The Chimera of Proportionality: Institutionalising Limits on Punishment in Contemporary Social and Political Systems.

The concept of proportionality has been central to the retributive revival in penal theory, and underlies desert theory's normative and practical commitment to limiting punishment. Theories of punishment combining desert-based and consequentialist considerations also appeal to proportionality as a limiting condition. In this paper we argue that these claims are founded on an exaggerated idea of what proportionality can offer, and in particular fail properly to consider the institutional conditions needed to foster robust limits on the state's power to punish. The idea that appeals to proportionality as an abstract ideal can help to limit punishment is, we argue, a chimera: what has been thought of as proportionality is not a naturally existing relationship, but a product of political and social construction, cultural meaning-making, and institution-building. Drawing on evolutionary psychology and comparative political economy, we argue that philosophers and social scientists need to work together to understand how the appeal of the idea of proportionality can best be realised through substantive institutional frameworks under particular conditions.

Recent scholarship on punishment in developed countries has been much preoccupied with the turn away from rationales of punishment based on its supposed rehabilitative effects, and with the revival of retributivism in the modernised form of ‘just deserts’ or ‘the justice model’. Like all forms of retributivism, desert theory purports to offer a clear criterion defining the fittingness of penalties by reference to a particular offence by a particular offender. On this theory, punishment is justified in response to, by reason of, and in proportion to, the offender’s desert. Desert, in turn, is premised on his or her blameworthiness, which is generally understood in terms of a combination of harmful or wrongful conduct and culpability for that conduct: crucially, the ensuing punishment must be proportional to or commensurate with[2] that culpability, thereby curtailing any tendency towards injustice in the form of punitive excess. A large literature has accumulated spanning a number of key questions about the justice model across a number of disciplines: philosophical questions about the conceptual shape and normative justification of desert-based punishment;[3] legal and criminological questions about the best ways of realising the aspiration to deliver just deserts within legislative, sentencing and other criminal justice arrangements;[4] and sociological and political science questions about both the origins of the revival of retributivism and its effects.[5]

Within this last genre of scholarship, it is widely acknowledged that, in countries such as the UK and, particularly, the USA, the aspiration of many proponents of the retributive revival[6] to place clear limits on punishment by ensuring proportionality has not been fulfilled. Moreover, punitive rationales have in practice continued to be shaped by consequentialist considerations such as incapacitation and deterrence, with consequentialist and retributive considerations often blurred not only in public debate and political discourse but also in sentencing practice. Yet other countries that embraced the justice model, notably Sweden, avoided the ‘grade inflation’ in sentencing that occurred in the US and the UK.[7] Views differ on just why this has been the case, and we are some distance from understanding the degree to which the turn to just deserts had any independent force alongside changing patterns of crime and the undoubted economic, cultural and social factors which have contributed to the upswing in punitiveness in many liberal market countries since the 1970s.[8]

In a recent paper,[9] we sketched an argument about the counter-productive implications of inviting ‘affective blame’ – the hostile, negative emotions such as hatred, anger, resentment, indignation, disgust, disapproval, contempt and scorn, which often accompany a judgment of ‘detached blame’ in the sense of a judgment of blameworthiness – into the sentencing and penal process. Drawing on a clinical model of holding patients responsible for harmful or wrongful behaviour without affectively blaming them, we made the case that a sentencing and penal process which equally sought to avoid affective blame would be entirely consistent with respect for an offender’s agency and responsibility for his or her offence. Even if a retributive rationale for punishment was accepted, we claimed, it would not follow that affective blame should be part of the criminal justice process, not least because there was reason to think that the retributive revival may have contributed to increasing punitiveness by legitimating hostile emotions against offenders without successfully institutionalising constraints on how these emotions should be expressed, acted upon and regulated. It follows that finding institutional mechanisms to distance affective blame is a key challenge for contemporary criminal justice systems; and we accordingly explored a number of ways in which such constraining mechanisms might be developed through adaptation of the protocols used to distance affective blame in the therapeutic process. Our paper also invited, but did not address, questions about the broader social and political conditions which might make the pursuit of penal responsibility without affective blame possible to achieve and, conversely, about the reasons why desert theory’s aspiration to limit punishment turned out to be hard to achieve in certain countries.

In this paper, we turn to these further questions by focusing on the issue of proportionality – one of the central concepts of the retributive revival, and the main idea underlying its normative and practical commitment to limiting punishment.[10] Indeed, proportionality stands as the key concept in a much longer history of efforts to modernise and temper punishment, occupying as it does a central place in the work of Enlightenment thinkers and reformers across many nations: Beccaria, Bentham, Jefferson and Montesquieu.[11] The promise of proportionality not only remains central to the appeal of desert theory, but also commands allegiance from many theorists who support ‘mixed’ theories of punishment that combine desert-based and consequentialist considerations.[12] But, as we shall argue in this paper, the retributive revival was founded on an exaggerated idea of what proportionality, in itself, could offer, and in particular failed properly to consider the question of what sorts of institutional conditions are needed in order to foster robust limits on the state’s power to punish.[13] The idea that appealing to proportionality as an abstract ideal can deliver limits to punishment, we will argue, is a chimera: for while proportionality has a clear formal meaning – indicating the existence of a broad moral equivalence or comparability between two different phenomena such as a particular crime and a particular penalty – it only has substantive upshot where there is agreement about, or effective enforcement of, substantive criteria of equivalence or comparability. Thus, ideally, adequate limits to punishment need to be grounded in substantive judgements about fair and appropriate penalties which are meaningful to, and regarded as legitimate by, the populace in whose name they are imposed. The challenge, we suggest, is therefore for philosophers and social scientists to work together to understand the ways in which the undoubted appeal of the idea of proportionality – an appeal which is reflected in its currency in a wide range of prevailing social and legal discourses, many of them concerned with the limitation of power[14] – can best be articulated within such a substantive framework under contemporary conditions. What has been thought of as proportionality, in short, is not a naturally existing relationship, but a product of political and social construction, cultural meaning-making, and institution-building.

The paper falls into three parts. First, we describe the emergence of neoclassical penal philosophy in the guise of ‘just deserts’, sketching the main causes of its resurgence and the main ethical and political aspirations of its supporters, before proceeding to examine the impact of the just deserts movement, and to evaluate its contribution to justice and humanity in punishment. In the second part of the paper, we review certain features of the operation of punishment in early modern systems, arguing that their capacity to coordinate punishment in such a way that it was perceived as fitting to the crime derived from a hierarchical social order, an association of certain forms of penal and political authority with the sacred, and the currency of a distinctive symbolism of equivalence. Though quite inconsistent with modern ideas of proportionality, particularly in the discretionary power which implied uneven application of penalties, we suggest that these three features of the context in which penal practice went forward in very different societies sheds light on how substantive criteria of fittingness or equivalence depend upon background social and cultural conditions. We then draw on this analysis to argue that the neoclassical revival of the late 20th century was problematic from its inception, because the metaphors of ‘desert’ and proportionality, particularly in certain countries, are no longer so obviously grounded in the widely shared symbolic systems representing agreed social norms, or in forms of political or religious authority, which previously animated and stabilised substantive judgments of equivalence or fittingness.[15] Proportionality, in other words, does not have an independent effect: where it ‘works’ to limit punishment, this is because of its articulation of, and resonance with, deeper conventions, normative systems, political institutions, and social structures.

In the third part of the paper, we explore what can be learnt from existing research – notably from the comparative political economy of punishment and from evolutionary psychology – about the conditions under which meaningful limits on punishment can be institutionalised in contemporary polities, which are configured differently not only from early modern systems but also from one another. Most of the institutional mechanisms, and many of the evaluative ideas, which generated limits on punishment in early modern systems, lack moral and practical purchase under late modern conditions, particularly in liberal market countries. Hence, unfortunately, the revival of retributivism has all too readily translated, under certain conditions, into a politics of anger, affective blaming and ‘othering’ of offenders – a process against which appeals to proportionality have little bite, and in which retributive policies are in fact counter-productive to the very aims of many of the liberal supporters of the retributive revival.[16] But we argue that much can be learnt about the conditions under which delineated limits on punishment command authority by examining the broader institutional structure of polities whose systems of punishment have indeed achieved stability. Drawing analogies between findings in evolutionary psychology and those in the comparative study of punishment about the broad conditions which foster reconciliatory as opposed to vengeful reactions to hostile or harmful conduct, we argue that the different impact of the just deserts movement – and of the ‘neoclassical’ revival of ideas of proportionality more generally – in countries like the US or England and Wales as compared with Nordic countries is in significant part to be attributed to institutional as well as cultural features of the economic and political system in those countries.

The structure of the argument is as follows. In the first section, we argue that the neoclassical claim that punishment is limited by proportionality is belied by the fact that recent years have shown a significant upswing in punitiveness in some but not all countries which adopted the ‘justice model’ amid the neoclassical revival. In the second section, we argue that proportionality is a chimera when appealed to in the abstract; the notion of a stable social sense of ‘fittingness’ of punishment to crime can be worked out only under very particular conditions – some of them illustrated by early modern societies, in which the conditions which underpinned that achievement are ones which we no longer have nor would we wish to have. We then show that the contemporary societies which have managed to sustain stable limits on punishment have done so not through appeals to proportionality as such, but rather through institutions and attitudes that foster reconciliatory dispositions between citizens, and we show that this finding is supported by a developing literature in evolutionary psychology about the conditions which orient people towards the choice of reconciliatory as opposed to retaliatory responses to hostile conduct. Irrespective of any appeal to proportionality, those systems which feature cultural and institutional arrangements which foster high ‘Associational Value’ – in other words, greater expected benefits of continued cooperation between individuals and groups – also foster reconciliatory dispositions.

THE NEO-CLASSICAL REVIVAL AND ITS EFFECTS

The articulated justification – or perhaps we should say rationalisation – of punishment in Western countries over the past two hundred and fifty years has followed a relatively clear, and widely discussed, trajectory.[17] To sketch this historical story only crudely: the early modern retributive corporal penalties gradually gave way, from the late 18th century, as a ‘civilising’ process issued in a preference for less overtly violent, and less publically visible, forms of sanctioning. By the same token, a demand arose for the highly discretionary forms of power, through which early modern penalties were typically delivered, to be rationalised and regularised as a result of the gradual modernisation and democratisation of the hierarchical forms of political authority from which they emanated. Beccaria[18] and Bentham[19] are, with good reason, thought of as the key figures in this modernisation of the rationale of punishment, though Beccaria is of particular importance in that his work stands as an important modern source of both of the two main ideas which have coincided and competed with one another as justifications of state punishment. These are, first, the ‘neo-classical’ argument that punishment is in some sense a morally appropriate equivalent to an offence, and is thus constrained by the requirement of proportionality (an argument to be found in pure retributive form in the work of, for example, Kant, who further saw the imposition of deserved punishment as obligatory rather than merely permissible); and, second, the utilitarian argument that punishment, as a prima facie evil, can only be justified by countervailing good consequences, achieved through specific or general deterrence, incapacitation, rehabilitation, restitution or moral education (an argument most rigorously worked out by Bentham).

The ideas that punishment is in some deep sense retribution for a blameworthy act or an institutionalised expression of affective blame, and that an offender’s blameworthy conduct is what justifies punishment, have deep historical roots. Yet the fortunes of retributive theories of punishment based on culpability – blameworthiness in the sense of responsibility for wrongful conduct – have been mixed, particularly since the development of consequentialist justifications for punishment, from the writings of Cesar Beccaria and Jeremy Bentham to the present day. The details of this history lie well beyond the scope of this paper. But a brief review of the relative fortunes of retributive and consequentialist rationales is useful in identifying the precise misunderstandings about the concept of proportionality as the modern expression of the retributive ethic, and about its capacity to anchor punishment within clear limits, to which we want to draw attention.

Until the 18th century, what we might call a retributivism of symbolic equivalence dominated both the theory and practice of punishment in most of Europe. To focus on England as an example, a panoply of (to our eyes) harsh corporal and social penalties – branding, maiming, whipping, pillorying, the use of the branks or ‘scold’s bridle’ and, ultimately, capital punishment – were deemed equivalent to the relevant offence, and inflicted in public, hence reinforcing spectators’ apprehension of that equivalence. In a pre-democratic world, the moral logic of the lex talionis – ‘an eye for an eye …’ – appears to have struck a deep chord, not least because of the resonance between penal authority and various Christian doctrines. Indeed the idea that penalties should be measured and proportionate finds its origins in the ecclesiastical law practice of imposing graduated fines[20] – a practice in stark contrast to princely, secular early modern penalties, in which the monarch’s power to punish or to pardon through an exercise of ‘grace’ was the antithesis of the classical aspiration to mete out even-handedly a ‘just measure of pain’.[21]

A robust form of retributivism – the idea that the application of a deserved penalty is, a priori, a self-evident good barely requiring justification, indeed something which is morally required – appears not to have posed significant problems of legitimation. It is only from the middle of the 18th century that this self-evidence of punishment seems to have been under challenge. This was for a number of reasons. They include the emergence of more egalitarian and democratic ideas in Europe and North America; the growth of sentiments opposed to the public display of violence; and the growing regulatory ambitions of a more organised state. For both moral and prudential reasons, the banal violence of the early modern penalties began to be questioned, and the seeds of a reformed penal system, focused not only on doing justice but also on disciplining the subjects of punishment in a more systematic way, notably through the modernisation and expansion of the prison, were planted.[22] But the declining self-evidence of the ancien régime in punishment may, ironically, also have had to do with the attempted rationalisation and modernisation of retributivism which Beccaria’s (and Kant’s) philosophy exemplified, with the growing acceptance that punishment required abstract moral justification itself a reaction to the declining force of early modern symbols of equivalence largely expressed through physical punishments. In particular, the recognition that desert is calibrated not only with the offender’s misconduct or manifested bad character but, importantly and additionally, with the degree of the offender’s responsibility, disrupted any simple symbolism of equivalence.

This was not, of course, in the nature of a clean break from retributivism to utilitarianism; in particular, religious overtones to punishment as a form of atonement persisted. The modern prison was, after all, constructed as a ‘penitentiary’, and early prison regimes were strongly focused on encouraging the offender to reflect on his or her wrongdoing and come to a state of repentance – an idea which it was possible to bring into conformity with the utilitarian goals of discipline and reformation which were becoming so influential.[23] Conversely, many pre-modern penalties – branding and whipping for example – must surely have had incapacitative and deterrent rationales and effects alongside their retributive foundations. But as the 19th century wore on, the moral project of modern imprisonment became increasingly tied up with the calculation of its consequences: initially with more discourses focused on deterrence; then on identifying, through social Darwinist ideas and, ultimately, Lombrosian criminology, certain ‘types’ in need of moral re-education or incapacitation; and, into the 20th century, with a welfarist philosophy applied to the rehabilitation of offenders.[24]

The first two thirds of the 20th century have been, aptly, characterised in Garland’s influential work as a period in which punishment in most advanced democracies was increasingly constructed in welfarist terms.[25] While punishment was certainly viewed as an appropriate state reaction to wrongdoing, and the panoply of responses deployed by most democratic societies included a range of severe sanctions, often including capital punishment, the infliction of hard treatment was tempered by a distinctively welfarist or ameliorative spirit. Probation and the development of various forms of training for young offenders were early signs of this penal welfarism. And after the Second World War, particularly in countries like Britain and the United States, a distinctive form of penal welfarism – widely known as the rehabilitative ideal – established itself as a central rationale for state punishment. Punishment was justified by appeal to its potentially rehabilitative consequences; and the variety of ameliorative and even therapeutic programmes and ‘treatments’ burgeoned.[26]

The penal practices and articulated penal policies and principles of national governments, of course, typically exhibit a mixture of potentially philosophically incompatible elements such as general and special deterrence, incapacitation, reform, denunciation, reprobation and retribution. So one should not exaggerate the completeness with which a welfarist or rehabilitative ethic prevailed across penal practices of the countries which followed the welfarist path. But it did, undoubtedly, enjoy a period of dominance; and, just as clearly, its influence declined as a result of a decisive change in public penal philosophy from the 1970s on. As recently as 1968, H. L. A. Hart, in his postscript to Punishment and Responsibility,[27] reflected something approaching a liberal consensus when he remarked that ‘Few people would now advocate so thoroughgoing a variety of retribution as the pure retributive view which regards the return of suffering for moral evil voluntarily done [as] itself just or morally good’,[28] let alone the further Kantian view that such punishment is not merely morally permissible but morally obligatory.[29] While Hart accorded some respect to the argument that the value of punishment consists in ‘the authoritative expression … of moral condemnation for the moral wickedness involved in the offence’ – an institutionalised communication of resentment – the idea that returning suffering for evil is intrinsically good is regarded as ‘a mysterious piece of moral alchemy’ or ‘a primitive confusion’ between punishment and compensation.[30] The prevailing sense of Hart’s book is a confident liberal consensus that while consequentialist approaches to punishment could be abused and subject to excesses, and that adequate limits on the distribution and quantum of punishment were a key concern, pure retributivism was a thing of the past. Yet a mere 20 years after the publication of Hart’s influential collection, the retributive tradition, repackaged as the ‘just deserts movement’ or the ‘justice model’, had captured the imagination of both policy makers and penal philosophers in the USA, the United Kingdom and many other countries. How did this happen?

The reasons for the demise of the rehabilitative ideal are well known,[31] and we will not rehearse them in any detail here. Two important concerns were uppermost. First, the rehabilitative ideal led to a rise in lengthy indeterminate sentences based on predictions of ‘dangerousness’ or need for treatment, and implied broad and unaccountable official discretion as to release date, based on expert judgments about prognosis, risk and ‘cure’. Alongside this concern with executive discretion sat a civil libertarian concern about the capacity of indeterminate sentencing to violate offenders’ rights. The concern to establish respect for agency and responsibility as core values of the criminal process was not, of course, restricted to proponents of ‘just deserts’,[32] it also characterised theories which seek to combine backward-looking and forward-looking considerations in the justification of punishment.[33] But the justice model successfully presented itself as the approach best able to generate an account of punishment compatible with both full respect for offenders as agents and the modern ambition to achieve even-handedness, constrained discretion and – most important to our argument in this paper – limits on state punishment.

The limiting principles celebrated by the justice model are underpinned by the further claim that a certain proportionality can be built in to the quantum of punishment; and that this in turn is guaranteed by the backward-looking orientation of punishment’s justification. The idea that punishment is conditioned on, and responds to, blameworthiness implies that the punishment must be proportionate to it. It was therefore widely assumed – and hoped – that the proportionality implied by desert would provide clear limits on punishment. By linking not only the justification but the distribution and quantum of punishment to the offender’s desert, in the sense of the level of blameworthiness appropriately to be attached to the offence, the justice model appeared to offer a clear limit on the extent of the state’s right to punish, and presented itself as a progressive, humane and liberal approach.

This aspiration is, however, very far from having been universally realised in practice. Thirty years on, the practical impact of the justice model presents a mixed picture, and in liberal market countries such as England and Wales, Scotland, Australia, New Zealand and – most spectacularly – the United States, the scale of punishment has increased relentlessly. This penal ‘grade inflation’ has a number of different dimensions, of which we here mention only the most obvious. The imprisonment rate per hundred thousand of the population has soared, as a result of both increasing flows of defendants through the criminal courts and a rise in sentence levels, particularly for certain categories of offence.[34] Mandatory sentencing systems have become a common feature of these liberal market systems, and many of the sentencing guideline systems designed to foster the determinacy sought by the justice model[35] in fact led to longer sentences as a result of the political choice to structure them around very high tariff scales, and to reduce judicial discretion to temper sentence severity.[36] In addition, prison conditions in many of these countries have deteriorated, with not only overcrowding but also deliberately punitive and restrictive regimes ousting longstanding ameliorative programmes focused on education, vocational training, drug and alcohol treatment, and therapeutic interventions.[37] Finally, stigmatising post-sentence conditions and disqualifications, which are inimical to reintegration, such as disenfranchisement, placement on offender registers, and ineligibility for certain jobs and public benefits, have been, particularly in the United States, on the rise.[38]

Furthermore, a number of countries in which the justice model had the most decisive influence on policy – notably the United States and England and Wales – have also seen a continuation or even acceleration of practices, such as indeterminate sentencing and preventive justice, which were thought to express the more extreme injustices and disrespect for the rights and agency of the offender that characterised the rehabilitative ideal, in principle and in practice.[39] It turns out, as Jeffrie Murphy has put it, that retributivism has two faces: one oriented to justice and respect for responsible agency; the other oriented to vengeance and to moral emotions such as resentment, anger and other hostile, negative attitudes typical of affective blame, which when allowed to issue in what Murphy calls ‘deep character’ retributivism can easily rationalise severe and ‘othering’ penalties.[40] And appeals to ‘proportionality’ have little limiting potential here, for while empirical studies indicate a noteworthy degree of consensus, even across different countries, on the relative seriousness of standard offences – so-called ‘ordinal proportionality’[41] – they reveal no such consensus about what this implies in terms of what penalty is suitable – ‘cardinal proportionality’.[42] In ‘core’ areas of criminal law, appeals to ordinal proportionality may therefore provide some basis for institutional arrangements such as sentencing guidelines;[43] and indeed the various programmes of sentencing reform launched in many jurisdictions in the wake of the just deserts movement provide plentiful examples of successful institutionalisation of stable relativities between penalties. But the lack of any comparable consensus about cardinal proportionality implies that appeals to proportionality are, under current conditions, unlikely to be a successful basis for institutionalising substantive criteria of a punishment’s ‘fitness’ – hence, as we argue, rendering the appeal to proportionality chimerical as a basis for limiting punishment. In the light of this difficulty, it is therefore not surprising that many desert theorists have focused primarily on ordinal rather than cardinal proportionality. But the very idea that a certain punishment is deserved – and with it, the whole basis for proportionality’s purported capacity to set upper limits on, or to generate substantial criteria of fittingness of, punishment – rests on cardinal rather than ordinal proportionality.

Strikingly, in other countries which also moved towards sentencing guidelines and a range of institutional arrangements counselled by the justice model’s version of neoclassical retributivism, which emphasises the importance of responsible agency – notably Sweden and the other Nordic countries[44] – the increase in the scale and intensity of punishment has been less marked. Clearly, both trends in crime and in public concern about crime, as well as distinct penological traditions and sentencing institutions,[45] are at issue here. But the very different impact of the justice model in different countries raises important questions about the cultural, political, social and institutional conditions under which the retributive revival invites an escalation of insatiable affective blame, and those under which it fosters – or can be made consistent with – determinacy and moderation in punishment. It raises questions, in short, about the cultural and institutional arrangements which can make the metaphor of proportionality meaningful, and punishment accordingly limited in real terms.

FIXING PUNISHMENT IN EARLY MODERN SYSTEMS: SYMBOLIC EQUIVALENCE, STATUS HIERARCHY AND THE SACRED

In addressing these questions, a useful place to start is by looking at certain features of the penal systems which predated the modern push towards consequentialism in punishment, or which continue to exist in parts of the world whose penal practices were less decisively affected by the processes of modernisation and democratisation which swept Europe and the United States in the 18th and 19th centuries. In many of their earlier forms, retributive ideas related to wider frameworks of social and ethical meaning which underpinned prevailing social and political systems and institutional practices – an environment which had fundamentally changed by the time attempts were made, in the 1970s, to revive and realise retributive ideas. An effort to understand the nature of the lost cosmology which leant meaning to early modern punishments – impressionistic as it must be within the confines of a single paper – may, we suggest, help to explain such capacity as they enjoyed to institutionalise a practice of retributive punishments within clear limits, the very ‘distance’ of the examples helping us to think more imaginatively about how the legitimation of penal practices is shaped by prevailing interests, institutions and ideas.

Both the historical and the conceptual foundations of retributivism are generally associated with the lex talionis. Taking this as our core example therefore – while bearing in mind the analogies with other systems such as the corporal penalties typical of the early modern criminal justice systems of Europe and the system of penalties prescribed by sharia law – we would suggest that early systems of retributivism had at least three distinctive features key to their legitimation and stability. The first was a social ontology of status hierarchy which rendered the legitimation of punishment a far less pressing issue than it is in liberal societies which place a high value on individual autonomy,[46] and which prescribed different forms of penalty calibrated with social class and oriented to widely understood marks of degradation.[47] The second was a vision of penal authority tied to the sacred. This was manifested in either religious doctrines of damnation, expiation, atonement, penitence and so on, or a vision of political authority marked by symbols of supra-human or highly particular human authority such as the ‘divinity’ or ‘majesty’, implying the princely power not only to punish but to pardon through the exercise of ‘grace’.[48] The third was what we will call a shared symbolism of equivalence. Of course, few systems have attempted to enact literally the symbols of cardinal proportionality contained in the lex talionis. But the fact is that, to contemporaries, prescribed punishments in many early modern systems related in some intuitively meaningful way to the wrong done (and typically did so independently of any psychological judgments of responsibility which complicate assessment of culpability in modern desert theory, and which emerge alongside modern, psychological conceptions of selfhood).[49] Each of these conditions made it possible to coordinate the expectations of the relevant actors so that they were in agreement on not only the need for a penal response, but equally the appropriate shape which that response might take. If we look back to the early modern English penalties which seem so cruel, horrifying and indeed disproportionate today, we can nonetheless acknowledge that they found their form, and took their place, within a composite view of political authority which itself drew on symbols, doctrines and common values, as well as being stabilised by a rigid status hierarchy and an authoritarian system of governance – conditions which also applied, of course, to the church, whose panoply of graduated fines represents a fascinating precursor to modern conceptions of proportionality.[50] This was, of course, by our standards an unduly hierarchical and undemocratic world; and the rituals which elaborated different forms of capital penalty, as well as the different forms of corporal penalty such as branding, pillorying and so on, strike us, for good reason, as deeply inhumane. Moreover the idea of punishment before the public gaze – a key condition for the diffusion of shared perceptions of symbolic equivalence – is now regarded as fundamentally uncivilised.[51] But it is worth remembering that the world view from which these penal practices proceeded was one in which the symbols of state and penal authority formed part of a broader cosmology of authority and right, itself often bound up with claims about divine or traditional legitimation, and one which commanded respect well beyond the elite.

Consider, for instance, the widespread popular participation in the drama of the scaffold – not only as crowds witnessing executions,[52] but as consumers of many forms of popular culture. These include the criminal autobiographies which, in the form of Defoe’s Moll Flanders,[53] finally transmuted into an early form of the realist novel;[54] the Newgate Ordinary’s widely read reports on the spiritual condition and conduct of the condemned in the run-up to execution; and the street ballads and pamphlets which were preoccupied with not only the drama of justice being done but also the condemned offender’s reception of the punishment. Often – and, within this literary genre, ideally – such reception included scaffold speeches that consisted of confession and penitence, transforming execution into a potential scene of redemption within Christian cosmology. Indeed, many historians’ accounts of the behaviour of the scaffold crowd and of the choreography of the procession and execution process suggest that the offender’s confession was key to the successful enactment of what amounted to a highly ritualised – and cathartic – form of drama.[55] Despite the hierarchical authority and power with which punishment was meted out, all the participants had a part to play in the script.[56]

The vision of legitimate power and authority which underpinned the early modern English penal system did not, of course, consist only or even primarily in Christian doctrines: here as in other early (and some existing) systems, social codes of honour and status were also important.[57] But there can be little doubt that this world of widely shared deference to authority, and vision of authority as vested with, broadly, sacred significance – think, for example, of the conception of ‘majesty’ – made a certain sense of penal practices, constituting penality as one among many social rituals in which that hierarchical authority was enacted. Within this system of meaning and the prevailing cultural, moral and political economy, as Douglas Hay has persuasively argued,[58] prerogatives such as pardoning and mercy, highly discretionary and unevenly applied though they were, made sense. But, as the source and form of political authority has been subject to a process of systematisation and rationalisation in the construction of the modern nation state, older forms of ordering and of meaning-making have been eroded.[59]

Equally important, the symbolism of equivalence which underpins the lex talionis, sharia justice, or the corporal and capital penalties perhaps most vividly exemplified by Foucault’s famous portrayal of the execution of the regicide Damiens in the opening pages of Discipline and Punish[60] have, for the majority of those living in Western countries, lost their persuasive appeal. But to contemporaries, not only the complex dramaturgy represented by spectacular penalties such as that of Damiens, but more routine corporal penalties such as maiming and branding, related in some intuitively meaningful way to the wrong done.

The effort to build a modern equivalent to the lex talionis has long taken the form of appeals to proportionality. Indeed, Thomas Jefferson’s 1778 Bill for Proportioning Crimes and Punishments in the early formation of the United States combines the appeal to proportionality with a continuing commitment to talionic punishments to a striking degree: In the neoclassical revival of the late 20th century, the appeal to proportionality tended to be realised rather in the technical form of procedural mechanisms such as sentencing guideline systems. But the stark fact is that those systems produce staggeringly different judgments of what counts as a ‘proportional’ penalty for, say, theft, rape or manslaughter, in countries such as Sweden on the one hand and the USA on the other – countries which, for all their differences, share many features of political and social culture and economic development. Moreover there are significant differences between what is now considered proportionate within particular countries as compared with 40 years ago – as exemplified by the emergence of mandatory sentencing in many jurisdictions, notably the United States and England and Wales.[62] We now live in a world which has largely moved away from attachment to physical symbols of penal equivalence, from the established markers of status hierarchy which underpinned deference to established political authority as such, and from a view of political authority as invested with sacred power – all features which contributed to the stabilisation of criteria of ‘fitting’ punishment.[63] The modern answer to the (unappealing) pre-modern equilibrium is an appeal to proportionality. But such an appeal can in itself contribute little to the construction of norms adequate to limit state punishment. Indeed, proportionality represents an intuitively shared starting point precisely because it is virtually indeterminate in its substantive implications: in other words, it simply defers the crucial and complicated processes of meaning-making, consensus-building and institutional development.

Whosoever on purpose and of malice forethought shall maim another, or shall disfigure him, by cutting out or disabling the tongue, slitting or cutting off a nose, lip or ear, branding, or otherwise, shall be maimed or disfigured in like sort: or if that cannot be for want of the same part, then as nearly as may be in some other part of at least equal value and estimation in the opinion of a jury …[61]

THE CONDITIONS OF ‘PROPORTIONALITY’ IN LATE MODERN SYSTEMS

To modern readers, the suggestion that there is anything to be learnt of relevance to penal reform today from early modern penal systems, let alone that such systems were better able to fix conventional values on punishment which were in some sense ‘proportionate’ or ‘equivalent’ in the way we think of these requirements today, may seem outlandish. Our implicit self-understanding, in both our penal thinking and our political thinking, is that the last three centuries have been, albeit with horrifying setbacks, an era of progress and of increasing civilisation, not least in our practices of punishment. That progress is associated with a recasting of what we might call the cosmologies, or systems of symbolic meaning, which animated and legitimised older practices of punishment and state authority, as atavistic, pre-modern, part of the ancien régime. In the process of modernisation, those symbolic systems were either fully rejected as irrational or uncivilised, or demoted in political importance. But if they were indeed important in calibrating, legitimising and stabilising punishment, efforts at penal reform may have been hindered by a failure to appreciate their role or to acknowledge the impossibility of reviving them by an act of political will. The neoclassical aspiration was to create modern mechanisms for fixing cardinal as well as ordinal proportionality; but the impact as well as the shape of such mechanisms has varied decisively even among otherwise comparable countries. This raises questions about the implications of the value pluralism and heterogeneity typical of individualistic and secularised western democracies, and particularly in the competitive political systems of the liberal market economies such as the United Kingdom and the United States, for our efforts to garner consensus about appropriate limits on punishment under contemporary conditions.

A useful goal is therefore to work towards an understanding of the institutional conditions and broader social arrangements under which a successful orientation to limited punishment and to reconciliation may be stabilised. In this section we begin this task, arguing that this capacity is greatest where – by broad analogy with early modern systems – appeals to proportionality connect with a larger shared frame of common meaning and consensus about legitimate authority;[64] and where they take place within a social and institutional context in which the relevant actors share expectations of both interdependence and the ability of social norms and institutions to coordinate effective cooperation between them in the medium to long term.[65] This thesis is based on two very different fields of research. First it draws on work in evolutionary psychology which explores the significance of two contrasting standard responses to exploitative behaviour: vengeance and reconciliation; and which draws a clear distinction between the evaluation of the level of seriousness of offence and the choice of response. As we shall argue below, this research has some interesting implications for our efforts to understand why some modern social orders seem better able to institutionalise moderation in punishment than do others. Second, our thesis draws on several paradigms in comparative political economy which have developed models of distinctive social, economic and political systems with differential capacities to coordinate behaviour, motivate consensus, and be guided by an orientation to long term relationships between members. Focusing on the implications of these literatures in evolutionary psychology and in the comparative political economy of varieties of capitalism, political systems and welfare regimes, we aim to produce an analysis of the political, social and institutional conditions which are most likely to stabilise and moderate the conventions of proportionality in punishment.

Literature in evolutionary psychology (as well as in game theory) has explored the ways in which group members react to hostile conduct and how these impact on relationships and group dynamics.[66] Two basic responses to hostile conduct exist: retaliatory and reconciliatory. Retaliatory responses aim to reduce motivation to exploit others by imposing a cost and thereby adjusting the expected benefit of any hostile conduct for the aggressor in future. Reconciliatory responses, in contrast, aim to reduce motivation to exploit others by seeking to restore cooperative relationships, thereby preserving the possibility of mutually beneficial interactions between group members in future. Both responses thus aim to protect against future exploitation, but also carry risks. Retaliatory responses risk escalating aggression and threatening the fabric of on-going relationships. Retaliation protects against future exploitation only in so far as the aggressor fears getting caught and so being subjected to retaliation: it fosters no intrinsic desire to end hostilities in the aggressor; indeed, it may undermine any such desire and hence increase aggression. The success of retaliation as a response therefore depends on adequate monitoring of the aggressor and possession of the power to effectively harm them in turn. Reconciliatory responses demand a willingness to forgive the aggressor and wipe the slate clean for the sake of future mutually beneficial relations. Reconciliation aims to foster an intrinsic desire in the aggressor to end hostilities by eliciting remorse and recognition of the value of the relationship. It therefore does not depend on monitoring and power. Reconciliatory responses do however risk leaving the forgiving party vulnerable to future exploitation, if the aggressor deceives them about their commitment to avoid hostilities and maintain good relations. Nonetheless, from the long-term perspective, reconciliatory responses are optimal when successful, for they reduce the risk of the aggressor perpetrating harm without incurring the cost of monitoring and the maintenance of coercive power in the future, while bringing the benefit of maintaining relationships so far as possible.

Conditions that foster both the likelihood and the viability of adopting a reconciliatory response include: being in a kin relationship; recognition of the aggressor’s work or other social productivity; the aggressor’s remorse or repentance; and the mutual dependence in various forms of the aggressor and other members of the group. Within evolutionary psychology, these conditions are understood as pertaining in one way or another to perception of the aggressor’s ‘Associational Value’: in other words, the value which can be expected to derive from future interactions with them. Where Associational Value is high, the orientation to forgiveness and reconciliation is accordingly enhanced; where it is low, either because of features of the aggressor themselves or the relationship they have with members of the group – or, as we shall suggest below, features of the socio-political environment – the orientation to retaliation will be stronger. Evaluations of how to react to a hostile act, holding evaluations of its seriousness constant, are accordingly shaped by Associational Value: by the expected value of future interactions with the offender.[67]

This research resonates with an extensive literature in comparative political economy. The Associational Value findings suggest that countries with higher levels of social solidarity and trust will be those in which it is easiest to create institutions and practices oriented towards reconciliation and hence to moderate retributive excess; and that it may be harder to generate barriers to retaliation and escalating aggression, and incentives to reconciliation, in large scale, anonymous, urbanised societies, where confident evaluations of high future Associational Value are harder to make. Indeed, comparative sociological evidence supports this suggestion, in that the countries which have experienced the most dramatic increase in punitiveness over the last 40 years are those with the lowest levels of solidarity and trust and the highest levels of heterogeneity along a range of social indicators.[68] More systematically, there may well be important social and institutional differences which structure the incentives and capabilities towards retaliatory versus reconciliatory reactions to crime. And it is here, we would argue, that evolutionary psychology research resonates with three related models in comparative political economy: Hall and Soskice’s characterisation of ‘liberal’ and ‘coordinated’ ‘varieties of capitalism’;[69] Esping-Andersen’s characterisation of ‘three worlds of welfare capitalism’;[70] and Lijphart’s characterisation of competitive versus consensus-oriented political systems.[71] Although a full and detailed exploration of these resonances is beyond the scope of a single paper, the main points of connection may readily be sketched.

As we have seen, the notion of proportionality generates in itself no concrete limits to punishment; hence the question of how much – and indeed how – to punish remains open to the sway of convention, political decision, or expediency.[72] Prevailing conventions; the institutional structures within which they develop, and within which penal policy is formulated; and the quality and intensity of social relationships, themselves premised in part on conventions and institutional structures, accordingly become crucial to the construction of meaningful limits on punishment. Liberal market countries, which have in recent years moved to increasingly flexibilised economies, do not invest heavily in their members’ skills. In these countries, labour market conditions typically produce a high proportion of short-term and insecure jobs and, significantly for our argument, their heavy reliance on monitoring to secure compliance is reflected in strikingly high levels of ‘guard labour’.[73] The conditions for establishing high future Associational Value – and for evaluating it with any confidence – would therefore seem to be weaker in liberal market economies than in coordinated market countries. For coordinated economies’ comparative advantage is built on long term investment in their members’ skills, an investment itself premised on the expectation of long term relationships of cooperation and mutual dependence between different groups. And these relationships help to foster intrinsic motivations for compliance with prevailing social norms (at least among insiders – an important caveat when it comes to penal policy).[74]

The coordinated market economies, moreover, typically feature the more generous welfare systems which, particularly in their Nordic, social democratic form, represent a recognition of mutual interdependence and belonging and express a culture of solidarity which fosters trust relations (and which is strongly correlated, as shown by comparative research, with lower levels of punishment).[75] These welfare systems, one might say, represent an institutionalisation of the collective expectation that Associational Value between citizens will – and indeed ought to – be high and widespread. Moreover, we might expect Associational Value to be weaker – and hence the appeal to proportionality (or other metaphors aspiring to limit punishment) a less effective constraint on retributive reactions – in the countries with highly competitive, first-past-the-post political systems, where policy horizons tend to be relatively short term and policy platforms accordingly volatile, than in the consensus-oriented political systems typical of the northern European and Nordic countries. For, in the latter countries, policy horizons tend to be longer term as interests within a proportionally representative system have to be bargained out in the process of coalition-formation: voters can accordingly have some confidence in the credibility of the policy platforms on which parties stand for office.[76] It follows that these political systems not only depend on, and set up incentives encouraging actors to, compromise, but also foster the sort of stability and group cohesion which underpins greater expected Associational Value. Their capacity to limit punishment, in short, is dependent not on abstract appeals to proportionality but on social and political cohesion which fosters high Associational Value between citizens and hence an orientation towards reconciliation – conditions which are met to a greater degree in the coordinated systems of northern Europe and the Nordic countries than in the Anglo-Saxon, liberal market countries.[77]

In other words, the psychological conditions which foster an orientation to reconciliation, through sustaining dense networks of mutual Associational Value, are better institutionally supported in coordinated market economies whose production regimes are premised on investment in long term relationships; in proportionally representative, consensus-oriented political systems in which there is a longer time frame for policy making; and in social democratic welfare systems which symbolise mutual dependency and which foster solidarity and relatively low levels of social inequality. To put this explicitly in terms of the Associational Value model set out above, a well coordinated labour market within a society in which mutual dependency and solidarity are also reflected in generous welfare state arrangements supports recognition of an aggressor’s work or other social productivity and expresses, at the level of the polity, an analogue of kinship relationships at the interpersonal level; the proportional political representation of all sectors creates both mutual dependency and stability which enhances expected Associational Value; and mutual dependence and a prevalence of long term relationships within both economic and social relations prima facie seems likely to enhance the scope for both expectation of and effective communication of remorse or repentance, so as to preserve those relationships. The perfect conditions for forgiveness and reconciliation, like Durkheim’s ideal ‘conscience collective’, exist nowhere.[78] But the institutional structure of different societies has a decisive impact on the extent to which they can be constructed, by affecting the opportunities and incentives of key actors such as judges, prosecutors, police officers, victims of crime – and indeed all of us who vote on criminal justice policy. The importance of trying to understand the nature of this institutional impact can hardly be exaggerated.

The question of what shapes the balance between retaliation and reconciliation has long been recognised by sociologists, anthropologists and criminologists as a crucial question about punishment. Even those who, like G. H. Mead, see desert as the basic reason for punishment, and the expression of vengeance and affective blame as socially useful, acknowledge that these generate no stable criteria for the level of punishment, and so inevitably carry the risks of the escalation of aggression and of the stigmatisation of offenders. As Mead put it: ‘We see society almost helpless in the grip of the hostile attitude it has taken towards those who break its laws and contravene its institutions’;[79] and, in a passage reminiscent of Durkheim; ‘hence … the attitude of hostility, either against the transgressor of the laws or against the external enemy, gives to the group a sense of solidarity … which consumes the differences of individual interests, the price paid for this solidarity of feeling is great and at times disastrous.’[80] In the sway of these powerful emotional dynamics – recently seen perhaps most vividly, among advanced democracies, in the United States – appeals to proportionality are little more than empty rhetoric. What is needed, rather, is an understanding of the institutional conditions which can structure the dynamics of collective action so as to inhibit the slide into stigmatisation and othering, and foster the reconciliatory dispositions.[81]

CONCLUSION

A certain form of retributivism, then, in the sense of a set of stable and effective criteria defining the fittingness of penalties by reference to a particular offence by a particular offender, makes most sense within the cosmology, or moral economy, in which state penalties are enacted within, and widely regarded as integral to, a political system rooted in a symbolically legitimated status hierarchy. Moreover the legitimation of those penalties may be underpinned in some important ways by theological symbolism and various forms of religious doctrine and belief. Early modern punishments look to us crude and extreme expressions of revenge; but their meaning to contemporaries – notably to the spectators at the scaffold or to the consumers of the forms of popular culture mentioned earlier – was importantly premised on an afterlife, on expiation, on a symbolism of penitence. And they resonated with moral sensibilities in a less violence-averse and a more overtly status-based hierarchical world than ours. But in our social and political world – a world no longer organised around a moral order structured in terms of symbolically anchored notions of desert or appropriateness – there is no agreed mechanism for anchoring the penalty scale according to cardinal proportionality, and actual penalty scales are driven by convention, calculations of consequences, and political dynamics. Particularly under conditions of a highly politicised climate for criminal justice policy-making, the commitment to just deserts all too easily produces insatiable demands for hard treatment. Desert theorists believed that retributivism could provide a limit to punishment by restricting its moral justification to the past-oriented criterion of proportionality with the seriousness of an offence. That its credentials on the first criterion are dismal is not surprising given that our prevailing culture is no longer premised on widely shared belief in a moral order structured around notions of desert or appropriateness grounded in widely shared symbols of equivalence.[82]

But this sobering analytic conclusion should not, in our view, be interpreted as a counsel of despair. Few proponents of the justice model would want to recreate the hierarchical and authoritarian conditions which made retributive punishments such as the lex talionis meaningful to contemporaries, and religious affiliations and symbols of the sacred cannot be organised in democratic societies by processes of deliberate institutional reform, even should they seem attractive. The task, rather, is to consider what early modern societies can tell us about the links between the legitimation of punishment and broader social conditions; to reflect on the theoretical and practical resources which resonate with modern moral and political discourse; and to ponder how best they may be institutionalised under particular conditions.

Both moral commonality and the sense of a real relationship between victim and offender as fellow members of a society have doubtless been attenuated by the increasing heterogeneity and moral pluralisation of social orders, especially in the more individualistic, competitive, liberal market countries. Those countries in which the retributive revival appears to have had the most baleful effect in eroding the institutional counterpart to reconciliatory dispositions in the criminal process are those in which inequality, conflict and heterogeneity are highest.[83] Conversely, scholarship on punishment in the Nordic countries associates their penal moderation with their relative homogeneity[84], their relatively small size and hence sense of social interdependence, solidarity, and trust, their consensus-oriented political systems, and their relative levels of equality – all factors which foster conditions for the institutional counterpart of forgivingness[85] and enhance expected Associational Value. The lesson for the more punitive – and less equal – liberal market countries is that avoiding polarisation and reducing inequality is likely to be as important as the effort to reconstruct the criminal process to effectively foster repentance and reconciliation; and that the key to penal moderation lies not only in reintegrative criminal justice policy, but in social policy and in political arrangements and institutional structures which maximise expected Associational Value among citizens.

In conclusion, it is important to clarify that our argument about the incapacity of the neoclassical revival to generate a robust sense of limits on punishment is not to be taken as suggesting that consequentialist theories of punishment are immune from analogous difficulties, let alone to make a case for penal consequentialism. (In fact, one of us is inclined to take a restricted consequentialist line on punishment,[86] but that argument is for another day.) Even leaving aside the well known distributive difficulties with purely consequentialist arguments for punishment,[87] it will be evident that while many of the potentially positive consequences of punishment – deterrence, incapacitation, reform and so on – are in principle measurable with the increasingly sophisticated tools of the social sciences, the question of how much of any of them adequately balances the infliction of a particular penalty remains an intractable moral question which cannot be reduced to any calculus, and one, moreover, on which views differ widely. In the realpolitik of penal practice, the key aspiration must be to determine the ethically optimal penal practices consistent with political legitimation. For modern consequentialists, as for modern retributivists, there is little alternative to the messy business of building political and social coalitions around agreed conventions specifying, and limiting, adequate penalties. Hence our knowledge, from comparative research, that the countries which have most successfully resisted the drift to penal severity since the 1970s are those whose social, economic and political institutions have given them maximum capacity to coordinate policy, in the public interest, over the long term is of key ethical significance.[88]

The value pluralism and heterogeneity of modern societies, sitting alongside our commitment to democratic politics, complicates the tasks of both legitimising punishment and institutionalising arrangements favourable to reconciliation within a stable set of symbols and practical arrangements. The task has been yet further complicated by the spread of criminal law into ever greater areas of regulation. The social task of legitimising and civilising punishment cannot be separated from that of legitimising and civilising the deployment of criminal law, and while our focus in this paper has been on punishment, we accept that, parallel to the argument which we have made about the institutional conditions under which appeals to penal proportionality can be made meaningful, we need to come to some understanding of the conditions for building support around institutions capable of delivering social coordination on criteria of fittingness or proportionality relevant to criminalisation. But whether we are focusing on punishment or on criminalisation, we should train our attention on the analysis and conditions of existence of institutions stabilising and implementing substantive criteria of fittingness, rather than placing our reformist faith in the chimera of appeals to proportionality.