The following post is an adapted version of a book chapter in a recent publication by UCL Thinking about Behaviour Change: An Interdisciplinary dialogue.
On July 19th 2011 Guardian Comment is Free published my response to Baroness Neuberger’s House of Lords Commission report on behavioural insight in public policy. The headlines the previous day proclaimed that ‘nudge is not enough’, but the real story was of peers pulling their punches. The report was keen to imply but curiously unwilling to say that Science, the presumed foundation of behavioural insight, doesn’t provide a legitimate case for shrinking the State. It felt exciting to be at the forefront of public debate, but there were no paparazzi following me home that evening. Instead, when I scrolled through the assembled comments, I found some pungent counsel from CorneliusLysergic posted at 3:20pm: “F**k off with your change our behaviour sh*t. Just f**k off.”
I cannot match Cornelius’s conviction, but my own qualms have gradually clarified over 6 years of working closely with behaviour change. There is one feature of the domain in particular I want to develop here – namely the slightly pernicious way that behaviour change gets excited by the idiosyncrasies of human cognition while simultaneously stripping us of interiority. (For a broader political critique of behaviour change, I recommend Will Davies’s excellent, if discomforting book, The Happiness Industry.)
We have access to forms of knowing and evaluating and justifying from our own lives that we cannot afford to do without, but which we tend to disavow or neglect. We speak of such content as being ‘anecdotal’, as if anecdotes didn’t shape and direct understanding; ‘subjective’, as if perspective and context wasn’t of primary importance; and ‘personal’, as if there is anything else that is meaningfully universal.
So here’s what I want to explore: what if those implicit perspectives on behaviour that stem from such ‘epistemic hinterlands’ – perspectives on knowledge that arise from formative aspects of subjective experience – turn out to be the sine qua non of good behaviour change research? What if the recurring challenges relating to research legitimacy, predictive and explanatory power, and model building, stem from coming too directly and too quickly at behaviour change as an applied tool, neglecting to stop to think what it means to us, and why?
I develop this argument below; but, in the interests of consistency, let me first apply my own conclusions and reflect on how I came to this view.
A personal account of epistemic hinterlands
I am currently the Director of ‘The Social Brain Centre’ at the RSA in London. We’re a small semi-autonomous unit of researchers focussing on the ways in which implicit understandings of human nature shape theories of change and impact in policy and practice. We are nested within a larger organisation that is part global ideas platform, part membership organisation and part policy research institute. I have been working on the theory and practice of behaviour change with public, private and third sectors organisations at this curious nexus of academia, media, practice and policy for about six years.
Neither ‘behaviour’, nor ‘change’, nor ‘behaviour change’ can withstand a canonical definition because they are too porous and contested, so to make sense of this work I had to find my own felt sense of these terms. Relevant influences on my view of behaviour change, some of which I unpack below, include the physiological know-how involved in being a type-one diabetic since I was six; the exacting cognitive, emotional and volitional process of becoming a chess Grandmaster; searching for an academic home and not finding one; and the spiritual conviction that people can not only change their behaviour, but also their practices, their values, their sense of themselves, and thereby in some ethically meaningful sense transform for the better.
The confluence of these biographical influences explains why I have never wanted to disentangle the empirical matter of what we do from the philosophical question of what we are, and the ethical question of what we should be.
Chess was my main vehicle of identity formation and development. I can see now that I used chess as a way to signal I was bigger than the perceived constraints of diabetes, and more profoundly as a daily escape from adolescent growing pains, parental separation and the harrowing incursions of family mental illness. The collateral benefit of having chess as a coping mechanism was a deep love for the beauty of ideas, the experience of learning from mistakes, and a felt sense for what it means to grow rather than merely change.
In the second decade of my life, by regularly analysing my games, I spent thousands of hours forensically examining how thoughts and emotions arose automatically and how they shaped decisions and results. Being good at chess, I now realise, is about being an instinctive choice architect, rapidly deciding what needs to be known to choose the next move, and framing complex problems for opponents in ways that maximise the probability of mistakes.
I became a Grandmaster, competed and trained with World Champions, and won the British Championship in three successive years (2004–6). Two of the books I wrote drew on sports psychology and cognitive science to make sense of the curiously difficult challenge of post-plateau improvement in adulthood (Rowson, 2001; Rowson, 2005). Through playing, teaching and writing about chess I became fascinated by the challenge of supplanting bad habits with good ones, and came to view this habituation challenge as a microcosm of the human desire to feel free.
I didn’t view it as such then, but in Kahneman’s terms my chess career was characterised by putting ‘system 2’ (the effortful, conscious system) to work on ‘system 1’ (the fast, automatic system) often enough to know that system 2 was the most important, even if system 1 was more powerful; just as in chess the queen is much more powerful than the king, but the king ultimately matters more. This precious space of inquiry and adjustment, where the part of us that is intentional kindly attends to and gently adjusts the part of us that is habituated, is our wellspring of vitality – our way of staying awake to ourselves. I came to see this encounter of conscious and automatic systems as a deeply valuable middle way, superior in a moral sense to abject surrender to the automatic system or hubristic underestimation of its defining role in our lives.
Chess was always plan B, but there was no plan A, and without any particular sense of direction I enjoyed eight years of intense but unconsummated flirting with a range of academic disciplines and university tribes. Politics, Philosophy and Economics at Oxford became Mind, Brain and Education at Harvard, which led to some research methods training and then a PhD thesis at Bristol under the supervision of Psychologist Professor Guy Claxton in what was substantively a mixture of philosophy and psychology, but was technically ‘Education’.
And it was an education, in the literal sense of ‘drawing out’. I began my PhD with an abstract and analytical mode of inquiry formed in chess tournaments, Oxford tutorials and Harvard seminars; trained to get to the point and make it as sharp as possible. Whenever I attended research classes and seminars at Bristol I was therefore dumbstruck by the extent to which my cohort – mostly comprising female teachers and therapists – answered academic questions very differently, by sharing stories from their own experiences and practices. I was at first highly uncomfortable and even a little embarrassed by this form of inquiry, but now feel deeply grateful for having had the chance to experience its legitimacy.
My planned methodology was a post-positivist experimental design on the precursors of wisdom in adolescence (‘proto-sagacity’) and how they would measurably change following classroom interventions. ESRC gave this research plan a score of 90/100 when awarding funding, but at this point I had given little thought to what wisdom meant to me. The more my cohort asked me simple questions like – Why wisdom? Why now? Why you? – the more my sanitised, pseudo-objective approach began to feel hubristic and inauthentic. Something is very wrong here, I thought. In an effort to appear like a serious researcher I was running away from myself.
My doctorate came into shape as a sustained reflection on the process of trying to make sense of what it might mean to become ‘wiser’, which involved looking at the automatic system from an ethical rather than competitive vantage point (as I had with chess). That process involved some conceptual analysis of definitions, critiquing the meta-theories and methodologies of a range of psychometric models of wisdom and trying to make sense – partly through my own meditation efforts – of what exactly is supposed to transform through spiritual practice. I also created a data stimulation method based on selected vignettes, to tease out some of the features of wise action. The shortest and sweetest illustration is the story of Mahatma Gandhi boarding a train that had just started to pick up speed. One of his sandals fell off in the process, and he instinctively removed the other one and threw it down, so that somebody else would have a pair to wear. That’s the kind of behaviour that makes me want to change.
By the time the thesis drew to a close I was an expectant father, and had grown weary of squeezing meaning out of 64 squares, so I started looking for a soft landing in the real world. I joined the RSA as a Senior Researcher in their ‘Connected Communities’ programme, funded by the Department of Communities and Local Government (Rowson et al., 2010). This work on how social norms spread through social networks in deprived communities made a deep impression, and sat alongside the time bound Social Brain project, which I took over within a year, after a range of staff changes. After expanding the range of our activity and raising some money I became Director of the Social Brain Centre; the change of title signalled an indefinite commitment to the RSA’s work on behaviour change.
The elision between brain and behaviour is the kind of casual ontological slippage that complicates the academic/policy interface; but the bigger issue was that my default settings on behaviour change were at odds with the prevailing mood of the time, suffused as it was Nudge, MindSpace, and the birth of the Behavioural Insights Team. I felt excited because rethinking human nature was now mainstream, but I watched in disbelief at the speed at which the potentially broad, deep, multi-faceted and interdisciplinary terrain of behaviour change was quickly framed (yes, really) as applied behavioural economics. There are many broader perspectives of course, but in my professional orbit it took only about three years, from roughly 2009–2012, for an impressive but relatively limited and limiting intellectual hegemony to emerge.
Nobody ever decreed that all behaviour change is applied behavioural economics, nor did they argue that human nature is merely that which is revealed through experiments about decisions, but in practice it felt like a fait accompli. As the rich and multifaceted ontology of behaviour was tacitly equated with the relatively hollow epistemology of decision theory within economics, ‘behaviour change’ became technocratic, apolitical, utilitarian and psychometric. It felt like the prevailing view of behaviour was such that it captivated us intellectually without challenging us ethically or spiritually, which I found vexing, and actually quite annoying.
Partly as a reaction to this misplaced hegemony, I found myself interpreting ‘behaviour’ very broadly indeed. The Social Brain team has worked on the cultural constraints on how police reflect on their decisions (Rowson & Lindley, 2011), fuel-efficient taxi driving (Rowson & Young, 2011), psychological foundations of The Big Society (Rowson et al., 2012), social marketing to reduce child abuse, the experience of risk and trust in informal care relationships, and relatively ‘conventional’ forms of behavioural insight to improve financial capability and reduce the attainment gap in education (Spencer, Rowson & Bamfield). Most recently we examined the limitations of behaviour change in the context of climate change (Rowson & Corner, 2015), and took a deep dive into ‘spirituality’ as the foundation of transformative behaviour change.
Reflexivity highlights why the personal is political
If there is a pattern that connects this work, it’s the conviction that reflexivity – circular relationships between causes and effects in thinking agents and most complex social systems – cuts in a deep and distinctive way when it comes to the idea of directing and measuring behaviour change.
The point is that humans do things differently as a result of knowing why they are doing them or being asked to do them – sometimes as a form of defiance, but mostly as a form of creative and generative intelligence. Our descriptive and explanatory accounts of how and why people behave as they do become part of the very thing they purport to describe and explain. Our models sit midway between what we believe to be the case and what we think we should believe. It is not anti-scientific to be normative when your subject matter oozes normativity in this way.
How we talk about, model and measure behaviour can either emphasise this point as central, or downplay it as peripheral. I see it as central. Reflexivity, I believe, is a feature rather than a bug in human behaviour, which is therefore only ever partially predictable in principle. The ill-fitting, multiply-influenced, volatile, situational and porous nature of human behaviour will always elude or subsume the best models designed to capture it. You might even say we are at our most human when we do not behave.
This point is not about gratuitous subversion, and it’s of huge societal importance in 2015. Collectively, we are just waking up to the power and ubiquity of defaults, salience, comparative judgments, loss aversion and so forth, many of which have shaped our lives for good or ill for many years, and not all of which are belief-independent i.e. knowing more about their effects on us will change how we respond to them.
The extent to which you face up to the inherent waywardness of reflexivity is an open choice for researchers, and one that matters politically. Although the idea of a new generation of ‘behavioural natives’ is not unimaginable, it is hard to see the majority of us becoming, for instance, immune to advertisements by detecting their underlying principles, or creating our own technological defaults as second nature. Still, we should be open to the possibility that behavioural models could be more democratic in spirit, and reflect on what behaviour change of the people, for the people, by the people might look and feel like.
Reflexivity alone won’t get us there, but it does highlight a form of freedom that is different in substance and spirit from the top-down model of a libertarian paternalist who conventionally seeks to maximise choice and direct it. Instead reflexivity points towards a more lateral model of influence in which informed mentors – not least behaviour change researchers – advise on the general principles that underlie choice and behaviour so that they can be refashioned for purposes that are as autonomous as possible.
At present, this shift of emphasis sounds abstract and utopian, but isn’t it closer to how we might want behaviour change to manifest socially and politically? Accepting reflexivity as a central feature of human behaviour makes a more reflexive approach to research productive and legitimate. For instance, if you want to measure the impact of mindfulness meditation on stress reduction, you can measure cortisol levels in those who meditate, but your explanatory account will be deficient unless you learn to meditate yourself.
If you are just tweaking contextual inputs to shape discrete actions (here I’m thinking of those witty spillage saving urinal flies in Schipol airport) you can get away with a dispassionate third person (‘he’, ‘she’, ‘it’) account, but that’s the easy stuff. If the behaviour you are looking to change is less discrete and more deeply connected to broader cultural, social, economic and political factors the behaviour will generally be experienced and enacted in the first (‘I, we’) and second (‘you’) person, in ways that need to be understood for an intervention to be worthwhile.
For instance, some personal behaviours appear to be obvious environmental gains and can be measured as such if the model starts and ends with a target behaviour alone, but with a bit more empathy, imagination and political honesty, such outcomes are often phyrric victories. To give a specific example, some apparently positive behaviour changes may actually increase national carbon emissions when factors like moral licensing (eg ‘I fly every week but drive very little’) single action bias (‘I know I need to do my bit for the environment – so I recycle’), consumption-based emissions (‘most of the products in my home were imported from China but their embodied carbon don’t effect national emissions targets’) and systemic forms of rebound (‘I saved £100 on my heating bill, so I’m getting a cheap flight to Spain’) are factored in.
In such complex cases, understanding the particularity of the people and context being studied involves (perhaps even demands) introspective, associative and empathetic forms of inquiry, not a third (‘he/she/it’) person account of apparent stimulus-response relationships, however sophisticated they might be. We need to know not only what people are doing, but why they are doing it, why they think they are doing it, and how those perspectives are shaped by socio-economic and cultural factors. Only that tangled interplay of influences will begin to make sense of what might follow from an intervention. Reflecting on that goal, how can we authentically and authoritatively make sense of what is called for unless we involve ourselves more fully in the inquiry?
To conclude, the biography and interiority of people working on behaviour change has methodological legitimacy and epistemic warrant, and we should give it greater salience. I believe both behaviour and research about behaviour are best conceived as fundamentally reflexive processes, and that doing so will ultimately help the empirical and theoretical fields of behaviour change converge in the complex cases we care about. Understanding behaviour requires psychological insight, yes, but it also requires skilful introspection, cognitive and emotional empathy and political awareness – factors that are shaped by our lives as a whole. As long as we remain strangers to our epistemic hinterlands, we don’t really know what we are trying to do.
Dr Jonathan Rowson is Director of the Social Brain Centre at the RSA. From 2016, after six wonderful years at the RSA, he will begin a new venture, co-founding his own institute – Perspectiva - which will examine public policy challenges by integrating systems, souls and society; connecting the evolution of the political economy to relationships and the inner world. You can reach him on Twitter @jonathan_rowson
Rowson, J. (2001) The Seven Deadly Chess Sins, Gambit, London
Rowson, J. (2005) Chess for Zebras, Gambit, London.