Participatory Action Research in OrganisationsWho Owns the Question

In the 1940s the Harwood Manufacturing Corporation made pyjamas at a plant in Marion, Virginia. Its head, Alfred Marrow, was a close friend of the psychologist Kurt Lewin, and under Lewin’s influence the factory became a site for some of the first experiments in how people respond to being managed. The best known of them was written up in 1948 by Lester Coch and John French. The company regularly needed to change how jobs were done, and the machine operators, most of them women recruited from the surrounding mountains, tended to respond badly: output dropped, people left, and hostility towards management rose.

Coch and French handled the same kind of change in different ways with different groups. One group was told about the change in the usual manner. A second elected representatives to work out the new methods and piece rates with management. In the remaining groups, every operator took part in the discussion. The group that was simply told saw its production fall by about a fifth and never recover. The representative group needed two weeks to get back to its old output. The groups that participated fully overtook their previous level by around fifteen per cent.

The paper became a founding text in the study of change at work, and it is still cited in management courses. Its title was Overcoming Resistance to Change, which tells you what the participation was for.

What Participatory Action Research Is, and Where It Came From

Lewin coined the term action research in 1944. In a 1946 paper, Action Research and Minority Problems, he described it as a spiral of steps, each made up of planning, action and fact-finding about the result of the action, and he argued that research which produces nothing but books will not suffice. Lewin had left Germany in 1933, and his work on groups was bound up with questions of prejudice and democracy. His best-known studies before the war, with Ronald Lippitt and Ralph White, compared groups of boys run under autocratic, democratic and laissez-faire leadership, and found the democratic groups more cooperative and more able to carry on working when the leader left the room. Action research was meant to let people study their own situation in order to change it, with the researcher as a collaborator rather than an observer.

In the 1970s another tradition took the idea somewhere more radical. The Colombian sociologist Orlando Fals Borda, one of the organisers of a 1977 conference on participatory action research in Cartagena, saw research as a way of building countervailing power for poor, oppressed and exploited groups. Knowledge, for Fals Borda, was never neutral. It had tended to serve whoever controlled its production, and participatory research existed to break that monopoly. Paulo Freire’s critical pedagogy pointed the same way. The people being studied were to become co-researchers, deciding which questions mattered and what the answers meant, and the research would belong to them.

This is the lineage that Michelle Fine works in, where participatory action research is first of all a claim about who gets to define knowledge. Its central question is political: whose account of a situation counts as true, and who benefits when one account wins.

The organisational strand took a different road. Through Harwood, through the Tavistock Institute in Britain, and through the growth of organisation development after the war, action research became a method for improving organisations. It kept Lewin’s cycle of planning, acting and reviewing, and it largely set aside the question Fals Borda placed at the centre. In 1999 the management historian Bill Cooke published a paper called Writing the Left out of Management Theory, about the way the history of change management had been told. Whatever one makes of the argument, the title names something real. The version of participatory research that reached most organisations arrived without its politics.

Why Participation in Organisations So Often Becomes Consultation

In 1969 Sherry Arnstein published A Ladder of Citizen Participation in the Journal of the American Institute of Planners. She was writing about urban renewal and anti-poverty programmes in the United States, but the article could have been written about almost any organisation. The ladder has eight rungs. At the bottom are manipulation and therapy. In the middle are informing, consultation and placation, which she grouped as degrees of tokenism. At the top are partnership, delegated power and citizen control.

Her sharpest sentence draws a line between going through the empty ritual of participation and having the real power needed to affect the outcome of the process. Participation without redistribution of power, she wrote, is an empty and frustrating process for the powerless.

The move into the workplace adds a complication Arnstein did not have to face. The residents she wrote about were dealing with agencies that could not dismiss them. An employee who speaks frankly inside a participation process is speaking to the organisation that pays them, reviews them, and may one day decide whether their role survives. Every comment offered in a survey box or a focus group is weighed against that, and people weigh it accurately. Anonymity is promised, and staff in small teams know how little it protects them: a remark about the only manager in a team of six identifies its author nearly as well as a signature. As The Illusion of Psychological Safety traces, an invitation to speak up is only as safe as the consequences people have watched others suffer.

Most participation in organisations lives on the middle rungs, and the engagement survey shows how. Its questions are written by the people whose leadership it will measure, or by a vendor they have chosen. Staff answer within those questions, in the form those questions allow. The results are interpreted by HR and senior leaders, who decide which numbers matter and what they mean. An action plan follows, chosen by the same people, and a few months later a poster appears under the heading You said, we did. Each step is presented as listening. At no step do the people answering have any say over what happens next. When response rates fall and organisations begin talking about survey fatigue, they usually mean apathy. It is frequently closer to an accurate judgement about how much difference answering makes.

Formal consultation over a restructure sits on the same rung in its purest form. Employment law in many countries requires it, and it is often run carefully and in good faith. It is also commonly run on a proposal that has been costed and signed off at board level before staff see it. People are asked for feedback on a structure in which their own roles may already have disappeared. Occasionally the feedback changes a detail. The proposal almost always survives. Everyone involved understands this, which is why consultation meetings are often so subdued: both sides are performing a process whose outcome both sides can predict.

The co-design workshop, convened once the decision has been made, belongs one rung higher, in placation. People are invited to shape how something will happen, never whether it will, and the invitation is real enough to make refusing it look obstructive.

Arnstein’s therapy rung is the most uncomfortable to read. She described programmes that treated powerlessness as a form of mental illness, so that citizens were treated for their supposed pathology rather than for the conditions producing it. She called this both dishonest and arrogant. Its organisational relatives are familiar: the resilience workshop offered in response to complaints about workload, the change curve that locates staff objections somewhere between denial and acceptance. As Organisational Trauma explores, the language of healing can turn a grievance about a decision into a condition the aggrieved are expected to work on.

Harwood belongs in this picture too, and it deserves to be seen fairly. Bernard Burnes, in a 2019 history of the company, warns against treating it as a workers’ paradise. Experimentation had to be subordinate to the needs of the factory. Like many progressive employers of the time, Marrow regarded unions as an obstacle, and Harwood resisted the unionisation of its Marion plant until the mid-1940s. After Marrow left, worker participation died out under new managers.

None of that makes the participation fake. The operators at Harwood did shape their own work, and they did better when they did. Burnes’s history makes something else clear, though. Participation was adopted because it worked by management’s measure: output went up and resistance went down. That is the paradox at the centre of participation at work. It is welcome for as long as it produces the results the organisation already wanted. The real test is what happens when it produces something else, and Coch and French were never in a position to run that test. They found a way to reduce the operators’ resistance. As Why Change Doesn’t Fail Because People Resist It argues, what gets labelled resistance is often information, and the experiment had no place to record what the resistance might have been saying.

Whose Knowledge Counts When Organisations Research Themselves

There is a familiar sequence in organisational life. A consultancy is engaged to diagnose what is wrong. Its staff interview people across the organisation, who tell them, often in considerable detail, what the problems are and where they come from. Some weeks later the consultants present their findings to the leadership team, who find them illuminating. Much of what is presented was known throughout the building before the consultants arrived. What changed was who said it. Front-line knowledge became credible once it had been translated by someone with standing, in a format leaders were used to receiving. Staff who later see the slides recognise their own words in them, now carrying someone else’s logo, and draw their own conclusions about whose knowledge the organisation is prepared to hear.

This is the problem participatory action research was invented to address. Fals Borda’s point about the monopoly of knowledge applies inside organisations with little adjustment. People who do the work know things about it that nobody above them can see. That knowledge counts only when it passes through channels the organisation recognises: the survey, the report, the consultant’s slide. Each channel strips out part of what was known, usually the part that implicates the people commissioning the channel.

Some organisations go further and commission research that is genuinely participatory. Staff are trained as co-researchers and given time to study their own workplace with their own questions. When it works, it produces knowledge no survey can reach, because people tell colleagues things they would never tell HR, and because insiders recognise patterns outsiders miss.

It also sets off tensions that the organisation rarely anticipates. The research is funded by the organisation, and what it finds may implicate the people who funded it. The co-researchers have put their names to the inquiry, and if it concludes that the problem lies in how the place is managed, they are the ones who said so. They are no longer simply staff, and they are certainly not management. Their belonging has been altered by the work, and the protection the project offered them lasts only as long as the project’s sponsor stays in post and stays interested.

Then there is the report. It belongs to the sponsor, who decides what is circulated and what is set aside for later. People who gave their time and their trust watch their findings pass into a process they have no part in. As The Learning Organisation notes, organisations are better at gathering lessons than at being changed by them. The next time someone invites these people to participate, they will remember what happened last time, and they will be right to.

Genuine participatory research also asks something of the sponsor that organisations find unusually hard. It requires them not to know the answer in advance, and to accept that they may become part of what is being studied. Most commissioned research is designed, consciously or not, to confirm a direction already chosen. Unknowing is tolerable when the subject is the market or the customer. It is much less tolerable when the subject is the leadership team.

The deepest shift in participatory research concerns meaning. In conventional organisational research, staff supply the data and others supply the interpretation. Participatory research moves interpretation to the people whose situation is being interpreted. That is its whole point, and it is also where it most often fails inside organisations, because the final reading reverts, at the last step, to those with the authority to act on it. The co-researchers may have written the findings. Someone else writes the response.

There is one further risk, and it is easy to miss because it looks like trust. When staff are asked to diagnose the problems of their own workplace, they are often expected to solve them as well. A team invited to investigate its workload may find itself responsible for redesigning its own processes to cope with a headcount it had no say in setting. Participation of this kind distributes responsibility downwards while keeping authority where it was. It feels like being taken seriously. It can function as a way of handing people the consequences of decisions they were never invited to make.

Reading Participatory Action Research Through The Fractured Lens

Lewin wanted research that would not end as books. Organisations have found their own version of what he feared: research that ends as reports, and participation that ends as a record of having participated. The difference between the two traditions that grew from his work comes down to a simple matter that is rarely simple in practice. In one, the people being studied help decide what the research is for. In the other, they help the research achieve what someone else has already decided it is for.

That distinction is not the same as sincerity. Plenty of consultation is sincere, and plenty of the people running engagement surveys want very much to hear what staff think. The question is structural. It concerns who sets the question, who reads the answer, and who is left holding the consequences, and these are matters of power whether or not anyone involved thinks of them that way. An organisation can be warm, consultative and well intentioned and still keep all three in the same few hands.

The operators at Harwood were invited to help decide how a change would be made. Whether it should be made was not open to them, and neither was who owned the factory, how the gains in output would be shared, or what would happen to the plant when the pyjamas could be made more cheaply elsewhere. Burnes records that a company which began in the garment sweatshops of New York ended its life in sweatshops in Honduras and Costa Rica, and the Marion plant closed in 1992. Every participation process has a question that is kept off the table. The published account of the Harwood experiment records what the operators produced, week by week, in careful figures. What they would have asked, had they been given the question to set, is not in it.