---
title: "Libertarianism Is an Incomplete Solution"
subtitle: "The truncated gradient: correct diagnosis, wrong prescription"
author: Elias Kunnas
description: "Libertarianism asks how much state to tolerate. That is not the whole design space. State scope and feedback quality are different variables."
canonical: https://kunnas.com/articles/libertarianism-incomplete-solution
url: https://kunnas.com/articles/libertarianism-incomplete-solution.md
date_published: 2026-02-16
date_modified: 2026-08-31
corpus_frame_url: https://kunnas.com/articles/how-to-read-this.md
---
## How to read this corpus

The corpus applies one lens to many domains: what mechanisms produce the outcome? It shares four methodological commitments and one explicit directional commitment. Each linked page argues for its part; the links are derivations and disputes, not evidence inherited by every page. The directional commitment does not by itself settle system boundary, distribution, sacrifice, or institutional authority.

1. **Mechanisms are what act.** Incentive gradients, selection pressures, feedback loops, and capital stocks produce the distribution of outcomes. Intentions, labels, official categories, and stated values are evidence about mechanisms, or are themselves coordination mechanisms. They are not causal substitutes. — [Mechanism Realism](https://kunnas.com/articles/mechanism-realism.md) · [Only Selection](https://kunnas.com/articles/only-selection.md)
2. **The reference telos is sustained flourishing.** The broadest achievable adaptive safety margin over deep time — not the continuity of any incumbent state, coalition, institution, or doctrine. A mechanism's own stated goal can still serve as a local proof obligation — showing that its incentives defeat even the purpose it claims is a bounded finding — but meeting that goal establishes nothing about the margin. — [Flourishing Is Maximum Safety Margin](https://kunnas.com/articles/flourishing-is-maximum-safety-margin.md)
3. **Law, rights, legitimacy, democracy, markets, and sovereignty are mechanisms under evaluation.** They are constraints, carriers, or proxies inside the analysis. None is a terminal value or a boundary of what is real. Treating one as terminal ends the mechanism search before it starts. Evaluation carries current function, replacement cost, path dependence, uncertainty, capture risk, reversibility, and who bears model error into the ledger. — [The Stack](https://kunnas.com/articles/the-stack.md) · [Mechanism Space](https://kunnas.com/articles/mechanism-space.md)
4. **Optimization is a system function.** A civilization has to build, exercise, and revise metamechanisms that search mechanism-space, discard dominated options, install, observe effects, and repair under uncertainty. Not running that loop leaves margin unrealized, and that is itself the failure. No single component — analyst, model, or institution — is presumed to contain a global optimum; the capacity is a property of the system. — [From Telos to Policy](https://kunnas.com/articles/from-telos-to-policy.md) · [The Three-Layer Architecture](https://kunnas.com/articles/three-layer-architecture.md)
5. **Uncertainty is preserved, not spent.** Partial orders, binding constraints, unknowns, and residuals stay explicit. An unmeasured effect is not a favorable default. — [The Compression Paradox](https://kunnas.com/articles/compression-paradox.md) · [Cargo Cult Epistemology](https://kunnas.com/articles/cargo-cult-epistemology.md)

*Each essay bears its own evidence. Links carry definitions, derivations, applications, and disputes; they do not transfer proof. Criticism is answered on its substance.*

Canonical: <https://kunnas.com/articles/how-to-read-this.md>

---

# Libertarianism Is an Incomplete Solution

*The truncated gradient: correct diagnosis, wrong prescription*

Elias Kunnas

## Thesis {#thesis}

Libertarianism correctly diagnoses state dysfunction as structural, then treats the option set as more-state versus less-state. Feedback-architecture quality is a third variable: keep the coordination engine, install a governor. Telocracy is that option.

## Standard objections addressed in this essay

- “Constitutional political economy already studies better state architecture.” — [§III](#iii-why-option-3-was-missing) (The claimed residual is continuous mechanism assurance, not first discovery of rules.)
- “The knowledge problem applies to telocracy too.” — [§III](#iii-why-option-3-was-missing), [§V](#v-what-option-3-looks-like)(Distributed information, plural models, and local action remain design constraints.)
- “State size and feedback quality are not independent axes.” — [§II](#ii-the-option-set) (They interact; the third-axis diagram is a search aid, not geometry.)
- “You have not shown the engineered state beats a minimal state.” — [§IV](#iv-the-correct-gradient-on-the-wrong-landscape) (“Global optimum” is a hypothesis requiring comparative operational evidence.)
- “Automatic rules can be gamed and become obsolete.” — [§V](#v-what-option-3-looks-like) (They need calibration, exceptions, correction, succession, and retirement.)
- “An explicit telos invites technocracy or tyranny.” — [§V](#v-what-option-3-looks-like) (The commitment, standing constraints, and residual closure must remain contestable.)
- “Markets also solve coordination failures through evolving institutions.” — [§VI](#vi-the-engine-problem) (The architecture should preserve and compare polycentric mechanisms.)

---

## I. The Diagnosis Is Real {#i-the-diagnosis-is-real}

The libertarian diagnosis of state dysfunction is substantially correct.

Hayek identified the **knowledge problem**: central planners cannot aggregate the dispersed information that markets process automatically. No committee can replicate what prices compute for free.

Public choice theory identified the **incentive problem**: politicians optimize for re-election, bureaucrats optimize for budget expansion, interest groups optimize for rent extraction. The state doesn't serve "the public interest" because [no mechanism ensures that it does](governance-alignment-problem.md).

Regulatory capture theory identified the **predator problem**: the entities that regulation is supposed to constrain end up writing the regulation. The referee joins the team with the most money.

These aren't ideological claims. They're *derived from the mechanism structure*. The libertarian analytical tradition—from Mises through Hayek through Buchanan through Friedman—didn't merely observe that states malfunction. They derived, from the architecture of available governance mechanisms, **why** states necessarily malfunction. Given self-interested agents + electoral incentives + information asymmetry, capture is deducible. Given distributed knowledge + central planning, failure is deducible. These are deductions from mechanism structure, not anecdotes.

And the deductions are correct—within the mechanism vocabulary they operate on. The diagnosis is real. What follows from it is where things go wrong.

---

## II. The Option Set {#ii-the-option-set}

When a state's feedback architecture is broken, three useful design directions exist:

1.  **More state.** Expand the broken architecture. Add more regulation, more bureaucracy, more redistribution. Hope that *this time* the incentives will align.
2.  **Less state.** Shrink the broken architecture. Remove regulation, cut bureaucracy, minimize redistribution. Let markets handle what the state handles badly.
3.  **Fix the feedback architecture.** Keep the state but rewire its control systems. Replace broken incentive structures with ones that actually align institutional behavior with outcomes.

The mainstream political debate oscillates between Options 1 and 2. The left argues for more state. The right argues for less state. Libertarians are the right's most intellectually rigorous wing, following Option 2 to its logical conclusion: minimize the state to the smallest possible footprint.

Option 3 is represented in part by constitutional political economy, public choice, cybernetics, and polycentric governance. The proposed residual is continuous mechanism assurance: tracing implementation, monitoring outcomes, and correcting failure across domains.

This is a search aid, not a literal geometry: state size, scope, decentralization, and feedback quality interact. When feedback architecture is excluded from view, reducing a dysfunctional state can be a locally attractive response. [Telocracy](telocracy.md) names a candidate architecture for testing the omitted dimension.

---

## III. Why Option 3 Was Missing {#iii-why-option-3-was-missing}

The libertarian tradition operates on a set of governance primitives: {central planning, markets, elections, bureaucracy, regulation}. Constitutional political economy, public choice, cybernetics, and polycentric governance already extend that set. Their remaining question is continuous assurance: whether institutions detect, challenge, and correct failures across their lifecycle.

It's like proving that no polynomial solves the general quintic (Abel-Ruffini theorem). The proof is airtight—*within polynomials*. Then Galois theory arrives and the problem transforms entirely. The theorem remains true. The problem space expanded around it.

Option 3 requires conceptual primitives that weren't in the governance vocabulary until recently:

**Mechanism design theory** (Hurwicz, Myerson, Maskin — Nobel 2007) formalized how to engineer incentive structures that produce desired outcomes. Before mechanism design, "fix the incentives" was a wish, not a discipline.

**Computational governance** — the idea that governance is an information-processing problem solvable by better algorithms, not just better people — only became thinkable with the computational paradigm. You can't propose "the state as [search function](telocracy.md#v-the-state-as-search-function)" before the concept of a search function exists.

**Selection pressure analysis** applied to institutions — understanding that [selection pressure, not moral failure](governance-alignment-problem.md#iii-selection-pressure-not-moral-failure), produces institutional dysfunction — requires evolutionary thinking applied outside biology. This cross-pollination is recent.

**AI alignment** formalized the problem of aligning a powerful optimization process with goals it wasn't trained to pursue. The governance alignment problem is structurally identical, but the parallel only became visible once AI alignment named the pattern.

Hayek had half the toolkit. He understood distributed information processing (markets). He understood the knowledge problem (central planning fails). But he stopped at markets. He didn't ask: *what if you could build institutional architectures with the information-processing advantages of markets AND the coordination capacity of states?*

He didn't ask because the question was premature. The engineering tools to answer it didn't exist yet.

---

## IV. The Correct Gradient on the Wrong Landscape {#iv-the-correct-gradient-on-the-wrong-landscape}

In optimization, **gradient descent** means: look at where you are, determine which direction improves the objective function, take a step in that direction. Repeat.

Gradient descent on a truncated landscape produces predictable failure. If the landscape only has two dimensions — more state, less state — then the gradient correctly points toward less state whenever the current position is "too much broken state." The algorithm is working. The landscape is wrong.

This is what happened to the libertarian intellectual tradition. The gradient is locally correct. Given {more state, less state}, less state produces better outcomes at the margin when the state is already dysfunctional. Fewer regulations means less capture. Less redistribution means less moral hazard. Smaller bureaucracies means less Parkinsonian expansion.

But the gradient does not establish a global optimum. A minimum viable state can outperform a dysfunctional larger state; an engineered state is a competing hypothesis that requires comparative operational evidence.

**The core claim:** Feedback architecture is a candidate design dimension alongside state size and scope. The third-axis diagram is a search aid; comparative evidence determines whether an engineered state improves on a minimal one.

The truncation produces a second failure mode at the rhetorical surface. A deductive system whose axioms have been certified internally (NAP, self-ownership, property as natural right) can derive policy conclusions of high logical validity, and the validity feels like execution. The argument inside the system clears every gate; the audit channel between the system's outputs and the institutional substrate they require — courts, registries, enforcement, political coalitions, the contingent institutional achievements that actually carry property and contract into the world — sits outside the system and is rarely tested. [*Nominal Execution*](nominal-execution.md) names this surface: the formal availability of the deductive answer is credited as the institutional fact it would have to produce.

The libertarian who derives "property rights are foundational" from the NAP is performing legitimate within-system reasoning; the move that converts the derivation into a political prescription treats the derivation's validity as if it had executed the enforcement substrate the prescription depends on. The truncated-landscape diagnosis identifies the missing dimension; the nominal-execution diagnosis identifies why the deductive move at the rhetorical surface produces consistent confidence without consistent contact with the substrate.

---

## V. What Option 3 Looks Like {#v-what-option-3-looks-like}

[Telocracy](telocracy.md) is governance with an explicit, measurable objective function — and institutional architecture to pursue it.

Its design constraints include distributed information, plural models, local action, constituent standing, and residual political choice. Automatic rules require calibration, exceptions, adversarial challenge, correction, succession, and retirement; without them they invite gaming and lock-in.

The key features that distinguish it from both "more state" and "less state":

**Explicit telos.** The state has a declared persistence commitment, with constituent standing and residual political closure. Physics supplies conditional viability constraints after the commitment and system boundary are specified.

**Mechanism modeling and measurement.** A [Fourth Branch](fourth-branch.md) that models incentive structures, stress-tests mechanisms, and continuously measures whether institutions produce their stated outcomes. Not process compliance — actual results. The mechanism that current governance entirely lacks: feedback from reality to policy.

**Automatic constraints.** Switzerland's debt brake is a prototype: a constitutional rule that fires automatically when fiscal parameters exceed thresholds. Politicians decide *what* to cut, not *whether* to cut. Extend this to demographics, institutional health, infrastructure. Hard rules that ground the democratic ratchet — the tendency of electoral systems to expand obligations faster than capacity.

**Sunset clauses.** Every institution expires unless it demonstrates continued function. The default is death. Survival requires proof. This inverts the current architecture where institutions are immortal regardless of performance — which is the failure mode that libertarians correctly identify.

**[Full accounting](full-accounting.md).** Every policy decision accompanied by its full cost across all capital types and time horizons. No [complexity laundering](complexity-laundering.md). No temporal displacement. The information asymmetry that enables dysfunction — the information asymmetry that Hayek identified — is attacked directly.

Note what this is NOT:

- **Not bigger state.** A state with proper feedback loops can be smaller than a dysfunctional one, because dysfunctional bureaucracies expand to compensate for their own failures.
- **Not technocracy.** Experts don't decide. The telos decides. Experts measure. The Fourth Branch audits results. Where multiple paths produce equivalent long-term viability, people choose — not because democracy is sacred, but because choice on the Pareto frontier is free and legitimacy-generating.
- **Not central planning.** The reference objective is publicly adopted. The means are distributed. Markets, local governance, individual initiative — all preserved. What's added is the measurement layer that tells you whether the distributed system is producing what it should.
- **Not democracy with a compass.** Telocracy is agnostic about decision mechanisms. It specifies *what to optimize for*, not *who decides how*. Physics constrains the option space. Within the viable region, the question of "who chooses" is an implementation detail — important, but secondary to "what's the option space."

---

## VI. The Engine Problem {#vi-the-engine-problem}

A car's engine is misfiring. The mechanic says: "The control system is sending wrong signals to the fuel injectors." Two responses:

**Response A:** "Remove the engine. We'll push the car."

**Response B:** "Replace the control system."

Response A is libertarianism. It correctly identifies that the engine is malfunctioning. It correctly identifies that the malfunction comes from the control system. It incorrectly concludes that the solution is to remove the engine rather than fix the control system.

The problem with Response A is that it **destroys the substrate that the actual fix requires**.

A state — even a dysfunctional one — provides coordination infrastructure that no market can replicate. Markets require the state to define and enforce property rights — libertarians know this (it's the "night-watchman state" concession). But the same logic extends: pollution, commons depletion, and intergenerational cost-shifting are coordination failures that fall outside any individual transaction — Hayek acknowledged this but didn't develop it. Infrastructure, demographic sustainability, and civilizational defense require coordination across timeframes longer than any market actor's discount horizon. And you need institutional infrastructure to *build* the Fourth Branch — dismantling the state before building its replacement is demolishing the house before the new one is ready.

That diagnosis does not displace markets or polycentric governance. A feedback architecture preserves them as competing mechanisms and compares their performance against centralized alternatives.

The libertarian who removes the state to fix governance is like the surgeon who removes the patient to fix the disease. The disease is real. The treatment kills the host.

These aren't just functions the state performs. They're **capital stocks** that require active maintenance: demographic capital (population sustainability), social capital (trust, cooperation norms), institutional capital (functioning legal and administrative structures), cognitive capital (accumulated knowledge, education systems), infrastructure capital (physical networks), and environmental commons. Each depreciates. Each requires investment. Each is invisible to markets because none carries a price signal.

A civilization that maintains only its financial and physical capital while letting demographic, social, and institutional capital erode is drawing down the balance sheet. The numbers look fine until they don't — the same way a company looks profitable right up until the deferred maintenance bankrupts it. **[Full accounting](full-accounting.md)** — across all capital types, across all time horizons — is the minimum condition for knowing whether you're building or consuming.

The libertarian tradition could afford this blind spot because these stocks were abundant when the tradition developed. Hayek could take social trust for granted. Friedman could assume functional institutions. The well was full. It isn't anymore. The libertarian who budgets for defense alone is maintaining one line on a balance sheet with a dozen entries. The other lines are depreciating unmonitored. Not because markets failed — markets were never designed to maintain capital stocks without price signals. This is the class of problem that requires institutional architecture, and it's the class of problem that "less state" cannot address.

---

## VII. The Convergence {#vii-the-convergence}

The surprising conclusion: libertarianism and telocracy share more diagnostic ground than either shares with the conventional left-right spectrum.

| Insight | Libertarian Version | Telocratic Version |
|----|----|----|
| State dysfunction is structural, not personal | Yes (public choice theory) | Yes ([alignment problem](governance-alignment-problem.md)) |
| Incentives determine outcomes, not intentions | Yes (Hayek, Friedman) | Yes ([implementation over disposition alone](ethics-is-an-engineering-problem.md)) |
| Information problems are central | Yes (knowledge problem) | Yes (feedback architecture) |
| Current governance is structurally captured | Yes (regulatory capture) | Yes ([alignment problem](governance-alignment-problem.md)) |
| Democratic ratchet is real | Yes (Buchanan) | Yes (obligations \> capacity) |
| Reform from within is almost impossible | Yes (government failure) | Yes (misaligned agents control reform) |

The divergence is narrow but critical:

| Question | Libertarian Answer | Telocratic Answer |
|----|----|----|
| Can the state's architecture be fixed? | No → minimize it | Yes → rewire it |
| Can markets handle coordination? | Mostly yes | Within horizons; not across them |
| Is there a third option beyond more/less state? | Not considered | Yes: better feedback architecture |

A libertarian who encounters telocracy faces a genuine update. The diagnosis they trust — public choice, knowledge problem, regulatory capture, democratic ratchet — all survive intact. What changes is the *prescription*: not "remove the engine" but "install a better control system."

In fact, telocracy delivers libertarian goals more completely than libertarianism does — by libertarian criteria:

| Libertarian Goal | Libertarianism Delivers | Telocracy Delivers |
|----|----|----|
| Minimize discretionary state power | Remove state functions | Replace discretion with automatic constitutional constraints — a debt brake removes more political discretion than defunding a ministry |
| Maximize feedback quality | Market prices | Market prices + Fourth Branch measurement of capital stocks that markets can't price |
| Institutional accountability | Competition (Tiebout sorting) | Competition + sunset clauses + mandatory outcome measurement — institutions that can't demonstrate function die automatically |
| Prevent capture | Minimize what can be captured | Make capture structurally unproductive — automatic constraints fire regardless of who controls the institution |

An automatic constitutional constraint is less discretionary state power than a politician making the same decision. A sunset clause that kills unproductive institutions is more libertarian than a deregulation bill that has to survive lobbying. The Fourth Branch that makes all costs public is more Hayekian than abolishing the department that hides them — because it fixes the information asymmetry rather than removing one source of it.

The obvious counter-example: the EU's fiscal rules were automatic on paper and suspended in practice. The difference is architectural. The EU's Stability Pact was **default-allow** — spending continued unless someone actively punished the violator, and the violators voted on their own punishment. The Swiss debt brake is **default-deny** — spending stops automatically unless revenue covers it. Politicians don't decide *whether* to cut; they decide *what* to cut. When inaction produces discipline rather than drift, the constraint doesn't need political will to hold.

Even the Swiss prototype has a blind spot: the debt brake constrains fiscal capital but doesn't distinguish consumption from investment. Under pressure, politicians cut infrastructure (deferred maintenance) rather than entitlements — depleting one capital stock to preserve another. This is why a single automatic constraint isn't enough. [Full accounting](full-accounting.md) across all capital types is the generalization: the same default-deny architecture, extended to every line on the civilizational balance sheet.

The update is: **the option set was incomplete.** "Less state" was the best available answer. It is no longer the best available answer. A better one exists — one that preserves every libertarian diagnostic insight while avoiding the substrate-destruction problem.

---

## VIII. Why This Matters Now {#viii-why-this-matters-now}

Libertarianism has been the most intellectually serious critique of state dysfunction for 80 years. Its prescriptions have been partially adopted — deregulation, privatization, monetary discipline — but as a complete program, no polity has achieved the minimal state. The reason is the engine problem: people intuitively understand that removing the state removes coordination capacity. "Who will build the roads?" is a meme, but it encodes a real objection. The libertarian response — "the market will" — is unconvincing for functions with externalities, long horizons, or coordination requirements beyond bilateral exchange.

Telocracy resolves this impasse. It takes the libertarian diagnosis seriously — more seriously than the mainstream left or right, which deny or minimize state dysfunction. And it provides a prescription that doesn't require people to accept the implausible claim that markets can handle everything.

The obvious libertarian counter: *who builds it, and what prevents the builders from being captured?* These are engineering problems with engineering answers — constitutional independence, international oversight, capital funding instead of annual budgets, mandatory transparency, automatic triggers that bypass political discretion. The [Fourth Branch](fourth-branch.md) details the structural requirements. A [concrete implementation proposal for Finland](https://mekanismirealismi.fi/mechanism-authority) demonstrates that capture-resistance can be designed into institutional architecture, not assumed.

The libertarian who discovers telocracy doesn't abandon their analysis. They *complete* it. The truncated gradient was pointing in the right direction — away from dysfunctional state architecture. It just couldn't see the peak that requires going *through* the state to reach: a state with feedback architecture that actually works.

**Diagnosis right, prescription wrong:** Libertarianism is the most important intellectual tradition that almost nobody follows, because it diagnosed the disease correctly but prescribed amputation when surgery was needed. The surgery — telocracy — requires tools that only recently became available. The libertarian who updates on the expanded option set becomes the most valuable ally the actual fix has.

## Synthesis {#synthesis}

**The argument in four sentences:** Libertarian public choice theory and the Hayekian knowledge problem are valid diagnoses, but the prescription to minimize the state destroys coordination infrastructure that markets cannot replace. Telocracy expands the option set by introducing a third dimension — feedback architecture quality — using mechanism design and computational governance. Constitutional constraints and an independent Fourth Branch minimize discretionary power more effectively than abolition. This preserves civilizational coordination capacity while rewiring the state’s control systems around an explicit public reference objective and audit loop.

---

*Governance series: [Diagnosis](governance-alignment-problem.md) → [Telocracy](telocracy.md) → [Institution](fourth-branch.md) → [Full Accounting](full-accounting.md) → **Libertarianism Is an Incomplete Solution***

**Related:**

- [Telocracy](telocracy.md) — The missing third option: governance with an explicit public reference objective
- [The Governance Alignment Problem](governance-alignment-problem.md) — Why the current architecture is structurally misaligned
- [The Fourth Branch](fourth-branch.md) — The institutional implementation of the feedback architecture
- [Ethics Is an Engineering Problem](ethics-is-an-engineering-problem.md) — Moral justification vs implementation reliability
- [Laws Are the Wrong Abstraction](laws-are-the-wrong-abstraction.md) — Why treating enactment as implementation fails

## Sources and Notes

**The libertarian analytical tradition ([§I](#i-the-diagnosis-is-real), [§III](#iii-why-option-3-was-missing)):** Ludwig von Mises and Friedrich Hayek on the knowledge problem — distributed information that no central planner can aggregate. James Buchanan and public choice theory on the incentive problem — politicians, bureaucrats, and interest groups optimizing for re-election, budget expansion, and rent extraction rather than the public interest. Milton Friedman on monetary discipline and the case for a minimal state.

**Mechanism design theory ([§III](#iii-why-option-3-was-missing)):** Leonid Hurwicz, Eric Maskin, and Roger Myerson, awarded the 2007 Nobel Memorial Prize in Economic Sciences for formalizing how to engineer incentive structures that produce desired outcomes.

**Automatic constitutional constraints ([§V](#v-what-option-3-looks-like), [§VII](#vii-the-convergence)):** Switzerland's constitutional debt brake as the prototype for default-deny fiscal discipline. The EU's Stability and Growth Pact as the contrasting default-allow failure case — automatic on paper, suspended in practice because violators voted on their own punishment.