The 2010s sold a seductive idea about flat organisations. Strip out the hierarchy, hand people a stake, and self-management would take care of the rest. A decade on, the mood is more sober. Holacracy experiments have quietly wound down, "flat" companies have rediscovered the need for someone to make the call, and the fatigue with self-management theatre is real. The lesson emerging from the wreckage is not that shared ownership fails. It is that ownership and decision-making are two different systems, and confusing them is how the experiments went wrong.
Shares are not a say
The most useful finding in this whole literature is also the most deflating for the flat-organisation romantics. Kruse and Blasi (1995) found that employee ownership alone does not automatically increase workers' perceived participation in decision-making. Give people equity and, by default, they do not feel any more involved in how the place is run. The share sits in an account. The meeting still happens without them. The felt experience of having a say has to be built separately, and the same review is explicit that participation must be combined with ownership for attitudes to move. Ownership is a claim on the returns. It is not, on its own, a claim on the decisions.
That single result explains a great deal of disappointment. Companies roll out an ownership plan, expect a surge of engagement, and get a shrug. They treated the certificate as if it were a governance change when it was only a financial one.
The confusion is understandable, because ownership in everyday language carries an implied power. To own something is to control it. But a fractional stake in a company held alongside hundreds or thousands of colleagues confers no practical control at all unless the firm deliberately grants a route to influence. Absent that route, the employee-owner is a small shareholder with a payout and no lever, which is materially different from the participatory ideal that ownership schemes are usually sold on. The word promises a say the structure does not deliver.
The scatter is the signal
Look across the studies and the pattern sharpens. Kruse and Blasi (1995) noted that results were widely dispersed from one firm to another, which they read as evidence that firm-level factors, the concrete choices a company makes about how it operates, drive outcomes more than the bare existence of ownership. This is easy to misread as noise. It is closer to a message. If two companies with the same ownership structure end up in different places, the ownership is not the deciding variable. Whatever they did differently is. And the thing they most plausibly did differently is how they allocated the right to decide.
The implication is practical. Stop asking whether ownership works in the abstract. Start asking how your particular firm distributes decision rights, because that is where the dispersion lives and where the outcome is actually determined.
Decision rights are a designable thing
The good news is that control can be engineered as deliberately as equity. Thomsen and colleagues (2025) describe steward-ownership precisely as the decoupling of control rights from economic rights, treating the two as separate instruments to be allocated on purpose. Their whole point is that you decide who steers as a distinct act from deciding who profits. That framing is useful even far outside formal steward-ownership. It says the question "who owns this?" and the question "who gets to decide?" are not the same question, and a serious organisation answers them separately and explicitly rather than assuming one settles the other.
And the design space is wide. O'Boyle, Patel and Gonzalez-Mulé (2016) found the small positive ownership-performance relationship holding across firm types, ownership vehicles, and sizes. Nothing in the evidence says decision rights must be arranged one specific way to keep the benefits. That leaves room to fit the decision structure to the company rather than forcing the company into a template. The freedom is real. So is the obligation to use it, because the default of leaving decision rights untouched is the failure mode the research keeps flagging.
The read
"Everyone owns it" is the start of the design problem, not the end of it. The evidence is unusually clear that spreading ownership without deliberately spreading decision rights leaves the felt experience of the workplace unchanged and the promised engagement unrealised (Kruse and Blasi, 1995). So the moment after you announce that the workforce owns the company, the real work begins. Which decisions move, to whom, at what level, and with what accountability. Answer that with the same rigour you brought to the cap table and shared ownership can deliver on its promise. Leave it to sort itself out and you will have redistributed the shares while leaving the power exactly where it was, which is the most reliable way to make people cynical about ownership altogether. A workforce that was promised a stake and handed a payout with no say learns a lesson the next scheme will not undo.
Sources
- Kruse and Blasi (1995). Employee ownership, employee attitudes, and firm performance (NBER WP 5277). Reviews find an overall positive ownership-performance association and higher commitment, but ownership alone does not raise perceived participation. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=225341
- O'Boyle, Patel and Gonzalez-Mulé (2016). Employee ownership and firm performance: a meta-analysis. A synthesis of 102 samples and 56,984 firms finds a small but positive and statistically significant link between employee ownership and performance, robust across designs. https://www.researchgate.net/publication/304459942_Employee_ownership_and_firm_performance_a_meta-analysis_Employee_ownership_a_meta-analysis
- Thomsen and colleagues (2025). Steward-ownership (curated research conversation), Business & Society. Defines steward-ownership by decoupling control from economic rights plus an asset lock, with Bosch as the canonical case. https://journals.sagepub.com/doi/10.1177/10564926251357812