Philosophical Roots of the Concept of Nudges

WORKING PAPER: DO NOT QUOTE Anne Brunon-‐Ernst, ECPR 20-‐24 Aug 2015, Montreal University (CANADA)

(French Nudge Project, funded by Ecole de droit, Sciences Po, Paris)

1

Philosophical Roots of the Concept of Nudges1

Anne BRUNON-ERNST

Institut Villey (Panthéon-Assas University) and Centre Bentham (Ecole de droit, IEP, Paris)

Introduction

As citizens,2 consumers and individuals, nudges3 are becoming an increasing part of our

daily lives. Nudges are however only one set of tools within a wider array of

paternalistic approaches to individuals, politics and economics. Indeed for a fuller

account of this new way of controlling individuals, other competing approaches will

have to be discussed such as choice architecture, soft paternalism, hard paternalism etc.

The crux of the matter is that these concepts have problematic implications in respect to

individual freedom, autonomy, and, by the same token, of potential choice manipulation

that could arise from them.4 There is therefore a vital need to understand both the

workings of these tools based on human psychology and to determine whether they are

legitimate interference to individual choices. All around the world, academics, but also

politicians and journalists join in to try work out an assessment of nudges.5 The present

talk wishes to contribute to the debate in presenting an intellectual history perspective

1 On 18th May 2015, at the University Panthéon-‐Assas, a draft of this paper was presented at the French Nudge Project Research Seminar as “Bentham-‐land or Mill-‐land : Nudges in Perspective”, a NUDGE-‐Extra-‐legal normativities and public policies project, funded by the Ecole de Droit, Sciences Po, Paris (FRANCE). 2 See for e.g.: Cass Sunstein’s position as the Administrator of the White House’s Office of Information and Regulatory Affairs in the Obama Administration, Richard Thaler’s involvement in the British Conservative Party in 2009 or the existence of a Behavioural Insight Team in the UK Cabinet. 3 The term was coined by Sunstein and Thaler, as repeatedly used in their publications, among which: Sunstein, Cass R. and Thaler, Richard H., ‘Libertarian Paternalism Is Not An Oxymoron’, University of Chicago Law Review, 70-‐4 (2003) : 1159f; Sunstein, C. and Thaler, R., Nudge: Improving Decisions About Health, Wealth and Happiness, New Haven, Yale University Press, 2008; Sunstein, Cass, ‘Nudges and Public Policy’, Paper at the International Behavioural Insights Conference, Sydney, 2-‐3 June 2014; Sunstein, C. Why nudge? The Politics of Libertarian Paternalism, New Haven, Yale University Press, 2014. 4 Rebonato, R., Taking Liberties: A Critical Examination of Libertarian Paternalism, New-‐York, Palgrave Macmillan, 2012; White, M. D., The Manipulation of Choice: Ethics and Libertarian Paternalism, New-‐York, Palgrave Macmillan, 2013 5 See for e.g.: numerous articles published in The Guardian (Mona Chalabi, ‘Does a government nudge make us budge?’, published on November 12, 2013, in The Guardian, Online edition, at http://www.theguardian.com, accessed 19/5/2015) or in The New York Times (Catherine Bennhold, ‘The Minsitry of Nudges’, published on December 7, 2013, in New York Times, Online edition, at http://www.nytimes.com, accessed 19/5/2015). But also in non-‐English media (Etienne Gless, ‘Le nudge marketing ou comment vendre sans contraindre’, published on July 16, 2013, in L’Express, Online edition, at http://lentreprise.lexpress.fr, accessed 19/5/2015).



2

(as opposed to psychology, economics or behavioural economics as are often used in

discussions on nudges) on these new regulatory instruments.

It is not the purpose of the present talk to look at the influence of philosophers in the

shaping of nudges, as indeed the notion of influence is difficult to trace; and as nudges

owe certainly more to the development of psychology and to the application of

behavioural insights to economics than to philosophy. This is not to say that heritage of

philosophers in nudges is not generally acknowledged both by academics and in the

media, as in the case of JS Mill.6 However, I contend that, in most of these cases,

philosophers are used with strategic rather than exegetic purpose.7 There is an urgent

need to put the theoretical underpinnings and the pragmatic implementation of nudges

into perspective and draw the lessons from such an experiment.

I will not look into JS Mill’s heritage, but into that of the British philosopher and legal

reformer Jeremy Bentham (1748-‐1832). Indeed Bentham’s concept of indirect

legislation is particularly relevant to our discussion here. At the beginning of the

industrial revolution, at a time when changing work relationships required a new

organisation of society, Bentham was at the forefront of the reflexion on new forms of

regulation. He will be remembered for his Panopticon schemes (which started out as a

Russian factory under the supervision of his brother Samuel, before being the well-‐

known panoptic prison) but also his Constitutional Code. The interest of academics in

Bentham’s investigation of alternative forms of regulation has first arisen from the

Bentham-‐Foucault debate on biopolitics,8 which highlighted the pivotal importance of

Bentham’s legal and non-‐legal forms of regulations encapsulated in the concept of

6 See for e.g.: David Brown, ‘Review : ‘Why Nudge?’, by Cass R Sunstein’, published on May 4, 2014, in the Financial Times, Online edition, at http://www.ft.com, accessed 19/5/2015, Cass R. Sunstein, ‘It’s For Your Own Good: Against Autonomy: Justifying Coercive Paternalism by Sarah Conly’ published on April 25, 2013, in the New York Review of Books, Online edition, at http://www.nybooks.com, accessed 19/5/2015, and Mona Chalabi, ‘Does a government nudge make us budge?’. 7 See paper by the present author on ‘JS Mill and Nudges’ in the Journal of European History of Ideas [under submission] 8 See S.G. Engelmann, ‘”Indirect Legislation”: Bentham’s Liberal Government’, in Polity, 35: 3 (2003): pp. 372f, C. Laval, ‘La chaîne invisible’, Revue d’études benthamiennes [e-‐journal], 1 (2006), posted on 1st September 2006, accessed on 15th January 2012. URL : http://etudes-‐benthamiennes.revues.org/63; and Brunon-‐Ernst, (ed.), Beyond Foucault. New Perspectives on Bentham’s Panopticon, Aldershot: Ashgate, 2012 and Utilitarian Biopolitics: Bentham, Foucault and Modern Power, London: Chatto and Pickering, 2012.



3

indirect legislation.9 Following the work of Michael Quinn, time has come to offer a more

balanced view of the mechanisms offered by indirect legislation. A more subtle picture is

now emerging of a regulatory system which offers new and unmapped forms of

regulation which are both legal and non-‐legal.

First, I suggest we look into the meaning of indirect legislation, trying to draw the line

between this form of regulation and its converse, direct legislation (Part II). Second, I

highlight the similarities between nudges and some types of indirect legislation (Part

III). Lastly, I see how the lessons Bentham draws from his experiment with indirect

instruments could contribute to the nudge debate (Part IV).10 The aim of my intellectual

history perspective on Bentham’s theory of social regulation allows a reappraisal of

nudges’ impact on individual freedom.

Nudges, choice architecture, soft paternalism, hard paternalism etc. as these normative

tools will be discussed all along the present talk, it might be better to settle early the

meaning of the terms, which will be discussed in the arguments which follow. There are

no standard definitions, as most writers have made up a working definition that fit the

purpose of their arguments.11 Nonetheless, for a clearer understanding of the issues

raised later in this talk, imperfect definitions might be better than no definition at all.

Part I thus opens up with a definition section.

1. Definitions and Comparisons

The first definition is that of nudges. Cass R. Sunstein & Richard Thaler’s coined the

term. Sunstein defines nudges as ‘initiatives that maintain freedom of choice while also

9 A. Brunon-‐Ernst, “Le gouvernement des normes : Jeremy Bentham et les instruments de régulation post-‐modernes”, in Archives de philosophie, 78 : 2 (2015), pp. 309-‐322 and Brunon-‐Ernst, A. and Van Waeyenberge, A., “Effects of the open method of coordination (OMC) in research and innovation: legislation in EU policy-‐making? ”, in Journal of Legal Pluralism and Unofficial Law, 2015, pp. 1-‐17; and M. Bozzo-‐Rey, ‘Back to Bentham: Indirect Legislation and Nudge’, Conference at the Bentham Seminar, UCL, London, 18 March 2015. 10 This method was also used by the same author in previous work on postmodern forms of regulation. Working from the perspective of intellectual history and Bentham’s work on postmodern normativities has allowed to highlight some controlling instruments common to different initiatives. See Brunon-‐Ernst, “Le gouvernement des norms” and Brunon-‐Ernst and Van Waeyenberge, ““Effects of the open method of coordination (OMC) in research and innovation”. 11 See Sunstein’s criticisms and changes to Rebonato’s definition of nudges in Rebonato, Taking Liberties (Sunstein, Why nudge?, p. 58-‐59).



4

steering people’s decisions in the right direction’.12 There are several important points

to make about this definition. First, nudges aim to alter people’s behaviour. It is

conceived as a tool to be used by people wishing to modify other people’s behaviour

(governments, businesses,13 parents etc). Nudges have the advantage of a scientific

method based on psychology, which makes the response of individuals to certain

initiatives predictable. In that respect, it appears as an ideal tool for governments and

businesses alike. The second point is that individuals nudged are not forced to modify

their behaviour. Even if the preferred option is more likely, they still have the choice to

select another option, which is not preferred by the nudger. A characteristic feature of

nudges is that they make their reversal possible, that is the nudgee can override the

nudger’s or the choice architect’s preference.14 As A-‐L Sibony and F. Alemanno note:

‘Nudge is therefore presented as a distinctive, alternative way, characterized as being

minimally burdensome, low-‐cost and choice-‐preserving, to help promote regulatory

goals.’15 The third point is that nudges are initiatives operatig in the interest of the

nudgee. You will have noted the use of the word ‘right’ in Sunstein’s definition. It is the

nudger who determines the right goals for nudgee. The goals are therefore only right in

the eye of the nudger, who can have a very different agenda from the nudgee. In this

respect, a nudger is a paternalist.

S. Conly defines paternalism as: ‘a practice wherein people are forced to perform actions

that bring about good consequences for themselves’.16 It is the beneficent aim of the

initiative devised by the paternalist that determines the paternalistic nature of the

intervention on choice. Nudges are paternalist as they are initiatives devised with an

aim. However not all paternalist initiatives aim at benefitting the individual. Some aim at

benefitting society or the financial interests of a company. Indeed when it comes to

paternalist initiatives operating in a business-‐related environment, the presumption of

12 Sunstein, Why nudge?, p. 17. 13 Alberto Alemanno and Anne Lise Sibony (eds.), Nudging and the Law. What Can EU Law Learn from Behavioural Sciences?, Oxford: Hart Publishing, 2015, Forthcoming. Alemanno and Sibony distinguish the purpose government initiatives from business-‐related ones. See Alemanno and Sibony (eds.), p. 18-‐19. 14 Nudges are referred to as the ‘mildest and most choice-‐preserving forms of intervention’, in Sunstein, Why nudge?, p. 17. 15 Alemanno and Sibony (eds.), pp. 5. 16 Conly, Sarah, Against Autonomy : Justifying Coercive Paternalism, Cambridge University Press: New York, 2013, p. 48.



5

benevolence is rebutted, as the nudge is likely enhance business profits at the expense of

effective consumer freedom of choice.17

There are two types of paternalism. Means/ends paternalism18 and soft/hard

paternalism.19 It is sometimes difficult to draw the line between soft and hard

paternalism. I believe the main distinction lies in the degree of paternalist’s

intervention: if the nudgee is given less opportunities to reverse the nudge then the

degree of nudging is more intense, thus qualifying the initiative for a hard paternalist

status rather than soft.

What is then the difference between nudges and choice architecture, another term used

in my talk so far? Choice architecture refers to the background against which choices are

made.20 Any environment which is organised by man and where choice is required

involves choice architecture. Choice architects can voluntarily (or involuntarily)

organise an environment to make one choice more likely. Nudgers are one type of choice

architects: those who voluntarily organise an environment to maximise choice according

to certain goals (political, health-‐related, economic, etc.), while minimizing cost.

In the terminological discussion above, a picture of nudges is emerging as an array of

utility-‐driven, cost-‐effective, beneficent initiatives based on scientific assessment of

motives individuals have to act. This description draws a likeness of nudges, which is

very similar to the utilitarian project. Bentham was the first (and the last) British

philosopher to endeavour to create a system of government based on the science of

motives. Bentham explains the founding principles of his system at the start of the

Introduction to the Principles of Morals and Legislation:

Nature has placed mankind under the governance of two sovereign masters, pain

and pleasure. It is for them alone to point out what we ought to do, as well as to

determine what we shall do. On the one hand the standard of right and wrong, on

17 On this point, see n. 13. 18 Sunstein, Why nudge?, pp. 61-‐71. Means paternalism refers to initiatives which provide steps in order to make the nudgee achieve the end promoted by the paternalist. 19 Ends paternalism which clearly steers the nudgee towards the preselected choice. S. Conly makes roughly the same distinction with the two concepts she uses: libertarian paternalism and coercive paternalism. See Conly, Against Autonomy, chapter 1. 20 Sunstein, Why nudge?, p. 14.



6

the other the chain of causes and effects, are fastened to their throne. They govern

us in all we do, in all we say, in all we think: every effort we can make to throw off

our subjection will serve but to demonstrate and confirm it. In words a man may

pretend to abjure their empire: but in reality he will remain subject to it all the

while. The principle of utility recognises this subjection, and assumes it for the

foundation of that system, the object of which is to rear the fabric of felicity by the

hands of reason and of law. Systems which attempt to question it, deal in sounds

instead of sense, in caprice instead of reason, in darkness instead of light.21

The basis for Bentham’s statement is scientific observation. The observation Bentham

makes helps him to isolate a psychological feeling shared by all sentient beings.22 They

are at the source of our actions: human beings avoid pain and seek pleasure. When

individuals decide to act or to refrain from acting, they do so on account of the pleasure

they will derive or the pain they will avoid. Bentham’s utilitarian system is rooted in a

scientific analysis of the psychological springs of human action. Bentham’s philosophical

endeavour goes beyond this initial psychological assessment, as he derives from this

scientific observation a moral standard. Pleasures and pains are not only what motivate

us to act, they are also the standard of right and wrong. In this quote, Bentham not only

gives the regulator tools to frame initiatives which alter behaviour, but also the goal

towards which behaviour should be steered: pleasure rather than pain. Bentham’s

purpose is even clearer in the following quote:

The principle of utility is the foundation of the present work […]. By the principle

of utility is meant that principle which approves or disapproves of every action

whatsoever according to the tendency it appears to have to augment or diminish

the happiness of the party whose interest is in question: or, what is the same

thing in other words to promote or to oppose that happiness. I say of every action

whatsoever, and therefore not only of every action of a private individual, but of

every measure of government.23

21 J. Bentham, Introduction to the Principles of Morals and Legislation, in Burns, J. H., Dinwiddy, J. R. et Rosen, F (eds.), The Collected Works of Jeremy Bentham, Oxford : Clarendon Press, 1996, [1789], p. 11 22 R. Harrison, Bentham, London: Routledge & Kegan Paul, 1983, p. 177. Note that in Introduction to the Principles of Morals and Legislation, Bentham did not restrict the feeling of pleasure and pain to human beings only; he clearly said that animals could experience pleasure and pain. 23 Bentham, Introduction, p. 11.



7

There seems to be enough similarities between Bentham’s utilitarianism and nudges to

warrant the more thorough comparison we are now presenting.

Now that the linguistic pitfalls of any study of nudges have been cleared, the present talk

can move on to its core argument on the philosophical underpinnings of the nudge

debate.

2. Direct and Indirect Legislation

a. History of the Text and Bentham’s Terminology

Bentham is remembered as a legal reformer, who stands in the shadows of one of his

most famous disciple: JS Mill. One of the reasons of this neglect is that the sources of

Bentham’s work are not easily accessible (although thanks to the work of the Bentham

Project, this is becoming less so).24 Writings on indirect legislation are to be found in

many different manuscripts, written in two languages (French and English) and

(re)published by two different editors. Bentham starts writing on indirect legislation in

1782, in a document named Plan. He does so as part of his reflexion on what would have

been chapter 1825 of the Introduction to the Principles of Morals and Legislation referred

to in the quote above.26 This work is then continued in French, as part of a series of

manuscripts entitled Projet. Etienne Dumont, his Genevan editor, then takes up some of

the material on indirect legislation to include it with the Traités de legislation civile et

pénale, which is first published in French in 1802.27 The Traités are then translated into

French and included in the edition of the Works of Jeremy Bentham in 1838-‐43, compiled

24 When Bentham died, a first edition of his works was compiled by his secretary John Bowring. Unfortunately the edition was far from scientific, aggregating parts of manuscripts with others, publishing versions by unapproved editors, publishing English translations from French editions, omitting the most unpalatable writings on religion and sex. A second edition is now underway at the Bentham Project (UCL, London), publishing volumes directly from the manuscript sources. Since 1968, thirty volumes (of an anticipated 70 volumes) of the new collected works have been published. 30 volumes in 48 years. There is another 50 years of estimated work to be carried out before the final complete edition of Bentham’s writings be made available to the wider public. Until the scientific edition is completed, the only accurate source of Bentham’s works is to be found in the manuscripts in the UCL or the British Library fund. 25 Bentham’s Introduction contains sixteen chapters. Chapter 17 was extended into Of Laws in General, now published as Limits of the Penal Branch of Jurisprudence, and deals with the differences between civil and criminal law, basing itself on the analysis of what ‘a law’ is. Chapter 18 was meant to be on means to prevent crimes, that is the ‘preventive’ branch of legislation, otherwise known as ‘indirect legislation’. 26 Bentham, Introduction, p. 11. 27 J. Bentham, Traités de legislation civile et pénale, pref. M. Bozzo-‐Rey, A. Brunon-‐Ernst, E. de Champs, Paris, Dalloz, 2010.



8

by John Bowring, Bentham’s last secretary and executor.28 The present talk investigates

the manuscript and published sources available on indirect legislation. However, the

present speaker believes that the corpus of indirect legislation should be extended to

include Bentham’s writings of poor law,29 the Police Revenue Bill,30 political economy31

and civil law32 in an attempt to bring a fuller picture of the various applications of

indirect legislation.

A second reason for our philosophers and intellectual historians to have overlooked

Bentham’s work and heritage is that Bentham creates a new self-‐sufficient system,

which uses it own terminology to describe its objects. This is all the more true of the

concept of indirect legislation, which needs to be defined. To understand it, the

converse, direct legislation also needs to be explained. In a nutshell, if direct legislation

corresponds to laws (although as we shall see later, Bentham’s understanding of a law is

more encompassing than its present acceptation), indirect legislation cannot be merely

equated with non-‐legal regulatory instruments. The legal/non-‐legal divide needs to be

explored.

28 /ADD REFERENCE/J. Bentham, Works of Jeremy Bentham, ed. J. Bowring, 1838-‐1843, 11 Vols. 29 J. Bentham, Writings on the Poor Laws, in ed. M. Quinn, The Collected Works of Jeremy Bentham, 2 vols (Oxford: Oxford University Press, 2001), vol. 1; J. Bentham, Writings on the Poor Laws, in. ed. M. Quinn, Collected Works, 2 vols (Oxford: Oxford University Press, 2010), vol. 2. 30 /ADD REFERENCE/Bentham, Works, p. XXX. 31 The available printed editions on Bentham’s economic writings are to be found in J. Bentham, Jeremy Bentham’s Economic Writings, ed. W. Stark, 3 vols (London: George Allen and Unwin, 1954) and in the Bowring edition; however, M. Quinn is editing a new scientific edition of Bentham’s economic writings. The reference to all the Bentham economics manuscript sources were given by Dr Michael Quinn (Bentham Project, UCL), currently editing the three volumes of the Economics Writings in the Collected Works. 32 Bentham’s writings on civil law have not all been edited. Those who are still in manuscript form are to be found in UCL boxes 4, 29, 30, 31, 32, 33, 37, 97, 98, 99, 100, 107a, 107b, 146 et 160. Bentham’s thought on civil law can be found in the following books: J. Bentham, Manual of Political Economy, ed. John Bowring, The Works of Jeremy Bentham, Edinburgh, 1843, vol. 3, pp. 31-‐84 ; J. Bentham, Principles of the Civil Code, ed. John Bowring, The Works of Jeremy Bentham, Edinburgh, 1843, vol. 1, pp. 297-‐364 ; J. Bentham, A General View of a Complete Code of Laws, ed. John Bowring, The Works of Jeremy Bentham, Edinburgh, 1843, vol. 3, pp. 155-‐210 ; J. Bentham, Pannomial Fragments, ed. John Bowring, The Works of Jeremy Bentham, Edinburgh, 1843, vol. 3, pp. 211-‐230 ; J. Bentham, Nomography, ed. John Bowring, The Works of Jeremy Bentham, Edinburgh, 1843, vol. 3, pp. 231-‐283 ; J. Bentham, Logical Arrangements, ed. John Bowring, The Works of Jeremy Bentham, Edinburgh, 1843, vol. 3, pp. 284-‐295 ; J. Bentham, Introduction ; J. Bentham, Of the Limits of the Penal Branch of Jurisprudence, in P. Schofield (ed.), The Collected Works of Jeremy Bentham (Oxford: Oxford University Press, 2010) ; J. Bentham, Constitutional Code, Vol. 1, ed. Fred Rosen, J.H. Burns, Oxford, Clarendon, 1983 ; J. Bentham, Legislator of the World, eds Philip Schofield and Joathan Harris, The Complete Works of Jeremy Bentham, Oxford, Clarendon, 1998 ; J. Bentham, Théorie des Fictions, trad. G. Michaut, Paris, Editions de l’Association Freudienne Internationale, 1996.



9

Bentham’s whole intellectual endeavour can be aptly summarized as the attempt to

control individual action. There are two ways to control individual or collective

behaviour: first, direct legislation, otherwise known as enacted legislation or law,

defined as the expression of the will of the lawgiver addressed to those who are

subjected to his/her will, with the aim of controlling their actions and backed by

sanctions33 in case of non-‐compliance;34 and second, indirect legislation, which is are the

acts taken by the lawgiver (or any other agent) to influence the conduct of others, either

in the place of legal punishment (non-‐legal forms of regulation) or as an accessory to it

(legal form of regulation).35

Control in Bentham thus operates through legal and non-‐legal means. However the

distinction is not necessarily relevant in Bentham’s plan as it is clear that all laws

contribute to the formation of an architecture of choice, that is to the shaping of an

environment in which individuals will make choices.36 There is a tendency to read

nudges as regulatory mechanisms used to replace legal instruments. Although there is

evidence of nudging in the law,37 the main difference remains one of degree. Nudgers or

soft-‐paternalists are interested in default options which are easy to overcome, and this

focus seems only compatible with non-‐legal means of regulation rather than legislation,

which is the preferred option of hard-‐paternalists. However for Bentham even direct

legislation mandates choices, which need not be complied with, as the would-‐be

offender might choose the pain of punishment over the pleasure of deriving a benefit 33 See J. Bentham, Of the Limits of the Penal Branch of Jurisprudence, in P. Schofield (ed.), The Collected Works of Jeremy Bentham, Oxford: Oxford University Press, 2010. However some commentators have shied away from this interpretation, finding in Bentham signs of a disposition to obey (see Constitutional limits and the Public Sphere: A Critical Study of Bentham’s Constitutionalism, Oxford, Hart Publishing, 2000, 49) to account for the complex reasons why individuals abide by a rule (HLA Hart, Essays on Bentham: Studies on Jurisprudence and Political Theory, Oxford, Clarendon Press, 1982, 252). Their arguments are convincing and complement a punishment-‐orientated law in Bentham (Brunon-‐Ernst, Utilitarian Biopolitics, 67-‐85), but cannot replace overwhelming evidence of the primacy of punishment. 34 Bentham, Traités, 24 and M. Bozzo-‐Rey, ‘Le statut du principe d’utilité dans la philosophie politique de Jeremy Bentham’, PhD dissertation, Paris Ouest-‐Nanterre-‐La Défense University, 2007. 35 S. Engelmann, ‘”Indirect Legislation”: Bentham’s Liberal Government’, in Polity, 35: 3 (2003): pp. 372f; M. Bozzo-‐Rey, ‘Le statut du principe d’utilité dans la philosophie politique de Jeremy Bentham’; Brunon-‐Ernst, Utilitarian Biopolitics. For instance, enacting legislation to criminalize the purchase of services offered by prostitutes is a clear example of direct legislation as it has the intended effect of prohibiting certain categories of individuals from behaving in certain ways by the threat of a legal sanction. However, It is also to be viewed as indirect legislation as its primary goal is to prevent prostitution by prostitutes (See Bentham, Traités, pp. 375-‐377). 36 M. Quinn, “Jeremy Bentham, Choice-‐Architect: Law, Indirect Legislation and the Context of Choice", Nudge Project Research Seminar, Université Catholique de Lille, 19 juin 2015. 37 Sibony and Alemanno.



10

from his crime. Therefore the distinction between legal and non-‐legal means of control

as it is understood by nudgers does not overlap the one made by Bentham.

b. Indirectness of Indirect Legislation

A better grasp of indirect legislation (and its converse, direct legislation) is needed to

understand the thin line dividing legal and non legal indirect provisions. To date,

courtesy of Michael Quinn, only explanation of the indirectness of indirect legislation

was unearthed from the manuscripts:

7. Seventh Division. Direct legislation, indirect legislation. I call direct laws those

which go straight to their target: that is prohibiting or ordering the act itself in

question in criminal law; expressly deciding that a title to property or a position

will belong to this or that individual in civil or constitutional law. All other means

would be called indirect. The same law which is direct in one respect in relation

to an end is indirect in relation to another end. This division was perceived by

some writers but not determined; and nobody has ever drawn this division to its

ultimate conclusion. However, it can be considered as complete.38

Bentham points the indirectness of new branch of legislation as opposed to the

directness of direct legislation. The obliqueness of indirect legislation refers not so much

to the non-‐legal nature of the tools used by indirect legislation, but to the indirect means

used by the legal provision to achieve a certain aim. This definition will nonetheless be

qualified later, as we will see that indirect legislation can also encompass non-‐legal

forms of control which are akin to modern-‐day forms of regulation such as nudges.

Notwithstanding this explanation of indirect legislation, a common ground of all the

definitions to be found in Bentham’s writings are that they are negative definitions, as

indirect legislation is defined by its opposite: direct legislation. Bentham explains in the

Plan:

38 UC xxxiii. 111-‐12. I am grateful to M. Quinn for this quote. ‘7. Septieme Division. Législation directe, législation indirecte. J’appelle loix directes celles qui vont directement à leur but: savoir par exemple défendre ou ordonner l’acte même, dont on a affaire quand il s’agit du droit pénal, arrêter expressement qu’un droit de propriété ou d’office sera en telles ou telles mains quand il s’agit du droit civil ou du droit constitutionnel. Toute autre façon d’agir peut être appellée indirecte. La même loi qui est directe à un égard, par rapport à tel fin, est indirecte par rapport à un autre. Division entrevue par quelques auteurs, mais non determinée, & que personne ne s’est encore avisé de pousser à bout. Cependant elle peut être regardée comme complette’ [the speaker’s translation]



11

To begin with delinquency, the sovereign may be said to act against delinquency

in the way of direct legislation, in as far as he contents himself with prohibiting,

under such penalties as seem competent to each case considered simply by itself,

the several acts which are intrinsically obnoxious to him: that is, supposing the

principle of utility to be that by which he governs himself, the several acts from

which he conceives mischief to arise. To the head of indirect legislation may be

comprised whatever else can be done in the way of law in subservience to the

same end.39

Let us spend some time explaining the meaning and purpose of this statement. Bentham

refers to the field of criminal law mainly, but as mentionned in the first quote above,

indirect legislation can also apply to the fields of civil and constitutional law. Here the

sovereign is given two tools to fight crime: direct legislation, which prohibits acts and

punishes any transgression of the said prohibition (our criminal laws proper), and

indirect legislation. Indirect legislation seeks to achieve the same aim as direct

legislation, but using different means (rewards rather than punishments, for example)

or different objects (accessory laws to use Bentham’s phrase, that is adjective law). This

is, I believe, the way the phrase ‘whatever else can be done in the way of law in

subservience to the same end’ should be understood.

c. Indeterminacy of Indirect Legislation

In our main quote above, indirect legislation is referred to ‘whatever else’, which points

to the indeterminacy of the means used by such a method. This indeterminacy is

exemplified in the indirect ways in which Bentham tries to define a field, which cannot

be defined in an unmediated way. Metaphors help to draw parallels with other

mechanisms. Bentham compares indirect legislation with the fluids circulating inside a

plant, thus reeling out metaphors originating in nature (storm, sap etc.).40 Bentham also

uses geometry to explain the workings of indirect legislation, whereby direct legislation

is the straight line where indirect legislation is an oblique line.41 He also refers to

39 J. Bentham, Plan, see Bentham Project transcript of Plan [my emphasis] 40 ‘[I]n the [case of direct legislation] [the legislator] endeavours to carry his point as it were by storm: in the [case of indirect legislation], by sap’, in Bentham, Plan. 41 ‘In the [case of direct legislation], [the legislator] aims directly at his mark, he attacks the mischief directly and in front: in indirect legislation he attacks it by oblique and sometimes scarcely perceptible approaches’, in Bentham, Plan.



12

theatrical and hunting metaphors, with the idea of the hunter catching game and

operating behind the scenes, screened from human eyes.42

The indeterminacy of indirect legislation cannot be settled by a straightforward

definition. Would a description of initiatives falling under the scope of indirect

legislation help settle the meaning of this new category? Bentham’s lists of indirect

initiatives can be referred to. In the Traités, Bentham mentions controlling people

through pay,43 awarding honours to reward certain types of behaviour,44 regulating

data-‐saving,45 promoting public education rather than private,46 freeing the press,47

indemnifying the victims of crime48 etc. In the manuscript source named Plan, which

pre-‐dates the Traités, Bentham notes that the following indirect initiatives: procedure,

prevention of crimes, benevolence, poor law, prevention of drunkenness, promotion of

culture, advocacy of activity on the Sabbath, rules to strengthen the administration of

justice, new remedies for crimes of honour, information, publicity of public proceeding,

marking of individuals, registers, standards of quantity and quality, stamps, account of

people (census) etc. These two lists gather what seems to be a hodgepodge of unrelated

fields. Bentham was well aware of the fact as he writes that:

If of the expedients which we shall proceed to mention under the head of

indirect legislation there should be some which may appear referable to the

other head […].49

Such statements are far from helpful to circumscribe the meaning and scope of indirect

legislation. However Bentham believes that ‘the sphere of indirect legislation will be

found to be immense in detail but still circumscribed by certain bounds.’50

The issue then is to find the bounds that circumscribe indirect legislation. Indirect

legislation could not be defined; neither could it be limited by open-‐ended list of 42 ‘In Direct Legislation, the Legislator puts on his armour of artificial terrors, and stands forth confessed: In Indirect he nets every thing by imperceptible wires, keeping himself behind the curtain’, in Bentham, ‘Appendix I’, in Bentham, Plan. 43 Bentham, Traités, p. 356. 44 Bentham, Traités, p. 356. 45 Bentham, Traités, pp. 369-‐70. 46 Bentham, Traités, pp. 388-‐90. 47 Bentham, Traités, p. 392f. 48 Bentham, Traités, pp. 404-‐5. 49 BENTHAM, Plan, op. cit. 50 Bentham, Plan, see Bentham Project transcript of Plan [my emphasis]



13

unrelated acts. However, it seems that, with the help of traditional legal divisions

(criminal, civil and constitutional), Bentham manages at least to situate indirect

legislation within the architecture of laws. There is a consensus among Bentham

scholars that indirect legislation deals with criminal law.51 The definition of indirect

legislation given at the beginning of the present talk52 identified criminal law as the

privileged field of action of indirect legislation. The issue of whether civil law can be a

field of operation of indirect legislation is more contentious. Some Bentham scholars do

not see in Bentham’s works any sign of the operations of indirect legislation in the field

of civil law. They are correct in the sense that Bentham did not develop the theory of

indirect civil law to the same extent as indirect criminal law. This statement does not

imply that, in practice, Bentham did not apply indirect means of control to issues related

to civil law. In the example above, mention is made of marriage. Within the traditional

division of law, marriage remains a civil issue.53 Other academics support the

contention.54 Constitutional law is rarely mentioned as allowing indirect legislation.

However Bentham is clear that there can be indirect constitutional law.55

It seems that our modern day categories are inadequate to make sense of what is

indirect legislation. It overlaps both civil, criminal and constitutional law; as well as

51 See in particular the work of D. Lieberman/ADD REFERENCE/. 52 ‘To begin with delinquency, the sovereign may be said to act against delinquency in the way of direct legislation, in as far as he contents himself with prohibiting, under such penalties as seem competent to each case considered simply by itself, the several acts which are intrinsically obnoxious to him: that is, supposing the principle of utility to be that by which he governs himself, the several acts from which he conceives mischief to arise. To the head of indirect legislation may be comprised whatever else can be done in the way of law in subservience to the same end’ (Bentham, Plan) see note 15. [my emphasis] 53 The statement is not true if we consider Bentham’s division of laws: civil law is a dictionary of legal concepts, whereas criminal law contains the punishment attached to each offense. For a discussion on the division of law in several branches and a bibliography, see A. Brunon-‐Ernst ‘Délimiter la branche civile du droit: Bentham et la quête impossible’, in Bentham, juriste. Guillaume Tusseau and Malik Bozzo-‐Rey eds., Paris: Economica, 2010, pp. 177-‐200; and A. Brunon-‐Ernst, ‘Of the Limits of the Civil Branch of Jurisprudence : Background, définitions and limits of Bentham’s discussion on civil law in Limits’, in The Legal Philosophy and Influence of Jeremy Bentham: Essays on 'Of the Limits of the Penal Branch of Jurisprudence', dir. G. Tusseau. London: Routledge, 2014, pp. 117-‐128. 54 S.G. Engelmann, ed., Selected Writings of Jeremy Bentham, /ADD REFERENCE/, p. 12. 55 Bentham, Traités, p. 392: ‘Constitutional law has its direct and indirect legislation; direct legislation consists in setting up the offices to which public power is allocated […]. Indirect legislation consists in general precautions to prevent the misconduct, incapacity or embezzlement among those who administer in chief or as deputies’ [‘Le droit constitutionnel a sa législation directe et indirecte : la législation directe consiste dans l'établissement des offices entre lesquels toute la puissance politique se trouve partagée : il n'en est pas question dans cet ouvrage. La législation indirecte consiste dans des précautions générales, qui ont pour objet de prévenir l'inconduite, l'incapacité ou les malversations de ceux qui administrent soit en chef, soit en sous-‐ordre.’] [Translated by the speaker]



14

procedural law and adjective law; it deals with issues related to poor relief, prostitution,

marriage, benevolence, religion etc. Bentham was aware of the difficulty of making

indirect legislation fit into any category. However Bentham did not think that the

efficacy of indirect legislation would depend on what he considers as problems related

to the organisation of knowledge. He writes: ‘it will be no great matter: nothing of

moment will turn upon the accuracy of the division’.56

If the modern scholar finds it as difficult today to identify clear-‐cut guidelines to

distinguish indirect legislation both from legislation and from any other form of non-‐

legal social control, Bentham indicates that in his own endeavour he also found himself

in a tight corner.

3. A Hierarchy of Norms

The main quote above illustrates also the fact that indirect legislation is always defined

in comparison with direct legislation. Examples of this are rife.57 This also means that

direct and indirect legislation are not to be understood as equal in value, as they do not

operate at the same normative level. Indeed, from the beginning Bentham created a

hierarchy between both regulatory instruments. This point is found in the quote above

but also reasserted in the quote below:

Indirect legislation may assist direct, but can not supersede it. The one and the

other must go hand in hand. The events which call for the infliction of

punishment may be more rare: but punishment must always be held up.58

Some consequences must be drawn from the hierarchy of legal norms thus highlighted.

The full meaning of the concept can only be understood if the converse concept, direct

legislation, is grasped. On account of the close relationship between both forms of

control, it seems now necessary to give a short outline of Bentham’s legal theory.

Bentham defines a law as:

56 Bentham, Plan. 57 ‘By Indirect Legislation, the physical power of taking the measure opposed is taken away. Direct suffers the physical power to remain, but against the interest plays off another interest. Indirect Legislation operates by causing men not to have certain desires, instead of thwarting them when they are arisen’, in Bentham, ‘Appendix I’, in Plan [my emphasis] 58 Bentham, Appendix B, in Plan [my emphasis]



15

A law may be defined as an assemblage of signs declarative of a volition or

adopted by a sovereign in a state, concerning the conduct to be observed in a

certain case by a certain person or class of persons, who in the case in question

are or are supposed to be subject to his power: such volition trusting for its

accomplishment to the accomplishment of certain events which it is intended

such declaration should upon occasion be means of bringing to pass, and the

prospect of which it is intended should act as a motive upon those whose conduct

is in question.59

The definition exemplifies the specificity of Bentham’s concept of law, which has a much

wider meaning than that given by other jurists, as it includes any mandate given by an

individual to another, where there is a relationship of obedience and where the order

will be enforced in case of disobedience. An order given by the Monarch or a piece of

legislation passed by Parliament is as much an instance of direct legislation as any other

form of delegated legislation, such as the order given by the master to his/her servant, in

the course of business, or a parent to his/her underage child. In the first series of

examples, the law will be said to be made by conception, and the second by adoption.60

What makes a law a law, is not only who orders whom, since (almost) anyone can be the

sovereign as long as s/he can see his/her will enforced.

What is paramount in defining a law is whether there are means to enforce the rule in

case of non-‐compliance. Thus punishment lies at the heart of direct legislation.61

However, if Bentham’s legal theory were to ground legal power in punishment alone, it

would provide insufficient tools to understand the complexity of today’s world.

Regulation in Bentham uses other sanctions than legal/political sanction/punishment to

help secure compliance. Indeed, in Bentham, there are four sanctions: the political/legal

sanction, which is used when legal punishment is applied, the moral/popular sanction,

when the people pass judgment on the acts of another, the natural sanction, when an act 59 J. Bentham, Of the Limits of the Penal Branch of Jurisprudence, in The Collected Works of Jeremy Bentham, Oxford, OUP, 2010, p. 24 [my emphasis] 60 Bentham, Limits, p. 45. See also M. Bozzo-‐Rey, ‘Le statut du principe d’utilité dans la philosophie politique de Jeremy Bentham’. 61 Other commentators have shied away from this interpretation, finding in Bentham signs of a disposition to obey (see Ben-‐Dor, p. 49) to account for the complex reasons why individuals abide by a rule (see Hart, Essays on Bentham, pp. 252-‐7). Their arguments are convincing and complement a punishment-‐orientated law in Bentham (see Brunon-‐Ernst, Utilitarian Biopolitics, pp. 67-‐85), but cannot replace overwhelming evidence of the primacy of punishment.



16

of nature makes the consequences of one’s actions bear upon oneself, and the religious

sanction, when individuals act in certain ways to comply with the commands of a

transcendent being.

The comparison with direct legislation highlights a common object (controlling

behaviour to maximize utility), common regulatory mechanisms (the use of one of more

of the four sanctions) and the diverse nature of the legislator, as the indirect legislator is

any person who acts as a choice architect. Indeed commentators are increasingly aware

of the combined role of direct and indirect legislation in controlling behaviour.62 Indeed

Stephen Engelmann writes:

Bentham conceives all legislation as control, and in this sense it is direct and

indirect at the same time. […] So indirect legislation recommends ways for the

legislator to enlist other sanctions.63

A more minute study of the ways sanctions are used will enable us to identify the

specificity of indirect legislation, as compared with other means of control.

4. Indirect Legislation and Nudges

When legislating directly, the legislator, or any other individual, uses motives and

sanctions to direct behaviour of the offender. However when operating at the level of

indirect legislation, the legislator acts either on the motives of a different player (i) or on

motives from a different angle (ii).

(i) In this first form of indirect legislation, uses the same sanctions as direct

legislation, that is the political/legal sanction. A specific act is forbidden. The

commission of the forbidden act will lead to the infliction of a punishment, the

amount of which is measured to outweigh the probable benefit to be derived

from the crime, thus with the hope of deterring the criminal from committing the

crime. The only difference is that instead of punishing the wrongdoer, the

legislator punishes another player who might benefit also from the act of the

wrongdoer, or a player who might encourage the criminal to commit his/her

62 See Engelmann, ‘”Indirect Legislation”: Bentham’s Liberal Government’, pp. 372-‐3, Bozzo-‐Rey, ‘Le statut du principe d’utilité dans la philosophie politique de Jeremy Bentham’, and Brunon-‐Ernst, Utilitarian Biopolitics. 63 Engelmann, ‘”Indirect Legislation”: Bentham’s Liberal Government’, p. 371, quoted by M. Quinn, in University College London, Bentham Lecture on Indirect Legislation.



17

crime, or be instrumental in the commission of the offence. For example, instead

of punishing the drunken man, Bentham’s legislator would inflict penalties on

the bar tender for the sale of alcohol or on the farmer who produces and sells

hops; instead of punishing the prostitute, the pimp or the consumer would be

targeted. In this form, indirect legislation defines accessory offences rather than

regulatory mechanisms akin to nudges. It is not the case with the second form of

indirect legislation.

(ii) In the second form of indirect legislation, the legislator uses different tools to

regulate behaviour. He/she will endeavour to nip any obnoxious desires in the

bud, by creating an environment that prevents them from flourishing, for

example. This is akin to the way in which nudgers use human psychological

flaws to make the choice of certain option more likely than others. One way in

which Bentham does this is in setting default options. He writes: ‘Indirect

Legislation operates by causing men not to have certain desires, instead of

thwarting them when they are arisen’.64 By limiting the range of options to

choose from, Bentham sets default options, thus gearing individual behaviour in

certain ways, which are more likely to maximise the utility of individuals.

Because it centres on human desires, and because it is a form of social control

that inflicts less pain than legal punishment, it can legitimately stray away from

the field of crimes per se (and the wrongful/criminal acts which are its object),65

to control almost all areas of individual and collective behaviour, such as

poverty, drunkenness etc.

The distinction made here between the first (legal) form of indirect legislation and the

second (non-‐legal) form of indirect legislation is central to our discussion today and

resonates with the nudge debate. The title of Alemanno and Sibony ‘s book is Nudging

and the Law, thus clearly stating that nudges can operate in a legal framework and

64 Bentham, ‘Appendix I’, in Plan. Bentham makes a similar contention in the later Traités, p. 341: ‘L’objet de la législation directe est de combattre les désirs pernicieux par des prohibitions et des peines dirigées contre les actes nuisibles auxquels ces désirs peuvent donner naissance. — L'objet de la législation indirecte est de contreminer leur influence, en augmentant la force des désirs moins dangereux qui peuvent entrer en rivalité avec eux’. 65 See the reasons given by Bentham to allow an act to be considered as a crime in Bentham, Limits, p. 7: ‘These cases, it may be remembered, are of four sorts: 1. Where punishment would be groundless. 2. Where it would be inefficacious. 3. Where it would be unprofitable. 4. Where it would be needless.’



18

within the law. Political scientists and some legal scholars have been interested in the

nudge phenomenon on account of its displaying non-‐legal forms of social regulation.

Nudging was seen as an effective and cost-‐effective policy tools, as it did not entail

costly legislative instruments (costly in terms of the task of managing the execution,

correct implementation and punishment in case of non-‐compliance with the initiative,

and costly in terms of the degree of intervention on people’s freedom of choice).

However, these tools can also be used in a high-‐cost environment (i.e. legislation) to

enhance the effectiveness of the goals it sets out to achieve. This legal approach to a

concept which was mainly discussed among political scientists up to now, can be

reconciled if one considers choice-‐altering initiatives (or in other terms, freedom-‐

restraining initiatives) as lying on a continuum of degrees of interference from mild

nudging to compulsory legal provisions. Initiatives need to be ordered on a scale from

milder forms of intervention on the individual’s freedom of choice (dubbed soft

paternalism or nudges) to more intrusive forms of intervention (dubbed hard

paternalism). The highest level of intrusiveness on individual choice is reached when

the initiative takes the form of a law.66

This distinction between the primary (legal) form of indirect legislation and the

secondary (non-‐legal) form of indirect legislation needs to be further explored by

looking into the series of parameters to be taken into account when acting: the

conditions under which the act is committed (which Bentham names motives for

action) and the sanctions (or rewards) imposed on the commission or the non-‐

commission of acts.

In Traités,67 there are three motives to act: power (to be understood as the legal,

physical or financial power, depending on the manuscript sources), knowledge, and

interest (that is the motive to act). The first aims to deprive an individual of the power

of behaving in a way the sovereign (in Bentham’s sense) does not wish him/her to act.

Bentham names it the ‘nanny policy’, as it is the attitude of nurses when looking after

66 Sunstein, Why nudge?, pp. 17-‐18. 67 Traités de législation civile et pénale was first edited by Etienne Dumont in French with Bentham’s manuscripts, and later translated back into English as Principles of Penal Law, and republished in the Bowring edition of the Works of Jeremy Bentham.



19

children, taking away the power to act.68 He does not believe it is a relevant control

mechanism for indirect legislation, in which indirect means of control target the

accessory or other motives to act. The second seeks to circulate knowledge, so that,

thanks to a wider understanding of facts and of the consequences of one’s actions,

individuals will be more likely to make the right decision for themselves. The third, and

most important tool of indirect legislation, are inclinations.69 Inclinations are indeed

paramount, as they are the reasons why individuals decide to act (motives). Inclinations

(i.e. interests to act) are generally triggered by the combined forces of power and

knowledge.

There are different ways of directing inclination. Each society, in a particular place and

at a particular time, has its own, since each generation has its own range of values and

modes of life. When Bentham gives specific instances of how the sovereign can control

behaviour, he clearly states that they are isolated examples, and that governments,

institutions and organisations should find their own indirect regulatory tools.70

Indirect legislation works mainly with knowledge. It supplies knowledge to prevent

crimes or to promote behaviour that will maximise utility. If individuals are aware of the

ways in which fraudsters defraud their victims, they will be more unlikely to fall for

their tricks. But knowledge can also be used to promote utility maximizing behaviour. If

I have information on population trends, I can set up a business that will tap into those

resources. Here with indirect legislation, as with contemporary nudges, the legislator or

any other choice architect tries to improve our cognitive limitations to fight crime or

maximise utility more generally. Knowledge is thus operative for both forms of indirect

legislation. Is knowledge always a nudge? Provision of information by economic players

or by government agencies must not always be considered as a nudge or as soft

paternalism. It only counts as a nudge when the information is provided with a

68 Bentham, Traités, vol. 3, p. 332. Idea also used twice in the Plan as ‘policy of the nursery’ (§ 2) and in the teachings of the foolish and the wise nurses (Appendix I). 69 Bentham, Traités, p. 333. 70 Bentham, Traités, p. 355.



20

particular aim or with a view of satisfying a particular need.71 Provision of random or

non-‐selected information is not a nudge.72

Indirect legislation channels motives in certain directions thanks to the combined or

alternate operations of power, knowledge and/or inclinations. However, in order to be

fully effective, the inducements provided by indirect legislation generally need to be

backed by sanctions, which act as motives for individual action. Of the four sanctions

described above, in the primary (legal) indirect legislation the political/legal sanction is

operative, and in secondary (non-‐legal) indirect legislation, the popular/moral sanction,

through the Public Opinion Tribunal (POT), is the one mainly used.73

The POT is a fictional and temporary tribunal, made up of members of the community

interested in a given issue and passing judgement on the behaviour of public officials or

other individuals. It thus hands down the popular/moral sanction against deviating

groups or individuals. Bentham trusts the POT, through the credit or discredit it will cast

on certain types of behaviour, to promote certain actions and discourage others. It is a

self-‐regulating principle in society, according to the values of a given community on a

given issue at a given time.74 The POT relies on the widespread desire of human beings

to fit in, thus to comply with a behaviour set out as a norm of the community.

Indirect legislation is understood as using three different tools – power, knowledge, and

inclination – in order control individuals. The use of public opinion, which corrects

behaviour deviating from the norm set by the community, is of particular importance to

71 Sunstein, Why nudge?, p. 64. 72 For comments on provision of information as an instrument common to many non-‐legal normative tools, see Brunon-‐Ernst, “Le gouvernement des normes : Jeremy Bentham et les instruments de régulation post-‐modernes”, pp. 320-‐321. 73 Bentham, Traités, pp. 380-‐2; 392-‐402 (for the constitutional use of the popular sanction through the Public Opinion Tribunal, POT). Human beings have no control over the natural events, thus the natural sanction cannot be used as the basis of legislation, that is of rationalised and scientific policy-‐making. The role of the religious sanction is more ambiguous. Bentham recognises that it can have an effect on controlling the behaviour of God-‐fearing individuals. However, the religious sanction as it does not aim to maximise utility but a body of rules encapsulated in religious dogma, cannot be the basis of Bentham’s utilitarian state. 74 For the operations of the POT, see F. Rosen, Jeremy Bentham and Representative Democracy. A Study of the Constitutional Code, Oxford, Clarendon Press, 1983; LJ Hume, Bentham and Bureaucracy, Cambridge, Cambridge University Press, 1980; P. Schofield, Utility and Democracy. The Political Thought of Jeremy Bentham, Oxford, Oxford University Press, 2009; and Brunon-‐Ernst, Utilitarian Biopolitics, 67-‐85.



21

add motivation to individuals to act in the way that is required from them.75 Through the

use of default options and the combined use of knowledge and the sanction of the POT,

Bentham plays on the psychological springs of individuals to conform to a behaviour

that is presented as a norm. In its second form, indirect legislation thus uses some of the

regulatory mechanisms as contemporary nudges as described by Thaler and Sunstein.

The table below summarises the differences highlighted between the different forms of

indirect legislation described by Bentham.

Type of

indirect

legislation

Target Motive Sanction Is it a nudge?

Primary form

(legal)

Accessory Power

Knowledge

Inclination

Political/legal

Moral/popular

No

Secondary

form (non

legal)

Behavioural

bias

Knowledge

Inclination

Moral/popular Yes

The present talk has so far highlighted regulatory mechanisms in Bentham’s theory

which are akin to contemporary nudges. The aim was to put the theoretical

underpinnings and the pragmatic implementation of nudges into perspective. Bentham

thinks he is the first to theorise regulatory mechanisms he names indirect legislation. As

an intellectual historian, the present talk has allowed to identify the philosophical roots

of nudges, which have wrongly been hailed as new regulatory mechanisms. Beyond the

issue of historical accuracy, the question arises of what can Bentham’s theorisation

bring to improve our understanding of nudges.

5. Lessons of Indirect Legislation

75 On the issue of the importance of public opinion and peer pressure in the operation of indirect legislation, see examples given in Brunon-‐Ernst and Van Waeyenberge, ‘Indirect legislation in EU policy making : The case of the open method of coordination’; and in Brunon-‐Ernst, ‘Le gouvernement des normes : Jeremy Bentham et les instruments de régulation post-‐modernes’.



22

In the 1780s and in the 1790s, Bentham develops very powerful method of ascribing

(unlimited) power over individuals. He calls this method: indirect legislation. The

present talk has shown that indirect legislation covered widely different areas of action,

some of which would fall under the heading of the law, and others not. Some of the

initiatives promoted by indirect legislations are rooted in secrecy as direct legislation is

in publicity. Some administrative processes (one example is registration etc.) can be

publicized without undermining their efficacy. Indirect initiatives based on knowledge,

to inform with the aim of preventing crimes, need to be publicized to be effective. Some

others cannot. For example, many of the indirect rules relating to the prevention of

poverty, addiction etc. would lose their power if they were expressed and explained to

the general public (public sham executions for example).

This talk has shown that similar methods of social control, backed by the development

of behavioural sciences, some of which are inherited from Bentham’s indirect legislation,

are now being increasingly applied by governments and companies all around the

world. However, at the turn of the 19th c, any mention of indirect legislation was

dropped by Bentham in his theoretical writings and practical projects. What is the

reason of this disappearance of any mention of indirect legislation?

One of the reasons which can be given is that the methods used by indirect legislation

conflict with the ultimate aims of Bentham’s utilitarian society. Indeed, Bentham wishes

to make individuals as happy as possible in their communities. To do so, he bets on

institutions that will promote happiness. Key to the success of his plan is the wide array

of safeguards against individuals or groups seeking to deviate the benevolent aims of the

institutions for their own selfish purposes. In Bentham’s plan, the best possible

safeguard against misrule76 is publicity. Secrecy hides misdoings, but publicity, at all

levels of public life, reveals any misdoing. Publicity in Bentham is not a prop. It is the

very core of his system.

76 Of note the fact that Bentham thought indirect legislation could also be applied to fight misrule: ‘There remains that branch of legislation which is or might be directed against misrule. Even under this head the distinction between direct and indirect legislation may be still preserved; though the limits between those branches are scarcely so distinct nor is the distinction itself of so much importance in this case as in the former’ in Bentham, Plan.



23

Indeed, the very nature of some initiatives falling under the heading of indirect

legislation is secrecy. Bentham was well aware of the fact and mentioned it regularly in

his writings, as here in Limits, at the time when Bentham was writing his indirect

legislation projects, were mention of secrecy is made:

Legislation is a state of warfare: political mischief the enemy: the legislator is the

commander: the moral and religious sanctions his allies: punishments and

rewards … the forces he has under his command: … direct legislation, a formal

attack made with the main body of his forces in the open field: indirect

legislation, a secret plan of connected and long-‐concerted operations, to be

executed in the way of stratagem or petite-guerre.77

Why is secrecy so essential to indirect legislation? For certain initiatives, their efficiency

is based on the fact that individuals are generally unaware that their desire is being

manipulated. Take the following telling example: behavioural science has established

that individuals are prone to comply with behaviour that is presented as the norm of

action. As a consequence, advertising campaigns make good use of the fact, referring

frequently to the wide number of people who approve of a product (99% of women are

happy with the effects of an anti-‐wrinkle cream, 79% of people donate in times of global

emergency etc). Would the same campaigns have the same effects if individuals were

forewarned that their assent or benevolence was being manipulated by the impression

they get that they need to comply with what everybody else does?

Strangely enough, if any mention of indirect legislation disappears from Bentham’s

writings and his project is neither completed nor published, Bentham continues using

forms of control akin to indirect legislation of the primary (legal) form, that is using

accessory law to further his aims. In that form, indirect legislation seems compatible

with a well-‐governed utilitarian State.

A utilitarian State could not, without endangering its structure and the safeguards

against misrule, promote at the same time publicity and secrecy. From the grave,

77 Bentham, Limits, p. 233.



24

Bentham asks modern-‐day regulationists to raise the same questions: can a government

created for the benefits of the governed be properly managed if it taps into the

behavioural failings people are unaware of? Can a State which does not publicize the

reasons of its action escape the threat of tyranny?

Philosophical Roots of the Concept of Nudges

Documents

Transcript of Philosophical Roots of the Concept of Nudges