Two Keys

I. Who Should Own the Firm?

In the coming years hundreds of thousands of German firms will change hands, or close because no one can be found to take them over. We described this in “The Silent Shutdown”. 28 percent of owners can imagine handing over to their workforce. That raises a question usually left to textbooks: who should actually own a firm?

We know three answers. The owner and his heirs: that is the family business. The shareholders: that is the stock market. And two answers, both invented by idealists: the workforce, which is the cooperative, and a purpose, which is the foundation-owned company. Both have shown that they can work. Both have also shown how they fail. And they fail at opposite points.

II. The Cooperative and Its Insiders

Mondragón in the Basque Country is the world’s largest cooperative group. We described it as a hope in “From Experiment to Model” and examined it critically in “The Experiment That Did Not Want to Become a Model”. At the beginning stood five young men and a priest, José María Arizmendiarrieta, who made clear to them that they would win in the long run if they overcame their selfishness. Because it worked, it carried on long after his death in 1976.

Then came what Franz Oppenheimer had described as early as 1896 as the “iron law of producer cooperatives”. When a cooperative succeeds, membership gains value. Those inside have an interest in keeping newcomers out, because every newcomer takes a share. So wage workers are hired instead. Today fewer than 40 percent of Mondragón’s employees are members. When the appliance maker Fagor went bankrupt in 2013, the members were placed in other cooperatives. The employees without membership lost their jobs.

The pattern is familiar from any golf club that becomes successful: first it looks for members, then it raises the entry fee, then it introduces a waiting list, and the work on the course is done by staff. Nobody decided this. It follows from the construction. The cooperative gives every worker a voice, but its assets are not bound. They belong to the insiders, and the insiders defend them.

III. The Foundation and the Missing Questioner

In 1889 Ernst Abbe transferred his shares in Carl Zeiss to a foundation. Robert Bosch laid down something similar for his company in his will. The principle: the firm belongs to a purpose, the shares may never be sold, nobody can pocket the assets. That disarms Oppenheimer’s law, because there is no value anyone would need to defend against newcomers.

The record is impressive. In Denmark, where foundations control companies such as Novo Nordisk and Maersk, Copenhagen Business School compared such firms over decades. 60 percent of foundation-owned companies reach the age of 50, against 10 percent of the others. Their results fluctuate less, they carry less debt and they pay better.

And yet Bosch and ZF show how this form fails. We described it in “The Long Game, Lost”: a monoculture built on the combustion engine, debt-financed acquisitions, a lost bet on hydrogen. At Bosch, members of the trust that exercises the voting rights also sit on the supervisory board. The controllers control themselves. The foundation form protects against the greedy shareholder, and therefore also against the awkward one. What is missing is someone who asks in time. Those on the shop floor probably sensed earlier than any board that diesel would not last forever. Nobody asked them.

IV. Two Keys

Put the two experiences side by side and a clear picture emerges. The cooperative has a voice but no bound assets. It fails because of its insiders. The foundation-owned company has bound assets but no voice. It fails for lack of oversight. Each form is blind exactly where the other can see.

A bank safe-deposit box opens only with two keys, one held by the customer and one by the bank. Neither can get in alone. That is how a firm would have to be built if it is to be at the mercy of neither its insiders nor its administrators. One key is held by the purpose, which protects the assets from the members. The other is held by the workforce, which protects the firm from its administrators. Each side checks the weakness of the other. This is not a matter of good intentions but of construction: a negative feedback loop that still works when the founder is long dead and nobody remembers his speeches.

V. What the Hybrid Would Look Like

The assets belong to the purpose. They cannot be sold, are not distributed on dissolution and cannot be released by amending the articles. Oppenheimer loses his lever.

The purpose is concrete. Not “promote the common good”, which can be stretched at will, but something like: “make precision parts in the region and keep the jobs that do so”. A concrete purpose can be checked.

Everyone who works there shares in the profits. Not in the assets, but in the profit, by the same rule for all. Whoever passes probation becomes a member without the others voting on it. A door that opens by itself cannot be bolted.

The workforce votes on the business, not on the purpose. It elects management and can vote it out. A separate body protects the purpose.

The guardian of the purpose does not appoint itself. It is mixed, partly elected by the workforce, partly appointed from outside, for example by a university or chamber, and everyone serves for a limited term. That way no new circle forms that controls itself.

There are rules for emergencies. Mondragón let Fagor go because more money would have endangered the other cooperatives, and caught the people with a solidarity fund. This ability to say no has to be settled beforehand, not fought over in the crisis.

VI. What the New Legal Form Can Already Do

A surprising amount of this is already on the table. On 4 March 2026 Germany’s justice and finance ministries published a framework concept for a “company with bound assets” (GmgV), which we introduced in “Europa und die Megamaschine”. The asset lock is meant to be irrevocable, even on conversion, dissolution or insolvency. It is one member, one vote. From 20 members a supervisory board is mandatory, and cooperative audit associations are to monitor the lock. So the first key already exists.

For the second key the concept has three gaps. First, joining requires the company’s consent. That is exactly the door through which Oppenheimer’s law entered Mondragón. Second, founders may receive multiple voting rights and a veto, as the German Chambers of Commerce point out critically. That may ease the transition, but it should be time-limited. Third, distributions to members are excluded, including profit-sharing. The lawyer Marvin Reiff considers this unrealistic, and for a workforce taking over a firm it is an obstacle. Those who carry responsibility should also share in success.

The German Trade Union Confederation points to a further danger: if no member shares in the profits, nobody looks closely. Assets could drain away through excessive salaries, rents or licences. It therefore calls for strong co-determination regardless of company size. That is the second key, approached from a different side.

VII. Where Every Form Reaches Its Limit

No legal form protects against a bad market. Britain’s John Lewis Partnership has been wholly owned by a trust since 1950, and all employees share in the profits as “Partners”. When retail hit crisis, the bonus disappeared for years. Only in March 2026 was it paid again, at two percent. After a loss of £234 million in 2023, management even explored selling a minority stake to raise one to two billion pounds for investment.

This shows the downside of bound assets: whoever cannot sell shares struggles to raise fresh capital in hard times. ZF therefore financed its acquisitions with debt, with well-known results. A hybrid therefore also needs an answer to the capital question, such as non-voting participations with a fixed return that bring money but no power. The form does not decide whether a firm does good business. It only decides who decides and who bears the consequences.

VIII. Handover as Refounding

What Arizmendiarrieta taught the five cannot be inherited. Every generation that never heard his speeches eventually sees the rules as mere regulations, and regulations can be voted away. That is how Mondragón’s pay ratio went from 1:3 to 1:6 and then to 1:9. All by majority, all democratic.

That is why the handover is the decisive moment. Whoever hands his firm to the workforce becomes, for a few years, a small Arizmendiarrieta himself. In the tandem year he passes on not only the craft but what the firm stands for. And he can make sure the rules outlive him by building them so that they do not depend on good will.

Goethe said it of inheritance, and it applies to every form of shared ownership: “What from your fathers you received as heir, acquire if you would possess it.” A cooperative, a foundation, a firm in workforce hands stays alive only if each generation acquires it anew.

IX. Two Keys Instead of a Saint

Mondragón had a priest, Zeiss had Abbe, Bosch had Robert Bosch. All three firms lived for decades on the character of their founders. That is fine, but it is not a model, because saints cannot be ordered.

A firm that belongs to its purpose and is run by those who work in it needs no saint. It needs two keys that are not in the same hand.

The economy becomes democratic only when its construction enforces it. Good intentions last a generation, good rules last longer.