Organisations that abolish hierarchy

Several arrangements coexist: holacracy, formalised by Brian Robertson from 2007, which replaces posts with roles and decisions with written processes; so-called liberated companies, which remove intermediate layers of management; and worker cooperatives, whose democratic functioning is long-standing and governed by law.

These formulas share nothing but a critique of pyramidal hierarchy. Their decision rules, their relation to ownership and their legal framework differ entirely. The term liberated company was popularised in France by the book by Isaac Getz and Brian Carney, published in English in 2009 and then translated, which is closer to testimony than to research.

An American online retailer adopted holacracy in 2013 and then partly abandoned it. The episode was followed by the press and by a few academic studies.

The difficulties reported concern the burden of coordination, ambiguity over responsibilities and the departure of a notable share of employees, some of whom took a severance payment offered at the time of the transition.

The case is often cited as proof of failure, which says more about those citing it than it can establish. A single case cannot separate what is due to the arrangement from what is due to the company, the pace of its transition or the economic climate of the moment.

A text by Jo Freeman published in 1972, arising from the American feminist movement, argues that an organisation without formal structure is not without structure: it has one, informal, unelected and impossible to challenge.

That argument is regularly invoked in analysing flat organisations, where field studies find positions of power based on seniority, access to information and closeness to the founders.

It echoes an older thesis by Robert Michels, who described in 1911 the tendency of democratic organisations to produce an oligarchy. The opposite argument also exists: making power explicit and open to challenge is precisely what the most formalised formulas claim to do.

Available research consists mainly of case studies, with the limits that implies: no comparison, selection of volunteer organisations, and accounts supplied by participants.

The rare quantitative comparisons concern cooperatives, whose productivity and survival have been studied over long periods, with results that do not put them at a disadvantage relative to comparable conventional firms. These results do not carry over to the other formulas, since a cooperative changes the ownership of capital and not only the organisation of work.

Labour law presupposes an employer and a power of direction, which sets limits on experiments: legal responsibility does not dissolve into an organisation of roles. The employer's duty to ensure safety, for instance, remains whatever the internal distribution of decisions.

The most stable arrangements are therefore those that change how decisions are distributed without removing the legal structure, a configuration found in most of the cases that last.