Democratisation of Work

A central aspect of sociotechnical systems thinking is the notion of self-organisation, where complexity is best handled by collaborating self-governing teams. This may sound a bit academic and theoretical, which it is, but this way of thinking can have important and highly practical outcome for us as individuals, teams, companies, even society as a whole. A great example of this is studies done in the mines of post-war Britain on how changes in coal getting affected the people working there actually led to Section 4.2 of the Norwegian Work Environment Act stating: “…emphasis shall be placed on giving employees the opportunity for self-determination, influence and professional responsibility,…”

Why then is the thought of having self-governing teams in IT still so controversial?

“Control leads to compliance; autonomy leads to engagement.” ― Daniel H. Pink

In 1962 the Norwegian psychologist and researcher Einar Thorsrud was asked to lead a national research program on the democratisation of industrial work with the goal of finding ways to counter the increasing alienation and declining engagements of the workers. In order to give the project gravitas and support, a collaboration with the Tavistock Institute of Human Relations (TIHR) in London was initiated. TIHR was well-known for their research in social sciences, also in an industrial context with key figures like Eric Trist and Fred Emery. They had done some groundbreaking work with workers in the English coal mines, studying the effect that changing production system had on the worker productivity and engagement. Paradoxically they found that improved technology neither led to better efficiency nor improved worker satisfaction, even with better pay. A new term was coined, “sociotechnical system,” where the claim is that the technology and the sociology cannot be seen as independent parts, that the system as a whole can only be improved by joint optimization of those parts. Productivity and wellbeing are seen as emergent properties of the system, either in a positive or negative direction.

Einar Thorsrud and Fred Emery took this hypothesis and ran a number of experiments on four significant industrial companies in Norway to develop new organisational principles. They wanted to see if participative democracy could help counter the negative trend on productivity, quality, wellbeing, and health of the workers. Numerous social experiments in other contexts had shown that the democratic leadership style is superior to the autocratic one, replacing command and control by the leaders with group self-government. Thorsrud and Emery successfully replicated this, showing increased productivity, lowered costs, and higher quality work across all sites.

The learnings made in this program has had a lasting effect, culminating in a separate section in the Norwegian Work Environment Act in 1978 on the worker rights of involvement. Many companies around the world still choose to use the lesser effective autocratic approach though, where responsibility for coordination and control is located at least one level above where the work is being done. In the democratic model, each person is seen as more than a simple cog in the machine and the control is help by the group that the person is part of. This group is self-governing and sets its own goals and way of work. It is a move from hierarchies of control to hierarchies of coordinated functions. The former induce competition and tend to demotivate people over time, while the latter leads to cooperation and heightened motivation.

In the IT industry today, there is a lot of focus on finding new ways of working that are more efficient and effective, finding organisational designs that works better with the creative profession of software development. It may seem odd to spend half of this post on learnings from the industrial domains, but the thing is that many of the companies struggling to find new ways of operating is carrying with them the genetics from this scientific management that has proven problematic even for the?manufacturing industries. When doing organisational design for more knowledge-work based domains, it therefore makes sense to take a look at the learnings made in sociotechnical systems theory, especially the democratisation of the organisation and focus on self-governing teams. Research clearly shows that is better for both the business and for the people working there as we have seen.

In the agile community there is a lot of focus on the teams and what they are meant to be. Every organisation that wants to move in this direction would be well served to start by defining what “teams” mean for them before trying to copy success stories from other companies, be it Amazon’s pizza teams or Spotify’s squads. In companies that have the industrial genes, this transition is an especially troublesome as they are often structured in a coercive social structure and a lot of problems is rooted in how people are grouped and controlled. The teams are often organised by functions, where people doing the same things work together, sized according to the management reporting hierarchy. These teams are very different from the cross-functional teams in the agile organisations that they seek to copy. They also often have other types of teams though, that do fit the bill better, namely the project teams. They are put together with cross-functionality in mind and are given a clear and important business goal. The problem is that they are temporary and people go back to their functional silos when the project has run out of money. In here lies the irony: These project teams have a lot of what are needed in an agile team: Cross-functionality, a purpose, funding, agency, competency, and relatedness. They check pretty much all the boxes that is known to give people motivation and engagement as described by the self-determination theory. What is lacking is stability over time, ownership of the product created, and the important self-governing aspect. The first two is about defining team boundaries, e.g. using Domain-Driven Design, business capabilities, and Team Topologies, while the latter can best be achieved by giving the team democracy of work, letting them decide for themselves what works best for them in their context. The teams should self-organise and experiment, choosing their own processes, tools to use, roles needed, and even people to work with. Let the team be a self-contained entity by giving them responsible autonomy.

If democracy, humanisation, and self-governing was possible in the English coal mines and the Norwegian mechanical industry, it sure must be achievable in the modern IT industry. It will benefit the people as well as the business.