Sharing Power, Not Just Limiting It

4 of 6 · Clarity. Why your organization hangs on a few heads, and how to change that without tipping into formlessness.

When it comes to distributing power, organizations tend toward two extremes. In the first, everything runs across one desk. The company holds its breath whenever that one person is on vacation. It lives and dies with that person's skills, goals, willingness to reflect, health, and mood. In the second, people have tried to abolish exactly these dependencies, hierarchy gone, everyone equal, no more boss, and after a while discover that a hierarchy exists again anyway. Nobody's named it, and nobody controls it.

Interestingly, both extremes share the same problem. Neither has actually shared power. In one case it's concentrated, in the other it's hidden, a problem that shows up in miniature in almost every organization. For you as the person responsible, both are the same weakness: the organization depends on people instead of structure. An expensive single point of failure, built into the management model.

Four ways of handling power

There are four fundamental ways to handle power. You can concentrate it, the monocracy at the top. You can limit it, leaving it concentrated but building walls around it. You can share it, distributing it across many hands. Or you can try to abolish it and realize too late that this doesn't work. Power is a permanent property of organizations. They can't do without it. It doesn't disappear when you tear down the structure. It just becomes invisible. Most debates about flat hierarchies confuse the third path with the fourth.

Limiting is good. Sharing is more

The classical constitutions, the Basic Law and the American one, are masters of limiting: separation of powers, checks and balances, independent courts. That's a lot. But it has a limit. Power stays at the top. It's kept within bounds, not distributed. Abuse of power becomes subtler, not impossible.

Switzerland does something different. It shares power on two levels. Downward, through subsidiarity: authority sits at the bottom by default, with the municipalities, then the cantons, and only last with the federal government. That reverses the burden of proof: it's not decentralization that needs justification, but centralization. And at the top, through the collegial principle: the Swiss government has no boss. The Federal Council has seven equal members; the Federal President is only first among equals and changes every year. That's the constitutional model against monocracy at the top.

Sharing is not abolishing

This is where most power-distribution experiments fail. Sharing power doesn't mean tearing down structure. It means distributing power into clearly defined, accountable units, with structure, not without it.

Valve chose the fourth path: no fixed structure. What it got wasn't shared power but hidden power: informal circles whose influence nobody named and nobody limited. Buurtzorg, the Dutch home-care organization founded by Jos de Blok, shows the third path. Here too there are no classical managers; care is organized in small, self-steering teams. But these teams don't stand in a void. They have coaches, a shared method, and a lean platform that keeps their backs free. Power is genuinely distributed, but it rests on a structure that holds. The difference between Buurtzorg and Valve is the difference between a net and a hole. Both removed hierarchy. One provides support. The other takes it away.

Size is the wrong question

Eventually, when growth comes up, someone objects: shared power might work in a small team, but not in a large company. That's not true. And the counterexamples are bigger than most mid-sized companies. W. L. Gore deliberately caps its sites at around 150 people, so everyone still knows each other and trust, not bureaucracy, governs. Haier splits over 80,000 employees into several thousand micro-enterprises of roughly ten to fifteen people each, connected through internal markets. Both are large organizations built from small units.

That resolves the size question into a better one:

Not "how big is it allowed to get?" but "what holds the small units together, and is that connective layer itself constitutionally anchored?"

Three connective layers have proven themselves: culture and convention (Gore), internal market and contract (Haier), and law (cooperatives with a stated purpose and one-member-one-vote). Each one holds, and each one degrades without its own constitution: the internal market tips into exploitation, culture tips into clique. Human-scale units, then, aren't a limit on your ambition. They're a building block. You don't trade size for cohesion. You keep the unit small enough that trust still holds, and build federal structures upward from there.

The hard truth

Sharing power isn't the same as delegating power. Delegation keeps power at the top and lends it downward, revocable at any time. That's just a longer leash. Genuine sharing gives power away, binding enough that you don't pull it back at the first uncomfortable moment. Shared power you can reclaim at any time is only borrowed power, and at best gives a feeling of control. And you can't dump power on people who aren't prepared and call that sharing. That's abdication. Buurtzorg invests in method and coaching precisely so the teams can actually carry the power they're given.

What this means for you, the two questions

The first question, as always, is for your organization. Three tests show where it really stands: does power sit in many hands, or is it concentrated and merely controlled? What happens if the most important person is out for a month, does it keep running or stall? And does the smallest capable unit actually decide, or does it just execute? Only one answer points the right way: an organization that doesn't depend on the presence of individual people. That's the real point of sharing power, not less leadership, but leadership that doesn't tie the organization's survival to a single desk.

The second question is for you: where do you like being indispensable? There's almost always a spot where it secretly feels good that nothing runs without you, and that's exactly where letting go is hardest. What's stopping you from truly sharing power there: honest worry that it'll go wrong, or the quiet satisfaction of being needed? Here too, for now it's enough to notice it and name it.

With that, the statics are in place. Now comes purpose. There remains the biggest question every carefully built structure must eventually answer: what is all this actually for, and does that purpose hold even when it costs money? That's the next essay.

Resilient. Human. Future-Fit.

A series of seven essays on what organisations can learn from 800 years of constitutional history. All English essays.

Previous: Participation Is a Productivity Function, Not a Feel-Good Factor
Next: Sustainability as Purpose, Not as a Goal