top of page
  • Geoff Mulgan

Collective mental health: the case for new measures and maps and a DSM5 for groups

The COVID-19 crisis has seriously damaged mental health all over the world. A high price has been paid by many children who’ve missed out on school and time with friends and by many of the elderly and isolated (the challenges and solutions are being documented by the International Public Policy Observatory on COVID).

But the crisis is also shining a light on the stresses faced by employees. In the UK, new evidence on health workers shows that nearly half of those working in intensive care units suffered from severe anxiety, depression, post-traumatic stress disorder, or harmful use of alcohol. There may be similar patterns in other groups on the front-line. But we just don’t know. And this is highlighting a bigger problem in how we look at mental health. This is nearly always seen as an individual issue requiring individual solutions. Yet in reality it’s as much a problem for groups and organisations.

In everyday conversation we often talk of toxic cultures and of organisations that are depressed, deluded or prone to compulsive behaviours. We know that individuals who join such organisations are powerfully affected by them: collective mental health influences individual mental health.

The best starting point for addressing problems is usually to find ways to see or map them. Unfortunately, we lack rigorous definitions of what these collective mental states might be. We lack good and comparable categories and measures, despite an explosion of psychometrics and surveys.

Psychiatry has a powerful tool to help solve this at the individual level. The DSM is the ‘Diagnostic and Statistical Manual of Mental Disorders’ that is the bedrock of American psychiatry Revised periodically – the most recent version is DSM5 - it is meant to capture the state of knowledge.

There is much that can be challenged in it. It rests on an uneven evidence base (only a tiny percentage of psychology experiments are successfully replicated); it uses often arbitrary categories; it lacks much solid knowledge about causal mechanisms; and it’s prone to ideological biases (homosexuality was only taken out of the DSM in the early 1970s – before that it was treated as a disease).

But it does at least provide some coherence. For groups and companies we could do better. Recent years have brought rapid growth in survey data on anxiety levels, depression and optimism. There are regular surveys of employees, and feedback through sites like Glassdoor and Indeed. Through social media, a treasure trove of data on moods, beliefs and anger is now available. The data is rarely straightforward – and there will always be a gap between what people say and what they feel. But this sea of data provides a good starting point, and possibilities that couldn’t have been imagined a decade or two ago, some of them quite scary, from analysis of company emails to analysis of facial expressions in Teams meetings.

So what might the categories be for new maps of collective mental health? An obvious one is depression. The usual definitions of depression include being unhappy or hopeless, having low self-esteem and finding no pleasure in things you usually enjoy. Our interest should be in spotting this in groups or whole populations where, although there is a wide range of levels of depression, the normal distribution curve has been shifted.

Delusion also looks relevant. The Cleveland Clinic defines delusional disorder as ‘a type of serious mental illness in which a person cannot tell what is real from what is imagined’, associated with ‘an irritable, angry, or low mood, and hallucinations (seeing, hearing, or feeling things that are not really there)’. Again, this is quite common in groups and even nations.

A third example is obsessive compulsive behaviour which is when an unwanted, intrusive and often distressing thought, image or urge repeatedly enters people’s mind, causes them distress and is associated with repetitive behaviours. This certainly looks recognizable at a group or collective level.

Fourthly, PTSD is clearly relevant – particularly for groups like the military or medical staff, and emergency services, that often have to face traumatic situations. The US army, for example, introduced extensive programmes to deal with PTSD a decade ago, led by Brigadier General Rhonda Cornum, and using the language of psychological fitness and resilience. Again, however, the emphasis was very much on individual diagnosis and treatment.

Each of these conditions – depression, delusion, compulsive behaviour and trauma - could be diagnosed at an aggregate or average level. But we would expect there to be big differences within groups and organisations. Status and power correlate fairly well with measures of physical and mental health. So any new methods of measurement need to look at the differences as well as the averages. The importance of such differences was highlighted in recent research by Angus Deaton and Ann Case on ‘diseases of despair’ which showed the serious decline in both mental and physical health amongst predominantly middle-aged men in the US whose status had declined.

Many organisations use well-being indicators – covering levels of anxiety, satisfaction with job, pulse surveys and Net Promoter Scores as a measure of engagement. Two widely validated questions, which could be used as standard parts of future measures, are: “In the past 3 months, have you ever not felt well enough to perform your duties to your normal standard, but attended work regardless?” and ‘do you feel energised by your work’.

What Works Wellbeing has developed workplace wellbeing surveys, and there is data from ONS Annual Population Survey, Workplace Employment Relations Study 2011, Mind Index, Public Health England’s topic overview on Measuring Employee Productivity in 2015 which looks at links between wellbeing and productivity and presenteeism.

Other interesting sources include the World Health Organization Disability Assessment Schedule, Version 2.0 (WHODAS 2.0) which assesses a patient’s ability to perform activities in six areas: understanding and communicating; getting around; self-care; getting along with people; life activities (e.g., household, work/school); and participation in society. Here too, it’s not hard to imagine group of collective variants of most of these: how well does a company or nation get along with others.

A decade ago I oversaw an analysis of the UK’s social needs (working with Claus Moser, former head of ONS, and Rushanara Ali amongst others) which attempted to combine analysis of classic measures of material poverty with detailed analysis of psychological well-being, connecting quantitative analysis and qualitative and ethnographic analyses.[i] Our hope was that this would become much more normal.

If we had better measures we could use them not just to address negative patterns but also to promote positives - to understand what helps groups to thrive. This has been the message of the ‘positive psychology’ movement and many nations now measure happiness, offering important insights. The best predictors in the annual World Happiness reports are survey answers to the question of whether you had friends or relatives you could rely on in a crisis. We might expect that to matter for firms and groups too. An organization I helped found – Action for Happiness – for example, now has very powerful evidence on how to boost both well-being and feelings of social connectedness. There is also recent interesting evidence about how firms that followed some approaches sustained employee wellbeing through the crisis much better than others.

We need to start growing a more systematic, empirical and conceptual field to make sense of these questions:

· What do we know about how collective mental health effects collective intelligence and decision-making?

· What do we know about directions of causation, from individuals to the group and vice-versa?

· How can we best spot pathologies – and their positive equivalents?

One legacy of the COVID-19 crisis could be a revolution in how we measure mental health and recognition that our own mental health depends very much on the groups we are part of. Anecdotes about toxic employers and deluded nations could be more grounded in data. With new measures we could start to harness that data to improve not just our individual mental health but our collective mental health too.

A version of this piece has been published by the World Economic Forum:


bottom of page