Democracy and its Blueprints

The Problem That Does Not Go Away

In 1911, Robert Michels recorded what no one has since refuted: every organisation that grows large enough to change something develops an internal hierarchy that undermines the original goal of change. He called it the Iron Law of Oligarchy. Iron — because it cannot be suspended by good intentions. It is not a moral failure. It is system dynamics.

The German Greens are the most recent domestic illustration. They tried everything an organisation can try to avoid becoming like every other party: dual leadership, rotation principle, imperative mandate, grassroots democracy. Twenty years later they are a party like all others — with career politicians, backroom compromises and the built-in logic of re-election. That is not an accusation. It is a finding.

The finding forces an uncomfortable question: if even determined will to reform fails against the system — what can actually be changed? And how?

Over recent months three concepts have emerged that answer the same question — from three different directions, across three different time horizons, with three different levers. They do not contradict one another. But they do not replace one another either. Whoever pursues only one of them mistakes the depth of the problem.

First Answer: The Mandate

The simplest lever — and the only one that can be applied immediately without constitutional amendment — is the mandate itself. The Basic Law permits candidacies independent of party. There is no legal obstacle to sending citizens into parliaments who belong to no party, are subject to no parliamentary group discipline, and need not calculate their next re-election into a party career.

That sounds naive. It is not — it is minimal. It does not change the system. It changes the people within the system: one by one, mandate by mandate, from the municipal level upward. A voters' association takes on the logistical tasks a party fulfils — candidate selection, election campaigning, accountability — without the power structure a party inevitably builds.

What this concept achieves: it creates a different type of elected representative. Whoever does not need to be re-elected to sustain a party career, whoever does not depend on parliamentary group discipline, whoever does not derive their livelihood permanently from politics — that person decides differently. Not better in the moral sense. But under different incentives.

What it does not achieve: it changes nothing about the rules of the parliament into which this representative enters. Standing orders remain the same. Lobbying structures remain the same. Media power remains the same. The independent representative acts in an environment that continues to systematically reward adaptation. They can resist this environment longer than a party politician — but they will be shaped by it, sooner or later.

Michels' law applies here too. When enough representatives without party enter a parliament and want to work together effectively, a need for coordination arises — and from coordination comes structure — and from structure comes hierarchy. The new independence organises itself into a new party. The cycle begins again.

This is not a refutation of the concept. It is its limitation. The mandate is the right first step — because it is possible. Because it requires no constitutional change. Because it is demonstrable. But it is not enough.

Second Answer: Visibility

Corruption thrives in concealment. Not because all people want to be corrupted in the dark — but because concealment carries no consequences. This is not a moral assertion. It is an empirical observation about human behaviour under different conditions. Those who act in the light act differently.

A fifth constitutional organ — the Transparency Council — would have a single task: make power visible. Not evaluate, not sanction, not govern. Document. Every vote by every elected representative, cross-referenced with the electoral promise. Every lobbying contact, public and in real time. Every party donation, every supervisory board seat, every consultancy contract. Every enacted law with impact assessment: what was promised? What occurred? Who benefited?

The tool for this exists: AI infrastructure can do in seconds what journalists would need months to achieve — cross-reference 40,000 votes with 40,000 electoral promises, set 80,000 lobbying contacts in relation to 3,000 legislative texts. Technology is not the problem. The problem is institutional location: who controls the tool? Who is its guardian?

This is why the Transparency Council needs constitutional status — and a legitimacy that does not depend on the parliamentary majorities whose conduct it is making visible. Sortition for a portion of its members. Funding through constitutional law, not through the annual budget over which a majority can decide that prefers less transparency.

What this concept achieves: it changes the conditions under which all elected representatives act — whether party-independent or not, whether upright or not. Structural light makes no distinction between the good and the bad politician. It illuminates both.

What it does not achieve: it changes nothing about the fact that economic interests have structurally more influence than cultural or civil society ones. Transparency shows that financial corporations give 340 million euros in party donations. It does not end the power asymmetry that makes these donations worthwhile in the first place. Visibility is a precondition for democracy — but it is not democracy itself.

Third Answer: The Architecture

Johannes Heinrichs formulated in his reflective systems theory a diagnosis sharper than most: today's democratic system is not neutral toward interests. It is structurally built in favour of economic power. Parties that represent all subject areas simultaneously do so according to the weight of their financing. Those who pay dominate — and those who pay most have most to lose.

His answer: four independent parliaments, each responsible for a domain of values — economy, politics, culture, foundational values — with a priority ordering from top to bottom. Foundational values set the framework for culture, culture for politics, politics for economy. Subject-specific parties instead of all-purpose parties. Economic interests get their arena — but they may no longer colonise the others.

What this concept achieves: it thinks the problem through most radically. Not improve the mandate, not increase visibility — rebuild the architecture itself. If economic power is only effective in the economic parliament and the other three chambers form its counterweight, the fundamental dynamic of the system changes. This is the most far-reaching of the three approaches.

What it does not achieve — and what the criticism rightly notes: modern politics is not a four-chamber heart with cleanly separated circulatory systems. Migration, climate policy, digitalisation are not subject questions that can be assigned to one of four parliaments. They are by definition cross-cutting questions. A four-chamber system that provides no institutional arenas for these questions does not solve the problem — it shifts it into competency conflicts between chambers.

There is also this: subject-specific parties in specialised parliaments attract specialised actors who are organised in that field. In an economic parliament, those would be industry associations and corporations — possibly more effective than today, not less. The model could institutionalise the dominance of economic interests rather than break it.

And finally: a four-chamber system cannot be introduced from nothing politically. It requires a constitutional majority, social acceptance and a transformation strategy — of which Heinrichs says almost nothing. This is not merely a tactical weakness. It is a structural one: whoever jumps directly from diagnosis to architecture without describing the path between them is describing a utopia.

The Relationship Between the Three Concepts

Whoever places the three approaches side by side sees not competition — but time horizons.

The representative without party is what is achievable today. It needs no legislative change, no constitutional reform, no social majority for a new political architecture. It needs candidates willing to try — and citizens willing to vote for them. It does not change the system. It changes what is possible within the system when people with different incentives act within it. That is little. It is the only realisable first step.

The Transparency Council is what is achievable tomorrow — if the first step has proved that it works. The foundation accompanying representatives without party begins with transparency documentation for the representatives it knows. When citizens can see the difference between documented and undocumented mandates, pressure builds: why does this not apply to everyone? From the model comes a political demand. From the demand, in a favourable moment, an institution. The history of the Federal Constitutional Court shows that such moments come — in crises, when trust in existing structures is exhausted and new institutions become possible.

Value-tiered democracy is what is conceivable the day after tomorrow — as a reform target, not a daily programme. Heinrichs' architecture is too complex to be introduced in one move. But individual elements are extractable: the idea of subject competence, the hierarchy of values as a framework for economic decisions, the institutional separation of domains currently bundled together in the same all-purpose mandate. These elements can flow into a constitutional reform — when the political climate is ripe for it, and when the preceding steps have built trust.

This is not a ladder of history on which one automatically ascends from step one to step three. It is a sequence of possibilities that build on one another — but each of which can fail. The representative without party can become a party politician. The Transparency Council can be hollowed out by a majority that fears transparency. The structural reform can drown in competency conflicts.

What connects all three concepts nonetheless is a shared basic principle that does not refute Michels' law but answers it: do not change the people — change the structural conditions under which people act. The incentives, not the intentions. Shorten the half-life of corruption, not abolish corruption.

What Remains

There is no solution to the problem Michels described in 1911. There are only better and worse environments in which the Iron Law operates more slowly or more quickly.

Lorenzo il Magnifico was not a system. He was a person who happened to have the right disposition at the right time — and gave Florence for a generation what no system can guarantee: a government that placed the common good above its own power. His death in 1492 ended that. Savonarola came. The bonfire of the vanities burned. Machiavelli mourned and wrote down what he had seen.

What Machiavelli wrote down was not an instruction manual for corruption. It was a diagnosis: this is how power works. Not how it should work — how it works. Whoever does not understand that cannot build a better world. Whoever understands only that builds no better world.

The three concepts of this essay take Machiavelli's diagnosis seriously. They do not place their hopes on the good prince. They place them on structures that make the bad prince more costly — not impossible, but more laborious. That is modest. It is the most modest thing that remains honest after the failure of all utopias.

The mandate changes who acts.
Visibility changes the conditions under which action takes place.
The architecture changes which interests carry structural weight.
All three together are not a system that prevents corruption.
They are a system that makes corruption cost something.

That may be enough. It is certainly not little.