Exit to Protocol

What governance functions can migrate?

Look at the state the way an engineer looks at an inherited system, with the flags and anthems temporarily stripped away. Modern states are not frozen products of 1648; they are layered institutions shaped by war, taxation, democratization, administration, rights movements, markets, and technology. They nevertheless retain a distinctive territorial claim: final authority to make and enforce public rules within a jurisdiction. That architecture solved real coordination and security problems while institutionalizing coercion. Engineering language helps compare mechanisms, provided it does not pretend that law, legitimacy, identity, and physical force are merely software modules.

Everything else built in the seventeenth century has been replaced several times over. The state’s architecture has not. It still concentrates decision-making into hierarchies designed around scarcity, distrust, and obedience. It still optimizes for control rather than competence, because control is what a monopoly must maintain and competence is what competition would have forced. Its internal workings are still opaque to the people who fund it, and the funding is still not optional. Above all, it is non-forkable: if you think your government’s justice system is badly designed, you cannot take a copy of the good parts and run an improved version alongside it. You can vote, emigrate, or submit.

Global connectivity, cryptographic verification, and programmable contracts expand the space of coordination that does not require a territorial administrator for every transaction. The Conversion of Coercion framed civilization as a partial conversion of force into rule-governed coordination. The chapters since have not “cleared away” the objections: defense remains open, physical monopolies constrain exit, and protocols create their own maintainers, dependencies, and enforcement boundaries. This chapter asks which functions can migrate, under what conditions, and where the analogy breaks.

The Architecture of a Non-Coercive Order

Call the proposal protocolized governance: making some rules modular, auditable, interoperable, and open to competing implementations. Where appropriate, consortia can coordinate by contract and software can make performance easier to verify. Law does not become logic, and trust does not become code; interpretation, identity, bugs, remedies, force, and authority remain social and institutional.

A world without coercive government is not anarchy in the pejorative sense; it is post-political order — governance by consent rather than compulsion. The institutions of civilization remain entirely recognizable: law, defense, infrastructure, education, welfare, money. What changes is their method of organization and their moral foundation. Instead of monopolies enforced by violence, each function is provided by interoperable, reputation-driven networks that compete for trust rather than obedience. Authority becomes a function of reliability, not decree.

Laid out function by function, the candidate stack looks like this:

Two rows deserve a note. Defense is the hardest case, which is why it received the previous chapter rather than a table cell. Cryptographic currencies demonstrate global settlement outside a single operator, but states still regulate exchanges, custody, taxation, issuers, network access, and users; protocol persistence is not immunity from governance. The more a function depends on land, bodies, emergency response, universal access, or physical force, the less a digital analogy establishes.

The funding column tests whether contracts and exit can replace compulsory levies for particular functions. It cannot assume the answer: free-riding, adverse selection, poverty, network monopoly, captive infrastructure, and third-party harms may make nominally voluntary arrangements incomplete or coercive in practice. Competition and reputation can discipline providers when entry, information, liability, and exit are real. They complement rather than automatically replace regulation, and they do not render corruption unprofitable when users are locked in or the corrupt provider controls enforcement.

Measurable, Comparable, Forkable

The dynamics of such an order follow from four structural features.

Private consortia. Coalitions of specialized organizations provide justice, security, and infrastructure, competing on efficiency, reliability, and ethical clarity. Because membership is voluntary, performance becomes the only durable source of legitimacy — there is no flag to hide behind and no border to trap the dissatisfied.

Interoperable protocols. Common standards replace national regulation. Verification replaces trust; algorithmic transparency replaces bureaucratic opacity. Dispute resolution becomes a function of predictable computation rather than political bargaining. The standards themselves must be neutral infrastructure — carrying whatever their users route through them, the way a free-speech order carries speech its operators despise — a requirement whose architecture Civilizational Infrastructure has already built for the hardest case.

Voluntary jurisdiction. Individuals and groups affiliate with governance providers the way they affiliate with ISPs or insurers. Exit costs stay low, so accountability is enforced by freedom of association rather than by the fiction that a ballot cast every four years disciplines a monopoly.

Reputation and proof. Metrics of service quality, reliability, and fairness are auditable in real time.

Add these up and governance acquires three properties it has never had in the political stack: it becomes measurable, comparable, and forkable. Measurable, because a provider’s performance is a matter of auditable record rather than campaign rhetoric. Comparable, because multiple providers of the same function operate side by side, and their records can be read against each other — something territorial monopolies structurally prevent, which is why arguments about whether a state governs well never end. And forkable, because a governance system defined in protocol is a system whose rules can be copied, modified, and relaunched by anyone who thinks they can do better. The fork is the deepest of the three properties. It converts institutional reform from a political struggle — capture the monopoly or fail — into a market test: run the improved version and see who migrates.

This suggests a non-revolutionary test: let contestable alternatives take over functions where they outperform and preserve agency, while retaining or redesigning public institutions where coercive coordination still survives the comparative burden. Whether the state shrinks, changes, or remains essential in a domain is an outcome to discover, not a historical script.

The Transition Path

Any transition toward more protocolized governance would likely be incremental and uneven: some contracts arbitrated privately rather than litigated, some payments settled on new rails, some credentials or communities made portable. Constitutional change, public authorization, resistance, failure, and hybrid institutions would remain part of the path. “Voluntary” options must earn adoption and cannot be presumed better merely because they are new.

Experiments begin most easily in digital ecosystems and borderless commerce, where parallel implementation can be cheaper. Yet users have bodies, assets, counterparties, and network access within legal jurisdictions, so enforcement never loses all purchase. Decentralized arbitration, crypto-finance, and open-source governance are running systems, not demonstrated general successors. Some expose rules and records unusually well; they also face exploits, concentrated custody or voting, limited remedies, regulatory dependence, and selection effects. Comparative performance needs defined outcomes and baselines.

As people experience the superior performance of voluntary systems, legitimacy begins to migrate. This is the quiet variable that decides everything. States do not run on force alone — force is expensive, and no state can afford to compel everyone at once. They run on the presumption that their coordination is necessary. Every voluntary system that visibly outperforms its coercive counterpart erodes that presumption a little further. Governments, deprived of their monopoly on coordination, will retreat to the functions they still perform well — then to none at all. The political apparatus does not collapse; it fades into irrelevance, the way institutions actually die: not repealed but routed around, maintained long after anyone remembers relying on them.

The analogy to science displacing some religious explanations is suggestive but limited: politics allocates authority and force rather than merely explaining observations. The direction is not guaranteed. Protocols can widen cooperation and exit; they can also centralize surveillance, automate exclusion, and entrench rules that are difficult to appeal. Migration is a design hypothesis to test, not a law of history.

The Migration Is Already Visible

This would be comfortable speculation if legitimacy migration were merely predicted. It is observable now, and the clearest evidence is what is happening to classical liberalism — the political tradition of individual freedom, universal rights, free markets, and limited government.

Geographically, that project is in undeniable crisis, and the diagnosis eigenrobot has pressed¹ is hard to refute: the United States is already postliberal in practice, contested by two parties each hostile to liberal foundations from opposite directions — one rejecting neutrality, universalism, and merit in favor of identity frameworks, the other drifting toward centralized, personality-driven nationalism. The civic institutions that once transmitted a liberal identity have collapsed or turned into ideological battlegrounds. Europe’s technocratic managerial state feeds populist backlash; Britain regulates speech with an appetite that would have embarrassed its own censors a century ago. Anyone who loves liberty and looks only at territory sees a rout.

But territory is the wrong place to look, and Balaji Srinivasan supplies the missing half of the picture:² liberalism has not died; it has migrated to the Internet. Open exchange, voluntary association, strong privacy, global cooperation — the substance of the liberal package — are being rebuilt in a medium where they are secured by cryptography, blockchains, and smart contracts rather than by parchment and the goodwill of officials. Internet-native communities like the Bitcoiners are the first of many nations united not by geography but by shared values, a shared financial system, and shared governance — and unlike state liberalism, cloud liberalism is decentralized enough that no government can dismantle it by capturing a capital.

The pessimism and the optimism are not rivals; they are complementary halves of a single event. Eigenrobot accurately diagnoses liberalism’s geographic and cultural collapse. Balaji identifies its simultaneous digital rebirth. Put together, they describe exactly what the transition path predicts: legitimacy draining from coercive institutions that no longer perform, and pooling in voluntary systems that do. Liberalism’s geographic core is collapsing while its ideological core is re-implemented on infrastructure that is harder to seize than any constitution. The rights that fragile political institutions once guaranteed on paper are becoming properties that protocols guarantee by construction. As state liberalism retreats, cloud liberalism expands. This is not liberalism’s demise. It is an exit to protocol.

The same migration can be watched from the other side — in what has happened to the old knowledge class whose word once moved the political world. After World War II, intellectual elites were installed as civilization’s moral arbiters. Philosophers, historians, and legal scholars drafted the conventions that defined crimes against humanity; Nuremberg, the Genocide Convention, and the Universal Declaration of Human Rights all bear the fingerprints of academic expertise. To speak as a scholar was to borrow the moral capital of civilization itself. That capital has been spent — not because the subjects stopped mattering, but because the class holding it traded the posture of impartial guardianship for the role of partisan combatant. Humanities departments became one more faction in the culture war, and factions do not get to arbitrate.

So when a leading scholarly association now passes a solemn resolution declaring some state’s conduct a crime, the response that would once have been grave silence is a shrug. The declaration is not dead — courts still cite scholarly consensus, activists still build campaigns on resolutions, and the archives will preserve the record for a century of future judgment. What has died is command of the present. The scholar’s authority has been temporalized: it no longer sways the living; it haunts the future, an inscription in the fossil record of civilization rather than an intervention in its affairs. The scholar’s declaration is no longer a thunderclap; it is a press release.

That is what the twilight of an authority regime looks like, and it is worth staring at, because the state’s own legitimacy rests on the same mid-century foundations. Authority derived from institutional position — the credential, the office, the letterhead — is being repriced everywhere at once, and the repricing runs in one direction: away from trust me, I hold the position and toward verify me, here is the record. That is precisely the trade the protocol stack formalizes. Reputation systems and auditable performance are the institutional form of a civilization that has stopped taking letterhead as evidence.

The Honest Caveat

Nothing in this chapter should be read as the claim that voluntary at birth means voluntary forever. It does not, and the failure mode is already visible in miniature.

The moral engine of the entire replacement stack is exit. Every discipline described above — competition, reputation, the fork — works only so long as leaving a provider is genuinely possible at bearable cost. But network effects are a form of gravity. A dominant protocol, platform, or consortium accumulates users precisely because everyone else is there; and past some threshold, the alternative to membership stops being a competing provider and becomes exile from the network where economic and social life happens. At that point exit is nominal — formally permitted, practically ruinous — and a system with voluntary paperwork begins to reproduce the structural position of the state it replaced, no tax collector required. A guild that controls the only certification that employers recognize, a settlement network too dominant to leave, a reputation system whose scores follow you everywhere: each is a monopoly waiting to notice its own leverage.

So the fork-and-exit architecture is not a property that voluntary systems get for free; it is a property they must be engineered to preserve, against their own success. What exactly must be preserved — portability of assets, supremacy of exit, isolation of capability — is the design problem of The Limits of Leviathan; and the criterion for judging when any system, coercive or voluntary, has drifted past the boundary where agency remains viable is drawn precisely in The Admissible Region. Plant the flag here: the argument of this volume was never voluntary good, state bad as a taxonomy of labels. It is a claim about structure, and structure can rot under any label.

With that caveat entered, a narrower thesis stands. Converting some discretionary power into transparent, contestable protocol can reduce opportunities for domination when affected agents retain voice, remedy, and realistic exit. Constitutions, markets, firms, states, commons, and code can each convert violence into rules in some settings and conceal coercion in others. The measure is not how little politics remains, but how much unavoidable authority is traceable, bounded, corrigible, and answerable to the agents it affects.

The end of politics is not chaos — but it is not a destination either, only coordination we build to need less coercion.