On Nudging
FOR two years our daily lives in the face of Covid have been ruled by a coterie of experts in ‘behavioural psychology’, a breed of psychologist that seems to have appeared from nowhere. I speak as a psychologist who has been in the business for a good many years. It is worth asking, who are these people, and how did they come by these insights into the workings of the human psyche? Insights so profound that they merited our unthinking obedience?
Psychologists in the 1950s were given to claiming expertise in behaviour, in the rather circular sense that they would study any organism that ‘behaved’. Pavlov’s work on conditioned salivation in dogs was the paradigm case until B F Skinner extended this to include what he called ‘operant conditioning’, showing how pigeons could be induced to behave in quite extraordinary ways by scheduling their rewards. For a few years this kind of raw behaviourism had a powerful influence in psychology. It fell to a bright young linguist named Noam Chomsky to put paid to it with a demonstration that behaviourism could not account for the complexities of human language and thought. But this leaves my question unanswered: if the behavioural psychologists currently bossing us about are not behaviourists in Skinner’s sense, what on earth are they and where did they come from?
A useful orienting date for an answer is 2011 – the year the House of Lords Science and Technology Committee produced a report on ‘Behaviour Change’. A succession of experts from the world of social psychology had induced their lordships to swallow Nudge Theory hook, line and sinker (the infamous MINDSPACE protocol dates from 2010). Their report is still worth reading, if only to see how groupthink works. The committee accepted as hard fact a host of assertions which were, even then, patently questionable; for example, the proposition that our thought processes (and behaviour) are largely under the control of factors of which we are unconscious. We may be rational if we want to be, but it’s the environment that unconsciously nudges us, like prodded sheep, one way or another.
In a nutshell the noble lords seemed content to accept the proposition that systems of ‘unconscious thought’ determine how we act. Bizarrely, they even allowed themselves to be scolded for wrongthink by countenancing the idea that human beings might act as rational agents. Commit to that heresy, they were told, and you fall into the ‘fundamental misattribution error’ (which I accept sounds really bad). I can only conclude that their lordships thought all this stuff applied only to other people.
The report was caught up in (and possibly explained by) the intellectual fever surrounding the publication of Daniel Kahneman’s best-seller, Thinking, Fast and Slow. This was the book (a kind of Freud without tears) which allowed the notion of two systems of thought to enter popular culture. It was beyond doubt, Kahneman urged, that behaviour was readily altered without our being conscious of the determining factors. And indeed, the data at that time in support of ‘social priming’ were remarkable:
The range of these effects seemed constrained only by the imagination of the social psychologists carrying out the work. Countless articles appeared, many in respected journals, all broadly supportive. Nonetheless, if by now you are finding this whole bag of tricks a little fishy, you would be right. The brutal truth took far too long to emerge, but virtually none of the innumerable studies on which the elaborate edifice of social priming rested could be replicated. That’s none, as in zero.
The strange death of social priming was only part of a much wider replication crisis which engulfed the discipline of social psychology at that time, and it soon became evident that a replication problem could all too easily morph into something much worse. Even as the Lords’ report was being prepared, a misconduct hearing was set up in the Netherlands to examine the work of the Dutch social psychologist, Diederik Stapel, revealing an ‘astonishing lack of familiarity with elementary statistics’. Dr Stapel retracted dozens of published journal articles and admitted falsifying data. But that was not really the point: how could his many co-authors have failed to notice results so obviously dodgy? Why did successive generations of university doctoral students accept these plausible fictions at face value? Not to forget referees of Stapel’s publications: ‘It is almost inconceivable,’ the inquiry concluded, ‘that reviewers of the international “leading journals”, who are deemed to be experts in their field . . . did not notice the reporting of impossible statistical results.’
Kahneman, who received the Nobel Prize for his contribution to behavioural economics, wrote an open letter to social psychologists working on priming, warning them that ‘a train wreck’ was approaching. The statistical sleight of hand which had allowed investigators for years to fish for ‘significant’ results until they found them had precipitated a catastrophe. Significantly, he added: ‘The problem could well be more severe in your field than in other branches of experimental psychology, because every priming study involves the invention of a new experimental situation . . . I believe that you should collectively do something about this mess.’
It is reasonable to ask whether something was, in fact, done about the mess. And the answer is very little. Behavioural psychologists in some sceptical quarters have become hate figures but, crucially, any such criticism invariably runs along with acceptance that ‘nudging’ as a methodology is rooted in sound psychological principles. As I hope I have made clear, the evidence shows that this is far from the case. It follows that in setting up a Sage subcommittee drawing (by accident or design) exclusively on the expertise of ‘behavioural scientists, health and social psychologists, anthropologists and historians’ the UK government was guilty of serious misjudgment. Was it at all defensible to seek advice from a discipline still reeling from an insoluble replication crisis?
In light of Kahneman’s remarks it was equally perverse to ignore expertise from areas of psychological research uncontaminated by issues of replication failure and fraud. To take but one example (of hundreds), some of the most advanced experimental work in developmental psychology goes on in UK universities. Yet the technical advice of developmental psychologists on the consequences of lockdown measures on schoolchildren seems not to have been sought – or if it was sought, it was ignored. Were psychologist specialists consulted on the behavioural consequences of ‘masking’ on children (and their teachers) for long periods? In place of credible scientific advice, absurd nostrums such as ‘I wear my mask to protect my mate’ are simply embarrassing (apart from being false).
It will be argued, of course, that the proof of the pudding was in the eating and that extraordinary levels of public compliance during this pandemic reflect the power of the psychological techniques that brought it about. This seems to be at least part of the motivation for Dr Gary Sidley’s complaint to the British Psychological Society that social psychologists have betrayed their code of ethics. But claims about the power of unconscious ‘nudges’ are misguided because social priming effects remain little more than a scientific fiction. Thinking fast may have made for a snappy title, but arguably it is not thinking at all. Can we put to bed once and for all the notion that there is a mysteriously effective battery of psychological techniques, uniquely available to a new breed of psychological rulers? The grisly truth is more mundane: public compliance was secured with that same mixture of terror and virtue that Robespierre’s Committee of Public Safety would have found familiar. Lies, misinformation, intimidation and (in extremis) physical violence have served tyrants for as long as history – they have no more legitimate relationship with the discipline of psychology than physical torture has with the practice of surgery. No special justification for evil acts is purchased by defining them as ‘psychological’ with or without an appeal to elaborate ethical codes and the alleged mysteries of unconscious persuasion.
I find it impossible to believe my professional and academic colleagues (including Sage advisers) were unaware of a body of highly relevant experimental psychological research which could have improved the quality of advice (and almost certainly altered the actions taken). Someone surely must ask where the vast majority of professional experimental psychologists have been hiding for the last two years? Frankly, their silence begins to look like complicity.
.
FOR two years our daily lives in the face of Covid have been ruled by a coterie of experts in ‘behavioural psychology’, a breed of psychologist that seems to have appeared from nowhere. I speak as a psychologist who has been in the business for a good many years. It is worth asking, who are these people, and how did they come by these insights into the workings of the human psyche? Insights so profound that they merited our unthinking obedience?
Psychologists in the 1950s were given to claiming expertise in behaviour, in the rather circular sense that they would study any organism that ‘behaved’. Pavlov’s work on conditioned salivation in dogs was the paradigm case until B F Skinner extended this to include what he called ‘operant conditioning’, showing how pigeons could be induced to behave in quite extraordinary ways by scheduling their rewards. For a few years this kind of raw behaviourism had a powerful influence in psychology. It fell to a bright young linguist named Noam Chomsky to put paid to it with a demonstration that behaviourism could not account for the complexities of human language and thought. But this leaves my question unanswered: if the behavioural psychologists currently bossing us about are not behaviourists in Skinner’s sense, what on earth are they and where did they come from?
A useful orienting date for an answer is 2011 – the year the House of Lords Science and Technology Committee produced a report on ‘Behaviour Change’. A succession of experts from the world of social psychology had induced their lordships to swallow Nudge Theory hook, line and sinker (the infamous MINDSPACE protocol dates from 2010). Their report is still worth reading, if only to see how groupthink works. The committee accepted as hard fact a host of assertions which were, even then, patently questionable; for example, the proposition that our thought processes (and behaviour) are largely under the control of factors of which we are unconscious. We may be rational if we want to be, but it’s the environment that unconsciously nudges us, like prodded sheep, one way or another.
In a nutshell the noble lords seemed content to accept the proposition that systems of ‘unconscious thought’ determine how we act. Bizarrely, they even allowed themselves to be scolded for wrongthink by countenancing the idea that human beings might act as rational agents. Commit to that heresy, they were told, and you fall into the ‘fundamental misattribution error’ (which I accept sounds really bad). I can only conclude that their lordships thought all this stuff applied only to other people.
The report was caught up in (and possibly explained by) the intellectual fever surrounding the publication of Daniel Kahneman’s best-seller, Thinking, Fast and Slow. This was the book (a kind of Freud without tears) which allowed the notion of two systems of thought to enter popular culture. It was beyond doubt, Kahneman urged, that behaviour was readily altered without our being conscious of the determining factors. And indeed, the data at that time in support of ‘social priming’ were remarkable:
- Given something warm to drink before an experiment affects the warmth you will show towards other people afterwards
- Simply encountering words related to age makes you walk more slowly afterwards
- Having a pile of play money visible in a room influences how hard students will work to solve simple puzzles
- Reminding people of the concept of money makes them more likely to endorse free-market capitalism in a questionnaire administered later
- Having something with a fishy smell in the vicinity makes you suspicious
The range of these effects seemed constrained only by the imagination of the social psychologists carrying out the work. Countless articles appeared, many in respected journals, all broadly supportive. Nonetheless, if by now you are finding this whole bag of tricks a little fishy, you would be right. The brutal truth took far too long to emerge, but virtually none of the innumerable studies on which the elaborate edifice of social priming rested could be replicated. That’s none, as in zero.
The strange death of social priming was only part of a much wider replication crisis which engulfed the discipline of social psychology at that time, and it soon became evident that a replication problem could all too easily morph into something much worse. Even as the Lords’ report was being prepared, a misconduct hearing was set up in the Netherlands to examine the work of the Dutch social psychologist, Diederik Stapel, revealing an ‘astonishing lack of familiarity with elementary statistics’. Dr Stapel retracted dozens of published journal articles and admitted falsifying data. But that was not really the point: how could his many co-authors have failed to notice results so obviously dodgy? Why did successive generations of university doctoral students accept these plausible fictions at face value? Not to forget referees of Stapel’s publications: ‘It is almost inconceivable,’ the inquiry concluded, ‘that reviewers of the international “leading journals”, who are deemed to be experts in their field . . . did not notice the reporting of impossible statistical results.’
Kahneman, who received the Nobel Prize for his contribution to behavioural economics, wrote an open letter to social psychologists working on priming, warning them that ‘a train wreck’ was approaching. The statistical sleight of hand which had allowed investigators for years to fish for ‘significant’ results until they found them had precipitated a catastrophe. Significantly, he added: ‘The problem could well be more severe in your field than in other branches of experimental psychology, because every priming study involves the invention of a new experimental situation . . . I believe that you should collectively do something about this mess.’
It is reasonable to ask whether something was, in fact, done about the mess. And the answer is very little. Behavioural psychologists in some sceptical quarters have become hate figures but, crucially, any such criticism invariably runs along with acceptance that ‘nudging’ as a methodology is rooted in sound psychological principles. As I hope I have made clear, the evidence shows that this is far from the case. It follows that in setting up a Sage subcommittee drawing (by accident or design) exclusively on the expertise of ‘behavioural scientists, health and social psychologists, anthropologists and historians’ the UK government was guilty of serious misjudgment. Was it at all defensible to seek advice from a discipline still reeling from an insoluble replication crisis?
In light of Kahneman’s remarks it was equally perverse to ignore expertise from areas of psychological research uncontaminated by issues of replication failure and fraud. To take but one example (of hundreds), some of the most advanced experimental work in developmental psychology goes on in UK universities. Yet the technical advice of developmental psychologists on the consequences of lockdown measures on schoolchildren seems not to have been sought – or if it was sought, it was ignored. Were psychologist specialists consulted on the behavioural consequences of ‘masking’ on children (and their teachers) for long periods? In place of credible scientific advice, absurd nostrums such as ‘I wear my mask to protect my mate’ are simply embarrassing (apart from being false).
It will be argued, of course, that the proof of the pudding was in the eating and that extraordinary levels of public compliance during this pandemic reflect the power of the psychological techniques that brought it about. This seems to be at least part of the motivation for Dr Gary Sidley’s complaint to the British Psychological Society that social psychologists have betrayed their code of ethics. But claims about the power of unconscious ‘nudges’ are misguided because social priming effects remain little more than a scientific fiction. Thinking fast may have made for a snappy title, but arguably it is not thinking at all. Can we put to bed once and for all the notion that there is a mysteriously effective battery of psychological techniques, uniquely available to a new breed of psychological rulers? The grisly truth is more mundane: public compliance was secured with that same mixture of terror and virtue that Robespierre’s Committee of Public Safety would have found familiar. Lies, misinformation, intimidation and (in extremis) physical violence have served tyrants for as long as history – they have no more legitimate relationship with the discipline of psychology than physical torture has with the practice of surgery. No special justification for evil acts is purchased by defining them as ‘psychological’ with or without an appeal to elaborate ethical codes and the alleged mysteries of unconscious persuasion.
I find it impossible to believe my professional and academic colleagues (including Sage advisers) were unaware of a body of highly relevant experimental psychological research which could have improved the quality of advice (and almost certainly altered the actions taken). Someone surely must ask where the vast majority of professional experimental psychologists have been hiding for the last two years? Frankly, their silence begins to look like complicity.
.