no individual gains by deviating. Coordination games (e.g., which side of the road to drive on) are the paradigm. FACT that coordination games have multiple strict equilibria and that this is the standard game-theoretic model of convention, associated with the analysis in the tradition of Schelling and Lewis.

FACT In repeated games, cooperation that is not an equilibrium of the one-shot game can be sustained as an equilibrium of the repeated game through conditional strategies (the class of results known as folk theorems). Strategies such as tit-for-tat (cooperate first, then mirror the opponent's last move) were shown, in Axelrod's computer tournaments, to perform well against a wide field of strategies. FACT that folk theorems hold under stated conditions; [FACT that Axelrod ran such tournaments and tit-for-tat performed strongly — with the caveat that "won the tournament" is often overstated and is conditional on the field of entrants.]

INFERENCE The governance-relevant consequences: - A convention is a self-enforcing norm requiring no external enforcer; the enforcement mechanism is the incentive structure itself. INFERENCE - A focal point (Schelling) is a solution that players converge on because it is salient — and salience often derives from shared culture, not from the payoff matrix. This means equilibrium selection can depend on content that lives above the kernel. INFERENCE — flagged again in §05.8 and for Memo 11/12. - Cooperation can be manufactured purely from the shadow of the future (repetition + sufficient patience) without any moral content. INFERENCE

05.2.10 The political process can be modeled as exchange (public choice — the uncontested core)

FACT Public choice theory applies the methods of economics to political and collective decision-making, modeling voters, politicians, bureaucrats, and interest groups as agents responding to incentives. Its least-contested results include: the logic of collective action (concentrated benefits and dispersed costs favor small, well-organized groups over large diffuse ones); rational ignorance of voters (acquiring political information is costly and a single vote rarely decides); and Arrow's impossibility theorem. FACT that these are core public-choice results; the normative interpretation of them is contested and is deferred to §05.3.

FACT Arrow's impossibility theorem proves that no social welfare function can aggregate individual preference orderings into a collective ordering while simultaneously satisfying a small set of reasonable conditions (unrestricted domain, Pareto efficiency, independence of irrelevant alternatives, and non-dictatorship). FACT — this is a formal, proven result.

INFERENCE Arrow's theorem is the single most important formal result for this memo's relationship to voting-based sources (Memo 01): it establishes a hard limit on preference aggregation as a norm-generation mechanism. Any D1/D2 component that claims to compute a collective preference ordering from individual orderings inherits Arrow's impossibility. This is not an engineering inconvenience; it is a theorem. It is logged as a formal constraint at Handoff and revisited in Memo 11 alongside Gödel/Church/Turing limits.

05.3 COMPETING THEORIES

Live disagreements, presented without adjudication (M8: theories are evidence, not authority; all may be partially wrong).

05.3.1 Do markets discover norms or impose them?

HYPOTHESIS , two sides - Discovery view (associated with the Austrian tradition, Hayek). The market is an epistemic procedure that discovers information no central mind could hold; prices are the compressed output. On this view the emergent norm encodes genuine, otherwise-inaccessible knowledge. (Stated fully as setup in §05.9; its adversarial deployment against computable governance is reserved for Memo 11.) - Construction/embeddedness view (associated with Polanyi, and with economic sociology). Markets are not natural discovery devices but constructed, embedded institutions whose "spontaneous" outcomes reflect the prior distribution of power, property, and social relations. On this view the emergent norm largely re-encodes existing power.

Neither is adjudicated here. The disagreement matters to the kernel boundary: if markets discover content, that content lives above the kernel and cannot be recomputed centrally (Memo 11); if markets impose prior power, then the "norm from equilibrium" is not content-neutral and importing it uncritically would smuggle ideology below the line (JD Q8, Memo 12).

05.3.2 What is the firm — a production function, a nexus of contracts, or a governance hierarchy?

HYPOTHESIS , three live positions - Nexus-of-contracts (associated with Jensen–Meckling, Alchian–Demsetz). The firm is nothing but a web of contracts; "the firm" as an entity is a legal fiction and there is no bright line between firm and market — only contracts of different kinds. - Governance/transaction-cost (associated with Coase, Williamson). The firm is a distinct governance structure that supersedes the price mechanism with authority/fiat precisely because market contracting is costly under asset specificity and hold-up. - Property-rights / incomplete-contracts (associated with Grossman–Hart–Moore). What defines the firm is the allocation of residual control rights over assets when contracts are necessarily incomplete; ownership is the right to decide what contracts do not specify.

The disagreement is directly compiler-relevant: it is a disagreement about whether "authority inside a firm" is a different kind of norm from "obligation under a contract," or the same thing wearing different clothes. Memo 07 (organizational governance) inherits this question.

05.3.3 Are property rights pre-political or state-constituted?

HYPOTHESIS , two poles - Pre-political / natural-rights pole (associated with Locke, and in a different key with Nozick, Memo 09/11 threads). Property arises from action (labor-mixing, first possession) prior to and independent of the state; the state's job is to protect pre-existing rights. - Positivist / bundle-of-sticks pole (associated with the legal realists and much law-and-economics). Property is whatever the legal system says it is — a revisable bundle of state-conferred entitlements with no pre-political core.

Unadjudicated. It determines whether property is an input to the normative system (given, protected) or an output of it (defined, revisable) — a hierarchy question the compiler must not silently resolve.

05.3.4 Public choice: descriptive science or ideological program?

HYPOTHESIS The methods of public choice (modeling political actors as incentive-responsive) are widely accepted (§05.2.10). Contested is whether the characteristic conclusions — government failure, rent-seeking, regulatory capture, bureaucratic budget-maximization — amount to a neutral positive science or carry a built-in anti-state prior. Critics argue the framing pre-loads its conclusions; proponents argue it merely applies symmetric skepticism to state and market alike. We record the dispute and take no side (political neutrality, §0.7).

05.3.5 Are norms-as-equilibria the same thing as norms, or only a model of some norms?

HYPOTHESIS , two positions - Reductive (game-theoretic) view. A norm just is an equilibrium of a repeated or coordination game plus common knowledge of it; "oughtness" is dispensable surface (M4-flavored). The tradition associated with Lewis on convention, and Bicchieri on social norms, is often read this way. - Irreducibility view. Equilibrium explains stability but not normativity: an equilibrium tells you what people do, not what they take themselves obligated to do; the felt bindingness (the deontic residue) is not captured by the payoff matrix. On this view game theory models the scaffolding but misses the norm.

This is arguably the deepest disagreement in the memo for kernel purposes: if the reductive view holds, a large class of norms compiles down to game specifications and equilibrium selection; if not, there is a normativity remainder that no equilibrium representation captures. Logged as OPEN in §05.10 and handed to Memo 10/11/12.

05.4 Six-part analysis — SOURCE: MARKET NORMS (norm from equilibrium)

Per JD Q2 every source is analyzed under six headings plus computational implications. "Market norms" here means the ought-relevant regularities a market generates: prices as signals, trade usages and customs of a trade, standards of merchantability, and the meta-norm "honor the equilibrium price / do not transact away from it without cause." We treat the market as a source that emits norms by aggregating decentralized action into an equilibrium.

05.4.1 Origin

INFERENCE Market norms originate from the aggregation of many local, self-interested decisions into a public signal. No agent authors the price; it is the fixed point of supply and demand. In LNS terms the source is the interaction structure itself (the exchange mechanism), and the norm is the emergent regularity that participants then treat as a datum ("the going rate," "the market standard"). INFERENCE Trade usages (e.g., grading conventions, delivery customs, standard settlement periods) originate as repeated coordination that hardens into expectation — a convention in the game-theoretic sense of §05.2.9.

FACT Positive law in many jurisdictions recognizes these emergent norms: commercial codes routinely incorporate "usage of trade" and "course of dealing" as gap-fillers and interpretive defaults. FACT that such doctrines exist in commercial law generally; we do not cite a specific code section to avoid inventing a reference. This is a documented instance of one source (legislation) explicitly importing the output of another source (market custom) — a hierarchy relation (§05.4.5).

05.4.2 Legitimacy (vs validity vs efficacy)

Keeping the §0.6 distinctions sharp:

• Legitimacy. INFERENCE Market norms draw legitimacy from a procedural claim: the price is legitimate because it emerged from voluntary exchange among consenting parties, not because any authority declared it. This is legitimacy-by-consent-and-emergence, structurally different from legitimacy-by-authority (Memo 01) or legitimacy-by-revelation (Memo 04). HYPOTHESIS The perceived legitimacy of a market price is contingent on a background belief that the exchange was voluntary and reasonably competitive; where that belief fails (coercion, monopoly, fraud, desperation), subjects stop treating the price as normatively binding and appeal upward to another source (law, morality) for correction. Stated as hypothesis because it is a claim about belief-states that is only partially observable.
• Validity. INFERENCE A market norm has no formal membership test analogous to legislative validity (enacted by body B under procedure P). Its "validity" is nearer to existence: does the regularity in fact obtain and is it recognized as a usage? This is a weaker, more empirical notion of validity than the legal one — a point the compiler team must not paper over.
• Efficacy. FACT Market norms typically exhibit high efficacy by construction: because deviating (paying far above market, or offering far below) is immediately costly, compliance is largely self-generating. Efficacy and enforcement collapse into the incentive structure.

INFERENCE Note the structural inversion relative to legislation. Legislation is strong on validity (clear membership test) and often weak on efficacy (laws routinely ignored). Market norms are weak on validity but strong on efficacy. This inversion is itself a candidate classifier for the taxonomy (Memo 10).

05.4.3 Enforcement mechanism

FACT The primary enforcement mechanism is incentive / loss, not sanction-by-authority. A trader who ignores the market price loses money, loses counterparties, or fails to clear inventory. INFERENCE Secondary mechanisms layer on top: - Reputation. Repeated-game enforcement (§05.2.9): a party who defects on trade usage is remembered and future-punished by withdrawal of dealing. INFERENCE - Private ordering. Merchant communities historically enforced trade norms through private tribunals, exchanges, and exclusion (guild discipline, exchange membership rules). FACT that such private commercial adjudication has existed historically; [INFERENCE that it functioned as an enforcement layer parallel to state courts.] - State backstop. When private mechanisms fail, parties escalate to courts, which enforce recognized usages via contract law (§05.5). INFERENCE

Market-norm enforcement stack (weakest coupling at top):

INFERENCE The compiler-relevant point: L0 is self-executing — it needs no detector and no adjudicator, because the consequence is intrinsic to the act. L1–L3 progressively require external machinery (memory, institutions, courts). A market norm can therefore be enforced at radically different cost depending on which layer engages. This maps onto Williamson's governance-structure choice (§05.2.7).

05.4.4 Mutation mechanism

INFERENCE Market norms mutate continuously and without amendment procedure. Prices update tick-by-tick as information arrives; trade usages drift as technology and practice change; a new dominant practice becomes the new usage once enough participants adopt it (a tipping/critical-mass dynamic). There is no promulgation event, no effective date, no repeal. HYPOTHESIS Mutation is path-dependent and lock-in-prone: once a standard (a technical standard, a settlement convention, a dominant platform) achieves critical mass, switching costs can lock it in even if a superior alternative exists — the position associated with the economics of QWERTY / path dependence (David, Arthur). Tagged hypothesis because the strength and prevalence of inefficient lock-in is itself contested (§05.3).

INFERENCE Contrast with legislative mutation (discrete, dated, authored) and precedent mutation (incremental, authored by courts). Market-norm mutation is continuous, unauthored, and distributed. This is the hardest mutation profile to represent in a versioned normative store, because there is no commit and no committer. Flagged for D1 (§05.11).

05.4.5 Hierarchy

INFERENCE Market norms sit in a definite but contested position in the inter-source hierarchy: - They are subordinate to mandatory law: price controls, antitrust/competition law, consumer-protection rules, and prohibitions (usury caps, bans on trading certain goods) override market equilibria. The state asserts a power to void or cap market outcomes; the market has a corresponding liability. FACT that such overriding legal regimes exist. - They are superior to (i.e., supply defaults for) private agreements only as default rules: trade usage fills contract gaps but is displaced by the parties' express terms. So within contract, express agreement > usage of trade > course of dealing (roughly). FACT that this lexical ordering of interpretive sources is standard in commercial law. - They interact laterally with morality and social norms: a market price can be legally valid and efficacious yet be judged illegitimate on moral grounds ("price gouging" in emergencies), triggering the escalation dynamic of §05.4.2. INFERENCE

05.4.6 Conflicts

INFERENCE Characteristic conflicts involving market norms: - Market vs mandatory law. Resolved in favor of law by the hierarchy above, but at an efficiency cost (Coase, §05.2.5). INFERENCE - Market vs morality. Unresolved in general; produces recurring political contestation (just price, exploitation, commodification limits — "things money shouldn't buy"). This is a genuinely open normative conflict, not a technical one. OPEN - Market vs market (multi-equilibrium). When several equilibria or standards coexist, coordination failures and standards wars occur; resolution is by tipping, not by ruling. INFERENCE - Signal corruption. Fraud, manipulation, information asymmetry (the "lemons" problem associated with Akerlof), and externalities cause the emergent norm to be wrong — the price fails to encode the true social costs/benefits. FACT that market failures of these kinds are established in economics. This is the market analogue of a corrupted source: the mechanism still emits a norm, but the norm mis-encodes reality.

05.4.7 Computational implications

INFERENCE Consolidated implications of treating market norms as a source: 1. Self-executing norms exist. Some norms need no detector and no adjudicator because the consequence is intrinsic (L0 above). D2 should have a category for norms whose enforcement is the payoff itself. INFERENCE 2. Norms without canonical text. A price or an unwritten usage has no authored surface. D1 must be able to represent a norm whose content is a behavioral regularity + expectation, not a string. OPEN — this is a hard representation problem. 3. Continuous, unauthored mutation. The versioning model cannot assume discrete authored amendments for this source. OPEN 4. Legitimacy is conditional and revocable by upstream sources. A market norm's binding force is contingent on background conditions (voluntariness, competition) whose failure routes control to law/morality. Any representation must carry these defeaters explicitly. INFERENCE 5. Corrupted signals must be representable. The model must distinguish "the norm the market emitted" from "the norm it would emit absent fraud/externality/asymmetry," or it cannot represent market failure at all. INFERENCE

05.5 Six-part analysis — SOURCE: CONTRACTS (norm from agreement)

A contract is the paradigm of norm-from-agreement: parties exercise a legal power to create norms binding on a closed subject-set. This is the most directly compilable source in the whole series, because it already has near-formal structure (offer, acceptance, terms, conditions, remedies) and an explicit textual surface.

05.5.1 Origin

INFERENCE A contract-norm originates from a meeting of wills operationalized in law as offer + acceptance + (in some traditions) consideration + intention to create legal relations. The parties supply the content; a background power-conferring rule of the legal system supplies the bindingness. This is the two-layer structure introduced in §05.2.3:

INFERENCE This factorization — a public power-conferring norm enabling private content-generation — is, in this memo's judgment, the single most important structural pattern for D1/D2, because it shows how a content-neutral kernel could host arbitrary private content: the kernel need only implement the power-conferring layer (the enforceability predicate and the remedy machinery); the content layer is data supplied from above the line. Flagged prominently at Handoff.

05.5.2 Legitimacy (vs validity vs efficacy)

Contract is the source where the three concepts separate most cleanly and are therefore most instructive:

• Validity. FACT Sharply defined and testable: a contract is valid iff it satisfies the legal formation conditions (capacity, consent free of vitiating factors, lawful object, requisite form/consideration). Validity is a predicate over the agreement, exactly the kind of thing a compiler checks. This is the cleanest validity notion in the series.
• Legitimacy. INFERENCE Contract's legitimacy rests on autonomy/consent: the parties bound themselves, so the obligation is self-imposed (pacta sunt servanda — agreements must be kept, treated as a near-universal background principle). FACT that pacta sunt servanda is a foundational maxim across legal traditions. Legitimacy fails where consent is defective (duress, fraud, unconscionability, gross inequality of bargaining power), which is exactly when law refuses enforcement — showing legitimacy and validity are wired together but distinct.
• Efficacy. INFERENCE High but not automatic: most contracts are performed voluntarily because of reputation, ongoing relationship, and the shadow of enforcement; a minority are litigated. Efficacy rests on the enforcement stack (§05.5.3).

INFERENCE Note that a contract can be valid but illegitimate (a technically-formed but grossly exploitative adhesion contract) or invalid but efficacious (an unenforceable gentleman's agreement that parties nonetheless honor via reputation). These cross-cases are the strongest single demonstration in the memo that validity, legitimacy, and efficacy are independent axes and must be represented as separate fields (they cannot be collapsed to one "bindingness" scalar). This is a candidate invariant for Memo 12.

05.5.3 Enforcement mechanism

INFERENCE Layered, similar to market norms but with a stronger formal apex: - L0 self-interest / relationship. Ongoing-relationship value and reputation make performance self-enforcing in repeated dealings (§05.2.9). INFERENCE - L1 self-help remedies. Contractually specified: withholding performance, set-off, liquidated damages, security interests, escrow, termination clauses. These execute without a court. FACT that such self-executing clauses are standard. - L2 private dispute resolution. Arbitration, mediation, industry tribunals — chosen by the parties themselves (a contract about how to enforce the contract). FACT - L3 state adjudication + remedies. Courts award expectation damages (the default remedy in common-law systems — put the promisee in the position as if performed), reliance or restitution damages, and exceptionally specific performance. FACT that expectation damages are the common-law default and specific performance is exceptional.

INFERENCE Compiler-relevant: contract enforcement is remedy-oriented, not punishment-oriented. The default consequence of breach is compensation calibrated to the promisee's expectation, not a sanction and not forced performance. This is a distinctive consequence-type: the "ought" is backed by "or else pay damages," which the tradition associated with Holmes framed as an option to breach and pay rather than an absolute duty. FACT that the Holmesian "option" reading exists; [HYPOTHESIS that contract duties are best modeled as duty-plus-liquidation rather than categorical obligation — relevant to how D1 types obligation.]

05.5.4 Mutation mechanism

INFERENCE Contract norms mutate through: - Amendment / variation. The same power that created the contract can modify it (by mutual agreement), sometimes requiring fresh consideration. FACT - Novation and assignment. Parties can be substituted or rights transferred, changing the subject-set. FACT - Interpretation. Courts fill gaps and resolve ambiguity using default rules, trade usage, and course of dealing (§05.4) — so the effective norm-set expands beyond the written text. FACT that gap-filling is a standard judicial function. - Frustration / force majeure / impossibility. Supervening events can discharge or suspend obligations — a conditional self-mutation triggered by world-states. FACT

INFERENCE Contract has the best-defined mutation mechanism of the economic sources: change is authored (by the parties or a court), often dated, and traceable. This makes it the friendliest source for a versioned normative store — the opposite of market norms (§05.4.4).

05.5.5 Hierarchy

INFERENCE Position in the inter-source hierarchy: - Subordinate to mandatory law. Contract terms contrary to statute or public policy are void or unenforceable (illegal contracts, unconscionable terms, terms waiving non-waivable rights). Mandatory rules > contract. FACT - Superior to default law for the parties. The defining feature of default (as opposed to mandatory) rules is that contract overrides them: "freedom of contract" means the parties can displace the law's defaults with their own terms. So for the closed subject-set: mandatory law > contract terms > default law. FACT that the mandatory/default distinction structures contract law this way. - Creates a private hierarchy. A master agreement can subordinate schedules and statements of work; a contract can incorporate external standards by reference, importing other sources. INFERENCE

Contract — hierarchical position (for the parties):

05.5.6 Conflicts

INFERENCE Characteristic conflicts: - Term vs term (internal). Resolved by construction rules: specific over general, negotiated over boilerplate, later over earlier amendment, and express hierarchy clauses ("in case of conflict, the Master Agreement prevails"). FACT that such construction canons exist. These are mechanical priority rules and are directly compilable. - Contract vs mandatory law. Resolved against the contract (§05.5.5). FACT - Contract vs contract (third parties). Conflicting obligations to different counterparties (e.g., selling the same unique good twice) cannot both be performed; law resolves via priority rules (first in time, registration, good-faith purchase) and damages to the loser. FACT that such priority regimes exist. - Incomplete-contract conflict. The world produces a state the contract did not foresee; the gap is filled by default rules and interpretation, which is where the parties' expectations may diverge. INFERENCE This is the practical face of the incomplete-contracts theory (§05.3.2): no contract can specify all future states, so residual control and gap-filling are unavoidable. FACT that contracts are necessarily incomplete over sufficiently rich state spaces is a standard result.

05.5.7 Computational implications

INFERENCE Consolidated: 1. The power-conferring / content-generating split is the master pattern. A content-neutral kernel can host arbitrary private norms by implementing only the enforceability predicate + remedy engine; content is data from above (§05.5.1). This is the strongest evidence in Memo 05 that a content-neutral kernel with an above-the-line content layer is coherent. INFERENCE — evidence, not proof. 2. Validity is a decidable predicate. Contract formation conditions are close to a checkable schema — the friendliest validity notion in the series for D1. INFERENCE 3. Validity ⟂ legitimacy ⟂ efficacy. The valid-but-illegitimate and invalid-but-efficacious cross-cases prove these are independent axes (§05.5.2). Represent as three fields, never one scalar. INFERENCE — candidate invariant. 4. Consequence-type is remedy, not sanction. D1 must support obligations backed by calibrated compensation (expectation damages), including the duty-plus-liquidation ("efficient breach") reading, distinct from prohibition-backed-by-punishment. INFERENCE 5. Incompleteness is structural. No norm-set over a rich enough world is complete; the representation must have first-class support for gaps and default/residual-control resolution, not treat gaps as errors. FACT that incompleteness is unavoidable; [INFERENCE that gap-handling must therefore be first-class.] 6. Priority/construction canons are compilable. Intra-contract conflict resolution (specific>general, later>earlier, express hierarchy clauses) is already an ordered rule system — a concrete model for the kernel's conflict-resolution component (cf. Memo 02 on legal priority rules). INFERENCE

05.6 Compressed six-part analyses — PROPERTY and THE FIRM

05.6.1 SOURCE: PROPERTY RIGHTS (norm from allocated entitlement)

• Origin. HYPOTHESIS /contested per §05.3.3 Either pre-political (first possession, labor-mixing) or state-constituted (bundle conferred by law). In practice, mixed: a first-possession or custom baseline that positive law then codifies and revises. INFERENCE
• Legitimacy. INFERENCE Legitimacy claims split along the same fault line: natural-rights legitimacy (it is mine because I made/found it) vs institutional legitimacy (it is mine because the recognized system says so). Efficacy of property, by contrast, rests almost entirely on enforcement.
• Enforcement. FACT Combination of self-help (fences, locks, exclusion), private ordering (title registries, insurance), and state force (trespass/conversion actions, police, ejectment). Registration systems are a documented institutional device for lowering the transaction cost of verifying entitlement. INFERENCE that registries are transaction-cost-reducing, per Coase/North.
• Mutation. FACT Transfer (sale, gift, inheritance), subdivision, encumbrance, adverse possession (mutation by lapse of time + use — a rare unauthored mutation in law), expropriation/eminent domain (state override), and reform (redefinition of the bundle by legislation). INFERENCE that adverse possession is structurally interesting: a norm that mutates the entitlement through mere efficacy over time — possession hardening into title.
• Hierarchy. FACT Constitutional protection (takings/expropriation clauses) sits above ordinary property law, which sits above private arrangements (leases, servitudes). The state retains an override (eminent domain) but often owes compensation — a power coupled with a liability to compensate. INFERENCE
• Conflicts. INFERENCE Boundary disputes, competing claims (resolved by priority/registration), property-vs-property (nuisance: my use harms your use), and property-vs-public (regulation vs takings). The nuisance case is the Coasean reciprocal-harm problem (§05.2.5): both uses are "causes," and the question is which entitlement to assign.
• Computational implications. INFERENCE (i) Property is the cleanest Hohfeldian object (§05.2.2) and a natural test case for a typed-relation representation. (ii) Partial interests as subsets give set operations real meaning. (iii) Time-indexed and defeasible interests (future interests, life estates, defeasible fees, adverse possession) require temporal and defeasibility support in D1. (iv) The state-override-with-compensation pattern is a reusable template (a power + correlative liability-to-compensate).

05.6.2 SOURCE: THE FIRM / HIERARCHY (norm from authority-substituting-for-price)

• Origin. FACT /Coase The firm arises where the transaction cost of coordinating via market exchange exceeds the cost of coordinating via authority; inside the firm, the price mechanism is suspended and resource direction happens by command. INFERENCE The firm is thus a source that emits norms by fiat within a bounded domain — the employment relation converts a series of market transactions into a standing authority relation.
• Legitimacy. INFERENCE Derives from the consent embedded in the employment/agency contract (you agreed to accept direction within a zone) plus ongoing exchange (wage). Legitimacy is bounded: authority runs only over the contracted "zone of acceptance" (a notion associated with Barnard/Simon). Orders outside the zone are not legitimate and can be refused. FACT that the "zone of acceptance/indifference" concept exists in organization theory.
• Enforcement. INFERENCE Internal: monitoring, performance management, promotion/demotion, dismissal (the ultimate internal sanction). External backstop: labour law caps the firm's authority (non-waivable worker protections). Note enforcement is hierarchical and managerial rather than price-mediated — the whole point of the firm (§05.2.7).
• Mutation. FACT Reorganization, policy issuance, restructuring of reporting lines; far faster and more centralized than market-norm mutation. This is the authored, centralized mutation profile — the opposite pole from §05.4.4. Detailed in Memo 07.
• Hierarchy. INFERENCE Firm norms sit below law (labour, competition, corporate) and below the enabling contract, but above the individual employee within the zone of acceptance. The firm is simultaneously a subject (of external norms) and a source (of internal norms) — the rule/player duality of §05.1.
• Conflicts. INFERENCE Order-vs-law (an unlawful instruction: law wins, and refusal is protected), order-vs-order (conflicting instructions up the chain: resolved by hierarchy/escalation), and principal-agent conflict (the agent's incentives diverge from the principal's — the core problem of agency theory). FACT that principal-agent divergence is a foundational result.
• Computational implications. INFERENCE (i) Authority/fiat is a norm-generation mode distinct from agreement and equilibrium — the same substantive ought can be produced by command, contract, or price (Williamson, §05.2.7), confirming content/mechanism separability. (ii) Bounded authority ("zone of acceptance") is a scoped power — a power valid only over a delimited condition-set; D1 needs scoped, revocable powers. (iii) Principal-agent divergence means an enforcement mechanism can be gamed by the very agents meant to execute it — a hostile-emergence concern (M3) handed to D2.

05.7 SOURCE: THE COMMONS (norm from polycentric self-governance) — six-part

Ostrom's work (§05.2.8) is treated as its own source because it exhibits a norm-generation mode — self-governance by the affected users — distinct from market, contract, firm, and state.

• Origin. FACT Rules originate from the resource users themselves, through repeated interaction and explicit collective-choice processes, adapted to local ecological and social conditions. Not imposed from above; not the byproduct of price. INFERENCE The source is a bounded community exercising a collective power to bind its own members.
• Legitimacy. INFERENCE Strongly participatory: legitimacy derives from the fact that those bound by the rules participated in making them (design principle 3) and can revise them (self-determination). This is closest to democratic legitimacy (Memo 01) but at sub-state, community scale. Monitoring by accountable monitors (principle 4) sustains perceived fairness, which sustains legitimacy.
• Enforcement. FACT Graduated sanctions (principle 5): first violations draw small penalties, escalating with repetition/severity, administered by community monitors, with low-cost conflict resolution (principle 6) as a release valve. INFERENCE This is a documented, empirically-successful enforcement design and is directly relevant to D2: it shows that proportional, escalating, community-administered sanctions with cheap appeal outperform both no-enforcement and maximal-penalty enforcement in sustaining cooperation.
• Mutation. FACT Collective-choice arrangements (principle 3) are the mutation mechanism: users who are subject to the rules have standing to change them, typically in local, low-formality processes. INFERENCE Mutation is endogenous and participatory — the governed are the amenders.
• Hierarchy. FACT Principle 7 (minimal recognition of the right to organize) and principle 8 (nested enterprises / polycentricity) define the hierarchy: the larger political system must at least not undermine the local source (an immunity granted from above), and governance is layered — local rules nested within regional within national, each level with its own sphere. INFERENCE Polycentricity is the key structural import: multiple centers of decision-making with overlapping jurisdiction and no single apex, which is a genuinely different topology from the single-rooted hierarchy of JD Q4's politics→…→KPI chain.
• Conflicts. INFERENCE Appropriation conflicts (who takes how much) resolved by monitoring + graduated sanction; rule-interpretation conflicts resolved by cheap local conflict-resolution; cross-level conflicts (local rule vs regional/national law) resolved by the nesting arrangement, imperfectly. Ostrom's cases document that success is fragile and can collapse when any principle fails — especially when external authority overrides local rules (violating principle 7).
• Computational implications. INFERENCE (i) Polycentric (multi-root) topology must be representable — the dependency graph (JD Q4) cannot assume a single sovereign root; overlapping jurisdictions with no apex are real and successful. OPEN whether a compiler can resolve conflicts without a single apex — handed to Memo 08/12. (ii) Graduated, community-administered sanctions with cheap appeal are a concrete enforcement template for D2. (iii) The design principles constitute an empirically-validated checklist for a well-formed self-governing source; each maps to a compiler concern (boundaries→subject-set, congruence→condition-fit, collective-choice→mutation, monitoring+sanctions→enforcement, conflict-resolution→dispute sub-system, recognition→inter-source immunity, nesting→composition). (iv) Ostrom is direct evidence against the assumption that governance requires centralization — held here descriptively; its adversarial edge is Memo 11's (§05.9).

Polycentric (Ostrom) vs monocentric (JD Q4) topology:

05.8 SOURCE: PUBLIC CHOICE and GAME-THEORETIC NORMS

05.8.1 Public choice (politics as exchange)

FACT Public choice models collective decision-making with economic tools (§05.2.10). As a source characterization: it says the norms emitted by the political process are shaped by the incentives of the agents who operate it — voters (rationally ignorant), legislators (re-election-seeking), bureaucrats (budget/scope-seeking), and interest groups (rent-seeking). FACT that these are the standard behavioral posits; their empirical strength varies.

INFERENCE Governance-relevant imports: - Concentrated benefits, dispersed costs predicts a systematic bias in which norms get produced: policies with concentrated beneficiaries and diffuse costs are over-supplied. Any system that models norm-production (not just norm-content) must expect this bias. INFERENCE - Rent-seeking means resources are spent competing to capture the norm-generation process itself — the enforcement/production machinery is an attack surface. A direct M3 (hostile emergence) input for D2. INFERENCE - Median-voter and log-rolling models give (partial, assumption-heavy) accounts of which equilibrium the political market selects. HYPOTHESIS — heavily conditional on model assumptions. - Constitutional political economy (the "rules about rules" level, associated with Buchanan) distinguishes the constitutional stage (choosing the rules of the game, ideally behind a veil of uncertainty) from the in-period stage (playing the game). FACT that Buchanan drew this two-level distinction. This maps precisely onto the power-conferring / content-generating split (§05.5.1) and onto the meta-rule / rule distinction the compiler needs.

OPEN Whether public choice's characteristic conclusions are neutral science or carry a prior (§05.3.4) is unresolved and is not adjudicated (political neutrality). The compiler team should import the mechanisms (incentive-responsiveness, capture, rent-seeking as attack surface) and quarantine the conclusions (whether state or market "fails" more) as above-the-line content.

05.8.2 Game-theoretic norms (conventions, focal points, repeated games)

Building on §05.2.9. Four norm-generation sub-modes, each a distinct computational behavior for the taxonomy (Memo 10):

• Convention (coordination-equilibrium norm). FACT In a pure coordination game with multiple strict equilibria, any equilibrium is self-enforcing; a convention is a persistent selection among them. Enforcement cost ≈ 0 (deviation hurts the deviator immediately). Examples: measurement units, language, side-of-road. Mutation: rare, tipping-driven, with lock-in (§05.4.4). INFERENCE
• Focal point (salience-selected norm). FACT /Schelling When multiple equilibria exist, players converge on the salient one. INFERENCE Salience is frequently supplied by shared culture, history, or prominence — i.e., by content above the kernel. Consequence for the kernel boundary: equilibrium selection is not always computable from the payoff matrix alone; it can require above-the-line cultural content. This is a clean, specific limit and is flagged for Memo 11/12.
• Repeated-game cooperation (folk-theorem norm). FACT Cooperation unsustainable in one shot becomes an equilibrium under repetition + sufficient patience (discount factor high enough) + observability. INFERENCE The "ought to cooperate" here is manufactured entirely from structure (future shadow), with no moral content — evidence for the reductive view (§05.3.5) for at least some norms.
• Reciprocal strategies (tit-for-tat class). FACT Axelrod's tournaments found simple reciprocal strategies (be nice, be provocable, be forgiving, be clear) robust. FACT with caveat: robustness is relative to the field of competitors and to noise; tit-for-tat degrades under misperception/noise, where more forgiving variants do better. INFERENCE Governance import: a decentralized enforcement rule (mirror your counterpart) can sustain cooperation with no central enforcer, but is fragile to noise — a warning that decentralized incentive-enforcement (§05.4.3 L0–L1) needs error-tolerance built in (M3/M5).

INFERENCE Cross-cutting computational implication: game theory supplies the formal backbone for every "norm-without-an-author" source in this memo (market, convention, commons cooperation). It gives D1/D2 a candidate formal semantics for a whole class of norms: a norm as (game specification, equilibrium selected, common knowledge of it). The unresolved question is whether that triple captures the normativity or only the behavioral regularity (§05.3.5, OPEN).

05.9 The Hayekian knowledge / spontaneous-order argument (STATED as setup for Memo 11)

This section states the position associated with F. A. Hayek and does not rebut, defend, or adjudicate it. Its adversarial deployment — as an argument that governance may not be centrally computable — is reserved for Memo 11, per the conventions (§0.2, §0.3). It is presented here because it is the natural culmination of the "norm from equilibrium" thread and because the compiler team must see it in its economic home before meeting it as a threat.

05.9.1 The knowledge problem

FACT — as a summary of Hayek's stated argument The core claim (most associated with "The Use of Knowledge in Society," 1945) is that the knowledge relevant to economic coordination is: - dispersed — held in fragments by millions of individuals; - local and particular — knowledge "of the particular circumstances of time and place"; - tacit — much of it cannot be articulated, only exercised in action; and - dynamic — continuously changing.

Because of this, FACT — as Hayek's claim no central mind or central computer could ever assemble the knowledge required to allocate resources as well as the decentralized price system does. The price system is, on this view, a communication and computation mechanism that solves a problem no central agent could solve, precisely because it does not require anyone to possess the totality of knowledge: each agent acts on local knowledge and a few prices, and the system coordinates without central assembly.

05.9.2 Spontaneous order (kosmos vs taxis)

FACT — as a summary Hayek distinguishes spontaneous order (kosmos: order that emerges from many agents following abstract rules, e.g., the market, common law, language, money) from made order (taxis: order deliberately arranged by a designing mind, e.g., an organization, a command economy). The associated claim is that spontaneous orders can achieve a complexity and adaptiveness that designed orders cannot, because they harness dispersed knowledge that no designer can centralize; and that the rules underpinning spontaneous order are typically the product of cultural evolution (selection among group practices) rather than of deliberate design.

05.9.3 The calculation problem (the socialist-calculation lineage)

FACT — as a summary Related and older is the economic calculation argument (associated with Mises and carried forward by Hayek): without market prices for the means of production (which require private property and exchange in them), a central planner has no way to compute the relative scarcities needed for rational allocation — not merely a hard computation, but the absence of the very data (prices) that make the computation meaningful.

05.9.4 Why this is the memo's most important handoff to the adversary

INFERENCE The reason this belongs in the Institutional Economics memo, stated neutrally: - If the knowledge problem is correct, then a centralized computable governance that tries to compute allocations or optimal norms directly would fail for the same reason central planning is argued to fail — the relevant knowledge is dispersed, tacit, and dynamic and cannot be assembled into a central representation. [Stated as the structure of the challenge, not endorsed.] - Crucially, this is not obviously an argument against all computable governance. It bears most sharply on the content-computing ambition (computing what the norms/allocations should be) and least sharply on the content-neutral kernel ambition (executing norms supplied from above). INFERENCE — but this distinction is exactly what Memo 11 must test destructively (M5), not something to be comforted by here. - The focal-point result (§05.8.2) is a specific, formal echo of the same worry: even in pure coordination, the selected equilibrium can depend on non-computable-from-payoffs cultural salience. Hayek generalizes this to the whole economy.

OPEN The precise scope of the knowledge/calculation argument against Computable Governance — total, partial, or only against a particular (centralizing, content-computing) design — is left OPEN here and is the central business of Memo 11. This memo asserts only that the argument exists, is serious, and is the strongest single economic challenge to the enterprise.

05.10 OPEN QUESTIONS

Unresolved problems this memo cannot settle; D1/D2 must treat them as open.

1. OPEN The normativity remainder (§05.3.5). Does "norm = (game, equilibrium, common knowledge)" capture obligation, or only its behavioral shadow? If a deontic residue survives the reduction, no equilibrium representation is complete. This is the deepest open question in the memo and gates whether large classes of norms are compilable at all.
2. OPEN Norms without canonical text (§05.4.7). How does D1 represent a norm whose content is a behavioral regularity + expectation, with no authored string? The whole market-norm and convention family depends on solving this.
3. OPEN Continuous, unauthored mutation (§05.4.4). The versioning/provenance model assumes commits and committers; market norms have neither. What is the data model for a norm that drifts continuously with no amendment event?
4. OPEN Polycentric conflict resolution without an apex (§05.7). Ostrom's successful commons and overlapping jurisdictions have no single sovereign root. Can a compiler resolve norm-conflicts in a multi-root graph, or does conflict-resolution require a distinguished apex? (Shared with Memo 08.)
5. OPEN Property's pre-political status (§05.3.3). Is property an input to the normative system (given/protected) or an output (defined/revisable)? The compiler cannot silently pick a side without importing ideology (JD Q8).
6. OPEN Scope of the Hayek knowledge/calculation argument (§05.9.4). Total, partial, or design-specific? Deferred to Memo 11.
7. OPEN Is the firm a distinct norm-type or reducible to contracts (§05.3.2)? Whether "authority/fiat" is primitive or decomposable into a nexus of contracts determines whether D1 needs a primitive "command" modality or can synthesize it. (Shared with Memo 07.)
8. OPEN Enforcement-cost as a first-class quantity. Coase/Williamson show enforcement is costly and the cost drives institutional form. Should D1/D2 attach a cost model to every enforcement mechanism, and if so, in what units, given cross-domain incommensurability?

05.11 RESEARCH OPPORTUNITIES

Concrete places D1/D2 work could contribute or must decide.

1. Typed Hohfeldian bundle library, tested on property. Property (§05.2.2, §05.6.1) is the cleanest Hohfeldian object; build the typed-relation + set-algebra representation there first, including temporal/defeasible interests (future interests, life estates, adverse possession) as the stress cases. High payoff, well-scoped.
2. Power-conferring / content-generating compiler skeleton. Prototype the two-layer contract pattern (§05.5.1): a content-neutral enforceability predicate + remedy engine (kernel-side) hosting arbitrary party-supplied terms (above-line data). This is the memo's best candidate demonstration that a content-neutral kernel can host arbitrary content.
3. Three-axis bindingness representation. Implement validity, legitimacy, and efficacy as independent fields, motivated by the contract cross-cases (§05.5.2). Verify against the valid-but-illegitimate / invalid-but-efficacious examples as regression tests.
4. Graduated-sanction enforcement module (Ostrom). Encode design principles 4–6 as a reusable enforcement template: accountable monitoring + graduated sanctions + low-cost appeal. Test whether it sustains cooperation better than flat-penalty schemes in a repeated-game simulation (ties to §05.8.2).
5. Equilibrium-semantics experiment for conventions. Represent a convention as (game, selected equilibrium, common-knowledge predicate) and test where it fails to capture the norm (focal-point/cultural-salience cases, §05.8.2) — a direct, falsifiable probe of the §05.3.5 open question (M1/M5).
6. Noise-tolerant decentralized enforcement. Given tit-for-tat's fragility under noise (§05.8.2), design and test forgiving reciprocal enforcement for D2's self-executing-norm layer (§05.4.3 L0–L1).
7. Rent-seeking / capture red-team (M3). Treat the norm-production pipeline itself as an attack surface per public choice (§05.8.1); build hostile-emergence tests where agents optimize to capture rather than obey.
8. Multi-root (polycentric) graph model. Prototype a normative dependency graph with overlapping jurisdictions and no apex (§05.7); measure whether conflict-resolution degrades gracefully or requires an injected apex (feeds Memo 12's interface question).

05.12 Handoff to D1/D2

Concrete computational implications for the compiler architects, distilled from this memo. Each is tagged so the team can see what is load-bearing vs speculative.

1. Adopt the power-conferring / content-generating split as the master hosting pattern. INFERENCE — strongest structural result here A content-neutral kernel can host arbitrary private norms by implementing only the enforceability predicate + remedy machinery (Layer 1); all value-laden content is above-line data (Layer 2). Contract (§05.5.1) is the proof-of-concept; constitutional political economy (§05.8.1) shows the same meta-rule/rule split at the political level.

2. Represent validity, legitimacy, and efficacy as three independent fields — never one bindingness scalar. INFERENCE — supported by contract cross-cases §05.5.2 Valid-but-illegitimate and invalid-but-efficacious norms both exist and must be distinguishable.

3. Support norms with no canonical text. OPEN Market prices, trade usages, and conventions are norms without an authored string (content = behavioral regularity + expectation). D1 needs a representation for these or it cannot host the entire norm-from-equilibrium / norm-from-convention family.

4. Support multiple mutation profiles as first-class. INFERENCE At minimum: authored-discrete-dated (legislation, contract amendment), authored-centralized-fast (firm reorganization), participatory-endogenous (commons collective-choice), and continuous-unauthored-distributed (market drift, convention tipping). The last has no commit/committer and breaks naive versioning OPEN.

5. Model enforcement as a layered stack with attached cost, including self-executing norms. INFERENCE Some norms are enforced by intrinsic payoff (L0: no detector, no adjudicator needed); others escalate through reputation, private ordering, and courts at rising cost (Coase/Williamson, §05.4.3, §05.5.3). Enforcement cost is part of the design space, not free.

6. Provide scoped, revocable powers (bounded authority). INFERENCE The firm's "zone of acceptance" (§05.6.2) and the state's compensated eminent-domain power (§05.6.1) are powers valid only over a delimited condition-set, sometimes coupled to a correlative liability (to compensate). D1 needs scoped powers and power+liability couplings.

7. Make consequence-types plural: sanction and calibrated remedy. INFERENCE Contract's default consequence is expectation damages (compensation calibrated to the promise), not punishment or forced performance; the Holmesian duty-plus-liquidation reading (§05.5.3) must be expressible alongside prohibition-backed-by-punishment.

8. Treat incompleteness and gaps as first-class, not errors. FACT that contracts/norm-sets over rich state spaces are necessarily incomplete D1 must have native support for gaps + default/residual-control resolution (§05.5.7).

9. Provide compilable conflict-resolution canons. INFERENCE Intra-contract construction rules (specific>general, later>earlier, express-hierarchy clauses) and property priority rules (first-in-time, registration, good-faith purchase) are already ordered rule systems — concrete models for the kernel's conflict component.

10. Support polycentric (multi-root) topology and its open conflict problem. OPEN Ostrom's commons and overlapping jurisdictions have no single apex (§05.7). Whether conflict-resolution is possible without a distinguished root is unresolved and shared with Memo 08.

11. Inherit Arrow's impossibility as a hard formal limit on preference-aggregation. FACT Any component that computes a collective preference ordering from individual orderings cannot simultaneously satisfy unrestricted domain, Pareto, IIA, and non-dictatorship. This is a theorem, logged alongside the Gödel/Church/Turing limits of Memo 11.

12. Quarantine equilibrium-selection content that is not computable from payoffs. INFERENCE Focal-point salience (§05.8.2) can require above-the-line cultural content; the kernel must not assume equilibrium selection is always derivable from the game itself.

13. Build hostile-emergence tests for capture of the norm-production pipeline. INFERENCE , M3 Public-choice rent-seeking (§05.8.1) and principal-agent divergence (§05.6.2) mean the enforcement/production machinery is itself an attack surface; test agents that optimize to capture rather than comply.

14. Carry the Hayek knowledge/calculation argument forward untouched. OPEN The dispersed/tacit/dynamic-knowledge and calculation arguments (§05.9) are the strongest economic challenge to centralized, content-computing governance. This memo neither rebuts nor endorses them; Memo 11 must test the enterprise against them destructively (M5). The tentative (and to-be-attacked) observation is that the challenge bites hardest on content-computing and least on content-hosting designs — but that comfort is exactly what Memo 11 must not grant prematurely.

End of Memo 05 — Institutional Economics (LNS series). Cross-references: Memo 01 (political norm-generation, Arrow), Memo 02 (legal priority/conflict rules), Memo 06 (conventions, social norms, Luhmann), Memo 07 (firm-internal governance, agency), Memo 08 (polycentric/international conflict without apex), Memo 09 (institutional/cultural evolution), Memo 10 (taxonomy: norm-from-equilibrium / -agreement / -authority / -self-governance), Memo 11 (Hayek and the falsification program), Memo 12 (invariant interface).