Intended for healthcare professionals


Changing behaviour through state intervention

BMJ 2008; 337 doi: (Published 15 December 2008) Cite this as: BMJ 2008;337:a2543
  1. Theresa M Marteau, professor of health psychology1,
  2. Adam Oliver, RCUK academic fellow in health economics and policy2,
  3. Richard E Ashcroft, professor of bioethics3
  1. 1Department of Psychology, King’s College London, Guy’s Campus, London SE1 9RT
  2. 2LSE Health, London School of Economics and Political Science, London WC2A 2AE
  3. 3School of Law, Queen Mary, University of London, London E1 4NS
  1. theresa.marteau{at}

    When does an acceptable nudge become an unacceptable shove?

    The question of when governments should intervene to change our behaviour sits at the heart of political philosophy and bioethics.1 In their recent book, Nudge: improving decisions about health, wealth, and happiness, Richard Thaler and Cass Sunstein argue that public and private organisations can and should intervene more often than they currently do to alter people’s behaviour.2 Their thesis is based on the observation that people often behave in ways that are not in their best long term interests. The remedy, they argue, is to present options in such a way as to increase the likelihood that people will choose what they would, on reflection, most prefer. This approach has attracted much interest and support from politicians including Barack Obama and David Cameron.

    Thaler and Sunstein call their general approach “libertarian paternalism”—paternalistic in that governments, for example, play an active role in framing options, but libertarian in the sense that people ultimately remain free to choose. Examples of such approaches in the health context include placing fresh fruit and vegetables at the front of canteen displays,2 requiring people to purchase permits if they wish to buy cigarettes,3 and organising organ donation under an opt-out rather than an opt-in system.4 Related schemes include using financial and other incentives to encourage smokers to quit and to increase the likelihood that patients with schizophrenia will take long acting medication.5 6 These approaches can have a great impact on people’s behaviour.2

    The extent to which such interventions may be adopted depends not only on their effectiveness but also their acceptability. Although some people see such interventions as acceptable “nudges,” others consider them unacceptable “shoves.” The debate then revolves around whether the paternalistic or libertarian aspect of libertarian paternalism dominates. Its resolution turns in part on how we understand people’s behaviour, and whether it is possible or worthwhile to distinguish conceptually and empirically between these two forms of pushing.

    People’s behaviour is a function of their values and their environment. Even when people are not forced to act in particular ways they often choose options that, on reflection, they wish they had not. For example, most smokers would prefer not to smoke, and most overweight people want to lose weight. How do we understand this apparent irrationality? Two behavioural principles seem to be operating. Firstly, the environment has a far larger effect on our behaviour than we acknowledge. This lack of acknowledgment reflects a behavioural principle that has been termed the fundamental attribution bias,7 in which we perceive our behaviour as determined more strongly by our dispositions than by situations, a bias that is seen across cultures, patients, healthcare professionals, and policy makers.

    Behaviours that are immediately rewarding, such as eating chocolate cake, are more likely to occur than those that bring delayed rewards, like running across a park. The rewarding nature of immediate gratification reflects the second behavioural principle that helps explain why people often act in ways that do not reflect their long term goals or interests. Unfortunately, the health costs of many behaviours occur in the future, whereas their benefits are enjoyed in the present.3 The so called obesogenic environment reflects the operation of these two principles—that is, an environment that contains a plethora of readily realised options that offer immediate gratification but that cause and sustain obesity.8

    Any attempt to disentangle nudging from shoving requires operational definitions. Nudging could be defined as increasing the chances that people act in ways that, on reflection, they would have chosen themselves; this is variously described as acting on preferred preferences9 or acting consistently with deeply held values.10 Shoving can be defined as increasing the chances that people behave in ways preferred by the “shover” but not the “shoved.”

    Evaluating interventions against these definitions requires the development of measures of values, immediate and considered preferences, as well as coercion, to be used in the context of robust experiments evaluating the effect of interventions designed to nudge. So, for example, in the context of food labelling, different sets of nutritional labels might be compared for their effect on food purchased and, more importantly, the extent to which it matches the food people say that they would prefer to have bought after careful deliberation.

    Finally, it is worth considering in a health context whether distinguishing between nudges and shoves is useful. Thaler and Sunstein show that our choices are invariably structured, but that we are often unaware of this, and pay scant regard to the fact that the choices we make are frequently inconsistent with our “preferred preferences.”9 This is perhaps the key message. Policy makers and public health practitioners in particular should perhaps be less concerned with discussing liberty and paternalism and be more concerned with trying to structure environments to achieve choices that, on reflection, are endorsed by the chooser. At the heart of this is a paradox—acting paternalistically to achieve liberty. Failure to engage with this means that we may remain slaves to the environments we often have little part in shaping.

    Perhaps the best nudge provided by Thaler and Sunstein’s book is towards a debate about the political philosophy that should guide governments wanting to improve the health of their populations.


    Cite this as: BMJ 2008;337:a2543


    • Competing interests: None declared.

    • Provenance and peer review: Not commissioned; externally peer reviewed.


    View Abstract

    Log in

    Log in through your institution


    * For online subscription