---
title: "Mechanism Authority 2.0"
subtitle: "A stronger-capabilities design proposal"
author: "Elias Kunnas"
description: "Mechanism Authority 2.0 is a proposed, separately authorized corrective-powers module. Analytical upgrades (success specifications, stock review) can live in 1.0. Expiry, supermajority reenactment, and emergency suspension need their own authorization. Drafting, enacting, and activating are three events."
canonical: "https://mekanismirealismi.fi/mechanism-authority-2"
url: "https://mekanismirealismi.fi/mechanism-authority-2.md"
date_published: "2026-09-14"
date_modified: "2026-09-14"
llms_txt: "https://mekanismirealismi.fi/llms.txt"
---
# Mechanism Authority 2.0

A stronger-capabilities design proposal

*Elias Kunnas*

**Stronger-capabilities design proposal · Mechanism Authority 2.0**

**Prerequisite.** Assumes [Mechanism Authority 1.0](mechanism-authority.md). 2.0 does not replace 1.0. It is a proposed, separately authorized module of stronger corrective powers. Sections 1–7 remain in force as written. Frame: [Telocracy](https://kunnas.com/articles/telocracy) · [Trillions](https://kunnas.com/articles/trillions-are-managed) · Finnish twin: [Mekanismivirasto 2.0](mev-2.md).

**Relation to 1.0.** Mechanism Authority 1.0 is an ordinary-law specification: a persistent analytical agent and a stronger route into decisions. It is not a thermometer without a thermostat. Mechanism Authority 2.0 is a separate constitutional extension, not a hidden part of 1.0.

**Three events.** Drafting 2.0, enacting 2.0, and activating those powers are different things. A published protocol is not an installed protocol. Supporting 1.0 does not authorize 2.0.

**Version boundary.** Success specifications, prediction records, and systematic stock review belong in the 1.0 development programme. Sunset, two-thirds reenactment, and emergency suspension are additional legal effects that need their own authorization. This draft is not yet a sufficiently sound specification of those effects.

## Why 1.0 is not the whole system

Mechanism Authority 1.0 creates persistent analytical agency and a stronger route into decisions: a negative opinion returns a bill, requires a public reply, and keeps an accepted correction on a ledger. It does not guarantee that decisions produce adequate, resourced, coordinated action. What about statutes whose default is to continue? What about cases where exposing a defect does not change the default?

Mechanism Authority 1.0 §7 states: *Mechanism Authority cannot repeal laws or enact new ones.* That bound makes 1.0 installable. 1.0 is still not a mere meter. It couples analysis to existing decision-owners. 2.0 is a proposed extension of corrective powers — not proof that 1.0 is useless.

A stronger brake is not automatically a better steering system. If the default must be reversed or an institution given an effective intervention power, that must be justified as a capability, not as a version number.

## Adopted design requirements

Before additional power is selected, it must meet these:

- the trigger is independently reviewable and challengeable; it must not rest on Mechanism Authority's own classification alone;
- missing a target does not by itself select expiry or suspension;
- an intervention names its object, the comparison with other available actions, continuity of essential activity, a financing path for repair, a maximum duration, and an exit;
- parliamentary inaction is not an indefinite further authorization;
- the intervention's own harms and its effect on the activation indicators are tracked.

These are the selected requirements. Sections 8–12 below are an earlier candidate. They do not yet meet them. The analytical layer (success specification, stock review) can be built in 1.0 without expiry power.

## Earlier candidate: mechanism emergency

Crisis powers (§§10–12) are not permanently on. The present draft activates them when pre-defined thresholds are crossed. §§8–9 are always in force as analytical infrastructure; their legal effects (sunset, expiry queue) are the part that needs separate authorization.

**Why a formula at all?** Hungary's Ombudsman for Future Generations (2008–2012) used discretionary power and made over 200 interventions a year. A new government abolished the office. Discretion in a person's hands is a political target. Sweden's pension brake works the other way: when the balance ratio falls below 1.0, indexation is reduced by formula. Politicians said: *It is mathematics.* Nobody lost office, because nobody made a decision. The system did.

### 10 § Mechanism emergency

Mechanism Authority declares a **mechanism emergency** when at least two of the following thresholds are exceeded at once:

1.  **Debt threshold:** public debt to GDP exceeds 90% and the structural deficit is above 3% for at least 36 months.
2.  **Dependency threshold:** the economic dependency ratio exceeds the level at which existing pension and service obligations are mathematically impossible at the current tax rate. Current tax rate is a conditional policy conclusion, not an unqualified physical threshold.
3.  **Institutional threshold:** Parliament has overridden three successive Mechanism Authority 1.0 red alerts (§5) on a system-critical mechanism failure without corrective action.

Thresholds and methods are public. Mechanism Authority defines them in advance as part of 1.0's ordinary work (§5). Changing a threshold requires a public justification and approval by the international supervisory council.

**Who verifies?** Two conditions are not two independent checks if Mechanism Authority both specifies the warning and judges the response inadequate, and that judgment helps activate Mechanism Authority's stronger powers. Calling activation “mathematical” does not settle the choice of model, horizon, threshold, error tolerance, or intervention. The trigger must be independently reviewable, challengeable, and the decision to grant additional powers separately authorized. The present draft does not yet meet this.

Two simultaneous thresholds are a partial safeguard against a single shock. They are not an independent check on Mechanism Authority's own classification.

## Earlier candidate: sections 8–12

The package below is an earlier integrated candidate, not a selected replacement rule. In that candidate, §§8–9 are always in force; §§10–12 activate only in a mechanism emergency.

### 8 § Success specification and sunset

A **success specification** is analytical infrastructure: Mechanism Authority defines measurable criteria for each significant new statute. The specification can be written in the 1.0 development programme. It does not conceptually require expiry power.

The **sunset rule** is a separate legal effect. The earlier candidate moves a statute that misses its specification for three consecutive review periods into an expiry queue: the statute lapses after 24 months unless Parliament reenacts it by a two-thirds majority and answers Mechanism Authority's findings. That does not meet the requirement that a missed target does not by itself select expiry. A statute can be valuable but insufficient, poorly implemented, operating in a changed environment, serving conflicting objectives, or genuinely counterproductive. The comparison is the expected consequences of continuation, repair, replacement, and withdrawal, including transition costs.

Reenactment is not a formality. A two-thirds threshold and a reply duty block restoring a law in an omnibus vote without analysis.

**Review period:** three years as a default. Mechanism Authority may set a longer period (5–7 years) for statutes whose effects appear slowly (education reforms). The shortest allowed period is two years.

### 9 § Full accounting of the stock

Mechanism Authority conducts a sector-by-sector **full accounting** of the entire statute book. Each year one administrative branch is audited in full. The review covers all statutes, not only new ones.

For each statute:

- Does it still produce the intended effect?
- Is the incentive structure intact, or does it create reverse incentives?
- Does it interact with other statutes in a way that cancels its effect?
- Does it name a responsible party, a timetable, and consequences?

Stock review belongs to analytical infrastructure and can be built into 1.0. Entry into the expiry queue is a separate legal effect. The label “non-functioning” must not be an automatic consequence of missing a specification.

**Integration with the mechanism-case register:** A sector review must treat the 1.0 §4a register for that branch as a mandatory inspection item. Every open opening that concerns the sector is handled, including ownerless findings and improvement opportunities that a statute-by-statute review would otherwise miss.

**Pace:** designed to take 12–15 years for the whole stock. This is not an emergency measure. It is a standing process, like statutory audit of accounts.

### 11 § Emergency suspension

During a mechanism emergency (§10), Mechanism Authority has power to:

1.  **Suspend** the application of statutes that are shown to accelerate the crisis. Each suspension is temporary (at most six months at a time) and requires a public justification.
2.  **Freeze** the budget of a suspended statute. Funds do not return to the ministry; they move to a reserve until Parliament decides their use.

**Mechanism Authority cannot enact replacement law.** The present draft treats suspension as creating a vacuum that Parliament must fill, and continues the suspension automatically if Parliament does not act within 90 days. That confuses a temporary power with a persistent effect. The draft needs an aggregate duration cap, exit conditions, continuity arrangements, a financing path for repair, and what happens when both reinstatement and continued suspension carry serious risks. Frozen funds may be the very resource replacement delivery needs. People and other institutions may depend on the rule.

**Design requirement for any stronger intervention.** A decision to suspend, sunset, or freeze names the object of intervention, the comparison with other available actions, continuity of essential activity, the resource route needed to repair the defect, a maximum duration, and independent review and exit conditions. The intervention's own harms and its effect on the activation indicators are also tracked. Parliamentary inaction does not by itself create an indefinite further authorization. This is a requirement on the next draft, not a finished additional power.

The constitutional bound is clear: Mechanism Authority cannot receive legislative power (Constitution Section 3: state power belongs to the people, represented by Parliament). Enacting replacements is always Parliament's task. Mechanism Authority's power, if authorized, is to suspend and freeze — a brake, not a steering wheel.

### 12 § End of emergency

The mechanism emergency ends when the §10 thresholds are no longer exceeded. §§10–12 return to rest. The present draft leaves unaddressed suspensions in place until Parliament acts: consequences can therefore outlast the condition that authorized the emergency power. That is an open design point, not a finished temporariness.

## Rationale

### 1. Reversing the default

The present default: a statute remains in force forever unless someone actively repeals it. Repeal requires a bill, committee work, a vote. In practice, bad laws are not repealed, because repeal is politically more expensive than inaction.

The 2.0 candidate reverses that default: a statute that does not meet its aims lapses unless someone actively continues it — at a two-thirds majority. That moves the political cost onto the maintainer of a failed law, not onto its critic. It remains a candidate until the defect-to-remedy map is specified. Missing a target is not by itself a reason to create a vacuum.

The principle is not new. Finland's life-expectancy coefficient for pensions works the same way: the default is that pensions adjust to longevity. Political intervention is needed only to deviate from the formula.

### 2. Why Mechanism Authority writes the success specification

If the preparing ministry wrote its own success criteria, it would optimize for metrics that fill automatically — process metrics (meetings held, reports published) rather than the goal. That is Goodhart applied to self-supervision.

Mechanism Authority's independence (1.0 §6) makes possible a specification that measures effect, not activity. *Did outcomes improve?* — not *Was something done?*

### 3. Stock review is bookkeeping

Every firm audits its balance sheet annually. Public financial audit is statutory. Finland's statute book — thousands of laws that together are the state's operating logic — has never been audited as a whole. Nobody knows how many statutes are in force, how many still serve their aim, how many cancel another.

Full accounting of the stock is not a radical proposal. It is basic bookkeeping applied to the state's operating logic. The radical fact is the present: the state's most complex system without an audit. The radical step is attaching expiry to a missed label.

### 4. Emergency suspension is a fuse

In an electrical grid a fuse cuts current when load exceeds capacity. It does not repair the fault; it prevents a fire. Repair is the electrician's job. Mechanism Authority's emergency suspension, if authorized, is a fuse: it stops a mechanism that accelerates a crisis and gives Parliament time to repair.

Without a fuse, the alternative is to let the machine run until it breaks — and then the terms of repair are set by the IMF or the Commission, not by Finland's parliament. A domestic fuse is cheaper than external receivership. It is still not a steering wheel, and it still needs continuity rules.

### 5. Hungary's lesson and Sweden's model

**Hungary (2008–2012):** Ombudsman for Future Generations Sándor Fülöp used discretionary power effectively — over 200 interventions a year. A new government abolished the office. *Personal power is a political target.*

**Sweden (pension brake):** When the balance ratio fell below 1.0 in 2010 and 2014, real pensions fell automatically. Nobody lost office, because nobody made a decision. *Formula power is not a political target in the same way.*

**Slovakia (2011):** a constitutional debt brake with escalation steps: 50% debt/GDP → written explanation, 53% → ministers' pay frozen, 55% → automatic cut, 60% → mandatory confidence vote. Steps make escalation predictable and remove personal discretion.

The present 2.0 draft follows Sweden and Slovakia on formula triggers. That does not settle who reviews the formula, or what additional powers follow. Formula power that expands the formula-writer's authority is not the same as a pension brake.

### 6. Constitutional implementation

**1.0 vs 2.0 — constitutional stepping.** Mechanism Authority 1.0's mechanisms (mandatory return via rules of procedure, public reply duty by ordinary law, follow-up triggers) are designed to seek implementation without a constitutional amendment. Mechanism Authority 2.0's powers — especially a two-thirds reenactment threshold (§8) and emergency suspension (§11) — are placed here because they need constitutional-level authorization.

Those powers conflict with the constitution as it stands, especially Section 3 (state power belongs to the people) and Section 80 (limits on delegating legislative power).

Two routes under Section 73:

1.  **Ordinary Section 73:** the first parliament adopts the proposal by a majority and leaves it over an election; the next parliament adopts by two-thirds. Not two successive two-thirds votes.
2.  **Urgent Section 73:** a five-sixths majority declares urgency, then a two-thirds majority adopts in the same parliament.

The earlier candidate is designed so that Mechanism Authority does not *take* legislative power from Parliament. A brake without continuity, comparison, and exit is not the selected additional power.

## Objections

**“This is technocracy.”** — The present system is already hidden technocracy: consultancies prepare reforms, ministry economists write forecasts, models are secret. 2.0, if authorized, would make the technical analysis visible, public, and challengeable. Open mathematics is not technocracy unless a thermostat is tyranny. A brake that writes its own expansion trigger still needs an independent check.

**“Automatic thresholds are inflexible.”** — Inflexibility is a feature of a formula brake. Flexible rules are recommendations: a politician can always find a reason not to follow them. The 2.0 safety valve in the present draft is a two-thirds exception — not inflexibility without an exit. Inflexibility does not license the wrong remedy for a missed target.

**“Thresholds can be gamed.”** — Debt/GDP can be manipulated with off-balance-sheet arrangements. Two simultaneous thresholds are not enough if Mechanism Authority both computes the trigger and gains from activation. The trigger must be independently reviewable.

**“Parliament can just vote failed laws back.”** — A two-thirds threshold and a reply duty make that expensive. A politician who votes a documented failure back after Mechanism Authority's public analysis carries their name in the archive. That does not prevent reenactment; it makes it politically costly. Same family as CPB election scoring in the Netherlands: refusing open calculation is not illegal, but it is costly.

**“Who watches Mechanism Authority when it has this much power?”** — The 1.0 protections (§6, §7) are the starting point. They are not enough if a suspension remains after the emergency ends. A temporary power needs an exit condition, not only a trigger.

**“This is impossible to enact.”** — A stronger package needs its own coalition, authorization, resources, and continuity. Crisis is one possible opening, not the only one. The parallel route is to install 1.0 when that is possible, and at the same time look for principals who can carry a larger change without first installing the weaker package.

## Summary: 1.0 vs 2.0

|  | Mechanism Authority 1.0 | Mechanism Authority 2.0 (candidate) |
|----|----|----|
| **New statutes** | Mandatory return + public reply | Same + success specification at birth |
| **Existing statutes** | Monitoring and reporting | Stock review + candidate expiry queue |
| **Failed mechanisms** | Alert → committee | Candidate: lapse in 24 months unless reenacted 2/3 |
| **Crisis** | Reporting | Candidate: emergency suspension + budget freeze |
| **Stock audit** | Development programme (analysis without expiry) | Full accounting + candidate: entry into expiry queue |
| **Default** | Statute remains in force | Candidate: statute lapses unless shown to work |
| **Activation** | Permanently in force | §§8–9 always as analysis; §§10–12 in emergency, subject to independent authorization |

## Architectural integrity

Mechanism Authority 1.0 and 2.0 form one intended system:

- **1.0 (§§1–7):** measurement, analysis, pre-legislative review, mandatory return. Installed in ordinary time.
- **2.0 analytical layer (§§8–9 content):** success specification and stock review. Can be built in 1.0 without expiry power.
- **2.0 legal layer (§§8–12 effects):** sunset, expiry queue, mechanism emergency, suspension, freeze. Need separate authorization. Must not make Mechanism Authority the judge of its own expansion.

The move from 1.0 to 2.0 is not an automatic next step. Draft, statute, and activation are different events. A prepared protocol can be useful when a window opens — crisis is one possible opening, not the theory of change. Installing 1.0 does not oblige the future institution to recommend this 2.0 draft; it must be able to investigate its own bottlenecks and propose a better successor.

The strategic commitment is to continuing improvement. It is not to an irrevocable authority ratchet.

[![Mechanism Realism – A User Manual for Civilization](assets/cover_front_600.jpg)](/index.md)

MEKANISMIREALISMI: A USER MANUAL FOR CIVILIZATION

Mechanism Authority is one part of a larger frame. The book gives the diagnosis, the protocol, and the tools. (Finnish only)

[Buy the book](osta.md) [Read more](/index.md)