Thinking about Punishment across Space and Time

This short essay shows why the customary two-categories of punishment theories (retributive theories and consequentialist theories) should be increased to three to include a-normative theory and how that would aid understanding of differences in punishment thinking and policy over time and across space. Section I lays out conventional typologies of punishment theories and explains why a-normative theory should be added. Section II introduces the general literature on determinants of punishment policies and practices. Section III briefly discusses Eastern and Central Europe. Section IV illustrates how the normative and explanatory theories when combined enrich explanations and improve understanding.

Understanding of punishment policies and practices in a single country or generally can be enriched by looking across disciplinary boundaries and by combining ideas and arguments from disciplines that are usually kept separate.Most of the modern literature on the determinants of penal policy is written by social scientists and generally focuses on a single country (e.g., Garland 2001;Pratt and Clark 2005;Simon 2006;Lacey 2008;Tonry 2009).A few works offer generalisations about many countries (Cavadino and Dignan 2006;Tonry 2007;Lappi-Seppälä 2008;Nelken 2010;Snacken and Dumortier 2011).Such work is informative as far as it goes, but it could go further if it took better note of writings by philosophers and historians.
The ways people think and talk about punishment vary between places and over time and inexorably influence punishment discourses and practices.The development of indeterminate sentencing in the United States and of "measures" (as distinct from punishments; de Keijser 2011) in much of the Western Europe in the early decades of the twentieth century, for example, were products of positivist ways of thinking that were influential among academic and policy elites on both continents (Pifferi 2011).
Policies and practices could be better linked to changing ideas if resort more often were made to work in other disciplines.Writings by philosophers and legal theorists on normative justifications of punishment almost always categorize punishment ideas and arguments under the two headings "retributivism" (sometimes deontology) and "consequentialism" (sometimes teleology and in earlier times usually "utilitarianism" or "positivism") (e.g., von Hirsch, Ashworth, and Roberts 2009;Tonry 2011b).Support for those two ways of thinking varies with time and space.Policies and practices vary with them.Writings by historians document wide changes over time in prevailing ways of thinking about the purposes and functions of sentencing and punishment (e.g., Rothman 1971;Tonry 2011a).
When the literatures on normative punishment theory, determinants of penal policies, and the history of punishment institutions and practices are combined, a number of usually invisible relationships leap out.Here, for illustration, are two.
Firstly, the customary partitioning of normative ideas about punishment between retributivist and consequentialist perspectives is insufficient.They provide frameworks for discussing and proposing what may justly be done to people who have been convicted of crimes but they are fundamentally incomplete.They do not and are unable to encompass ways of thinking about punishment in which just or appropriate treatment of individual offenders is not a consideration at all.
These might be called a-normative theories.They are influential in many countries in our time, particularly in the United States, England and Wales, and parts of Eastern Europe.They were prevalent in South Africa during Apartheid and in the Soviet Union, and continue to cast long shadows in both.They were predominant in the Western countries before the nineteenth century as is demonstrated by Foucault's (1977) accounts of punishment under the ancien regime and historians' accounts of the use of capital punishment in England (e.g., Thompson 1975;Hay et al. 1976) and imprisonment throughout Europe and Britain in the sixteenth through eighteenth centuries (Whitman 2003).
The two customary ways of thinking about punishment focus fully or partly on offenders.Retributive theories link notions of blameworthiness, culpability, or wrongfulness to what happens to offenders.Consequentialist theories, although by definition concerned with effects, usually impose limits on what may justly be done to offenders: for example, that the expected amount of harm done to punished offenders does not exceed the harm thereby averted for others (Frase 2009).A-normative ways of thinking about punishment take no or reduced account of considerations relating to just treatment of offenders (e.g., Garland 2001;Tonry 2004;Simon 2006).That is the only way that extraordinarily disproportionate punishments-life without possibility of parole for 13-year-old robbers or three-strikes sentences of 25-years-to-life for adult shoplifters in the United States, or indeterminate, potentially life-time imprisonment of "dangerous" offenders in England and Wales-can be understood.Likewise, it is the only plausible explanation for the exemplary use of pretrial confinement and the increased punishments imposed on offenders involved in England's urban riots of 2011 (e.g., Lewis 2011;The Guardian and LSE 2011).Normative considerations are not absent from these policies and practices, but the norms involved have little or nothing to do with justice toward offenders.They center instead on denunciation of wrongdoing, provision of reassurance to citizens, acknowledgment of popular senses of outrage and insecurity, and demonstration of government resolve.All of these goals, or functions, involve communication about norms related to wrongdoing, but none of them is principally concerned with offenders and their circumstances or interests.There are thus three, not two major normative frameworks for thinking about punishment practices, policies, and decisions.
Secondly, historical and cultural contingencies have much more powerful influence than is ordinarily recognised.A number of sub-literatures on the determinants of penal policies demonstrate the influence of structural features of countries' legal and social systems and of distinctive cultural and historical characteristics (e.g., Tonry 2007Tonry , 2009;;Lappi-Seppälä 2008).These are no doubt important, but so is serendipity-exemplified by Michel Pifferi's 2011 writing on the widely different influence of positivist criminological ideas on American and European legal systems in the opening decades of the twentieth century.
In America, the result was indeterminate sentencing that led to near abandonment of retributivist thinking and concerns for proportionality as a consideration in dealing with most offenders, and shaped justice system institutions and practices that survive today.In continental Europe, the result was creation of the distinction between sanctions and measures, with measures reserved mostly for a small number of "dangerous" offenders.Fundamental ideas about proportionality and the Rechtsstaat, the "equality principle, " continued to strongly influence the punishments most offenders received, but had little or no salience for the measures reserved for the "dangerous." Those two ideas-a-normative punishment thinking and the effects of contingency-largely explain why imprisonment rates were so high and sentencing practices so severe in the USSR and remain severe in Russia (enemies of the people and social parasites' interests deserve no consideration) and South Africa (the interests of "Kaffirs" don't matter) and why some countries in Eastern Europe-still staggering under the legacies and cynicism of the communist period -are having a hard time achieving policies more like those in the rest of Europe.The two ideas also explain why America's imprisonment rate was 5 times higher in 2012 than in 1973 and why it was the world's highest by a wide margin in 2012.American adoption of indeterminate sentencing for most imprisoned offenders meant that long nominal prison sentences-5, 10, 20 years, life -became common.As a practical matter they were largely symbolic; parole boards set release dates for most prisoners after they had served small fractions of their nominal sentences.However, when a-normative thinking emerged as a potent influence in the 1980s, the policies that resulted were phrased in terms of the prevailing nominal coinage for prison terms.Ten-or twenty-year sentences in the 1950s or 1960s often meant as little as one year in prison and seldom more than a few years.Those formerly nominal numbers were converted in recent decades into ten-and twenty-year sentences that many offenders actually served, moderated only by small reductions for good behavior (for serious offenses in many states, typically only 15 percent of the sentence).
This short essay shows why the customary two-categories of punishment theories should be increased to three to include a-normative theory and how that would aid understanding of differences in punishment thinking and policy over time and across space.Section I lays out conventional typologies of punishment theories and explains why a-normative theory should be added.Section II introduces the general literature on determinants of punishment policies and practices.Section III briefly discusses Eastern and Central Europe.Section IV illustrates how the normative and explanatory theories when combined enrich explanations and improve understanding.

I. NORMATIVE THEORIES
From the early nineteenth century until the 1970s, there was a sharp distinction between prevailing ideas about punishment in the United States and continental Europe.Americans were mostly consequentialists; members of the theory class would have described themselves as "utilitarians", although in European terms "positivists" would have been more accurate 1 .Theories, institutions, policies, and practices in the English-speaking countries were importantly influenced by consequentialist ideas, most dramatically in the United States and significantly elsewhere.Continental European systems were principally retributivist.European opinion leaders were influenced by positivist ideas generally and rehabilitation in particular, from the time of Franz von Liszt onwards, but they were much more influential in relation to the youth justice systems than to adult systems.For adult offenders convicted of moderately serious or serious crimes, ideas associated with the Rechtstaat and the related principle of equality in treatment of citizens by the state, confined positivist and rehabilitative ideas to the margins (Pifferi 2010).The fully or largely indeterminate sentence, for example, was adopted nowhere in Europe and discretionary prison release systems had only limited scope in reducing times served.Since the 1920s, the primary influence has been on programming within prisons and in community sentences, and in relation to indeterminate sentences imposed as measures, not sanctions, for small numbers of offenders, 1 Jeremy Bentham (1830Bentham ( /2011)), the most influential popularizer of utilitarian punishment ideas, wrote mostly about deterrent aims.American systems were concerned with deterrence but gave primary emphasis to rehabilitation and incapacitation, as Enrico Ferri (1921), the most influential popularizer of positivist punishment ideas in Europe, would have done.
usually people believed to be especially dangerous.In principle, a "measure" is not a punishment for crime but an act of public protection.As a result, sanctions for crime -punishments -must be proportioned to the offender's wrongdoing, but measures need not be (de Keijser 2011)2 .Of course there were differences between countries and within countries over time, but the idea that the seriousness of the crime should be the primary basis of punishment or set the upper limits of punishment was influential and endured-as it has to this day.
In the United States, the academic and real worlds lined up nicely.Practitioners and policy makers may not have read Cesare Beccaria (1764), Jeremy Bentham (1830/2011), or Enrico Ferri (1921), or known who they were, but they were in broad agreement with them that the primary purpose of punishment is to minimize harms associated with crime and state responses to it.Most of the institutions that comprise contemporary criminal justice systems -prisons, training schools, reformatories, juvenile courts, probation, parole -were invented in the nineteenth century and premised on the pursuit of that purpose.So were individualised and indeterminate sentencing systems for dealing with adult offenders and the parens patrie rationale that underlies the juvenile court (e.g., Platt 1969;Rothman 1971;Mennel 1984).
Near the end of that 150-year period, the Model Penal Code (1962) laid out a blueprint for the mother of all consequentialist punishment systems 3 .Offenses were defined broadly and were categorised only into misdemeanors and three levels of felonies.Precise delineation of the seriousness of crimes was considered unimportant and unnecessary.The only important question was whether the defendant was guilty.Once that was determined, the judge was given broad discretion to decide what sentence to impose.Probation was available for any offense, including murder.If the judge believed that the sentences authorised for a crime were too severe, he or she could sentence the offender as if he had been convicted of something less serious.If a prison sentence was ordered, the parole board decided when the prisoner was released.The prison authorities could award and withdraw time off for good behaviour.Consistent with the utilitarian principle of parsimony, presumptions were created to ensure that offenders were not punished more severely than was necessary: judges were directed not to send people to prison, and parole boards were directed to release inmates when they first became eligible, unless specified conditions existed to justify some other decision.Allusions to retributive ideas appear only three times, and faintly.Two were provisions indicating that non-incarcerative penalties should not be imposed, or inmates released on parole, if doing so would "unduly depreciate the seriousness of the offense." The third was a provision that one of the overall objectives of the code was to ensure that disproportionately severe punishments were not imposed.In the specialized vocabulary of punishment philosophy, this is an avowal of "negative" retributivism: an offender may be punished in an amount proportioned to his blameworthiness (but not more), but need not be.This in contrast to "positive" retributivism, the contemporary view of Andrew von Hirsch (1976Hirsch ( , 1994)), Paul Robinson (1987Robinson ( , 2008)), and many other contemporary writers, and the nineteenth century views of Immanuel Kant (1999) and George Wilhelm Friedrich Hegel (1991), that offenders not only may be punished as much as they deserve, but must be.
Stop for a minute and think about that: Retributive ideas were almost absent from the most influential American criminal law document of the twentieth century.The code was developed under the aegis of the American Law Institute, then and now the most prestigious law reform organisation in the United States.Lawyers then could not, and now cannot, simply join.They must be nominated by current members and approved by a membership committee; admission is widely considered a great honour, the capstone of a successful career, and ostensibly limited to people of great accomplishment.The code was developed over a 13-year period by a group of influential practitioners, including judges, prison commissioners, prosecutors, parole board heads, and defense lawyers.Work was directed by Herbert Wechsler, the century's leading American academic criminal law scholar.Draft versions of sections were successively considered -and provisionally and then finally approved -at annual meetings of the American Law Institute.The voters were not primarily criminal lawyers or academics, but mostly commercial and business lawyers and federal and state judges.The Model Penal Code was drafted and approved by people who were neither radicals nor woolly-headed intellectuals.They were the most establishmentarian lawyers midcentury America had to offer, together with a group of leading criminal justice practitioners, and they did not believe that retributivism and proportionality should be central considerations in sentencing.
In our time, in contrast, retributive ideas seem an inherent part of popular and academic thinking about crime and punishment, even if they have had little recent influence on policy except in the vindictive sense that policy makers have generally preferred harsher punishments to milder ones.Social and experimental psychologists instruct that human beings are hardwired to react punitively to crime (Darley 2010).Evolutionary psychologists explain that natural selection has favoured human beings with that hard wiring.Individuals with clear senses of right and wrong and a willingness to act on them, it is said, are better community members, fostering cohesion, increasing the odds of community survival, and perpetuating the gene pool that predisposed people to be retributive (Robinson, Kurzban, and Jones 2007).Some influential philosophers of criminal law argue that those punitive intuitions justify retributive punishment theories (e.g., Moore 1993).
If retributive ideas and instincts are so common, how can it be that they had so little influence on the Model Penal Code?The answer is that rightthinking people in the 1930s through the 1960s believed that retributivism was atavistic.Commonly held intuitions and widely shared beliefs can be morally or ethically wrong, as is demonstrated by conventional views in earlier times about racial inferiority, homosexuality, and gender roles.That is what our midcentury predecessors believed about retributive instincts.They were wrong, should be acknowledged to be wrong, and should be ignored.Herbert Wechsler, and his mentor Jerome Michael, observed that retribution may represent "the unstudied belief of most men" but concluded, "no legal provision can be justified merely because it calls for the punishment of the morally guilty by penalties proportioned to their guilt, or criticized merely because it fails to do so" (Michael and Wechsler 1940, pp. 7, 11).A few years earlier Jerome Michael and University of Chicago philosophy professor Mortimer Adler explained that there are two incompatible theories of punishment: the "punitive" (retributive) and the "non-punitive" (consequentialist) and that "it can be shown that the punitive theory is a fallacious analysis and that the non-punitive theory is correct. . . .The infliction of pain is never justified merely on the ground that it visits retributive punishment upon the offender.Punitive retribution is never justifiable in itself " (Michael and Adler 1933, pp. 341, 344).The conventional ways of thinking, the zeitgeist, the prevailing sensibilities, rejected retribution and favoured rehabilitation and, to a lesser extent, deterrence and incapacitation.Indeterminate sentencing reflected that thinking.In its most extreme forms, in California and Washington State, judges did not set prison sentences; they merely sent people to prison … for a term limited in length only by the maximum specified in the applicable statute.A parole board decided when they were released (Rothman 1971).
Sensibilities, however, were changing while the Code was being drafted.The American Law Institute's timing in releasing the Code in 1962 could not have been worse.Harbingers of discontent with penal consequentialism had already begun to appear (e.g., Lewis [1949Lewis [ /2011]]; Allen 1959) and recurred with increasing frequency (e.g., Burgess 1962;Allen 1964;Davis 1969).By the mid-1970s, dissatisfaction was widespread.Policy makers rejected many features of indeterminate sentencing and favoured new approaches based on retributive ideas.Individualised, indeterminate sentencing was out.Retributive, determinate sentencing was in (Messinger and Johnson 1978).
Whatever it was that changed policy makers' and practitioners' minds also influenced theorists.Consequentialism lost ground and influence.Retributivism came into vogue.In the 1950s, Norval Morris (1953), John Rawls (1955), and H. L. A. Hart (1959) attempted to reconcile general utilitarian rationales for punishment as an institution with the salience of retributive ideas in making decisions in individual cases.Herbert Morris (1966) and Jeffry Murphy (1973) offered benefits-and-burdens theories, which, a bit obscurely, argued that the gravamen of crime is obtaining unfair benefit from others' law-abidingness and that punishment should balance things out.John Kleinig (1973), in the first book about punishment with "desert" in its title, assessed the relevance of desert considerations to the justification of punishment.Joel Feinberg (1970) and Jean Hampton (1984) argued for expressive theories and Herbert Morris (1981) and Antony Duff (1986) for communicative ones.Michael Moore (1993) offered an intuitionist account of punishment.Among the criminal lawyers, Norval Morris (1974) elaborated his theory of limiting retributivism, Alan Dershowitz (1976) his of "fair and certain punishment, " and Andrew von Hirsch (1976) his of "just deserts." University of Chicago law professor Albert Alschuler bewilderedly described the sea change: "That I and many other academics adhered in large part to [a] reformative viewpoint only a decade or so ago seems almost incredible to most of us today" (Alschuler 1978, p. 552).By the early 1980s, it was not unreasonable to believe that a corner had been turned and that policy makers, practitioners, and theorists would long march to the beat of distant retributive drums.
That did not happen, except for a few years.In the late 1970s and the early 1980s some legislatures' enacted statutes meant to encourage proportionate sentences and abolished parole release in order to ensure that offenders served the proportionate sentences they received.Sentencing commissions adopted guidelines based on retributive premises.The rhetoric of deserved punishment and desert entered the political lexicon in the United States, and shortly thereafter in Australia, Canada, and England and Wales (Blumstein et al. 1983, chap. 1;von Hirsch, Knapp, and Tonry 1986).
In the world of policy-making, the retributive moment quickly passed.Except in lip service, proportionality largely disappeared as an important goal.Many of the sentencing laws enacted in the United States in the 1980s and 1990s, including mandatory minimum, three strikes, truth in sentencing, and life-without-possibility-of-parole laws (LWOPs), paid no heed to proportionality.Drug laws mandated sentences for street-level dealers longer than sentences typically received by people convicted of serious assaults, robberies, rapes, and many homicides.Three-strikes laws mandated lengthy and life sentences for repeat property and drug offenders.LWOP laws meant what they said and by 2008 more than 42,000 people had been sentenced to prison under them (Nellis and King 2009).If principled rationales were implied by developments such as these, the principles were consequentialistdeterrence by means of threats of harsh punishment, incapacitation by means of lengthy periods of confinement, and moral education by means of the messages severe punishments ostensibly convey about right and wrong.
New, less overtly punitive initiatives also paid little heed to proportionality.Drug courts and other problem-solving courts targeting mentally ill offenders, domestic violence, and gun crimes began in the early 1990s.By 2010 they numbered in the thousands (Mitchell 2011).Drug courts are predicated on the beliefs that drug treatment can work, that drug dependence is causally related to offending, and that coerced treatment backed up by firm judicial monitoring can break drug dependence.Other problem-solving courts are based on parallel logic.Proponents of problem-solving courts regularly announce that they are influenced by ideas about therapeutic jurisprudence, a school of thought that urges incorporation of therapeutic ideas into legal doctrines and processes (Wexler 1995(Wexler , 2008a(Wexler , 2008b)).
Other proportionality-defying approaches proliferated.Prison reentry programs predicated on risk prediction and rehabilitative programming proliferated.Throughout corrections systems, increased investments were made in cognitive-skills, drug-abuse, sexual offending, and other treatment programs.Throughout the world, including in the United States, thousands of new restorative justice programs were established.All of these initiatives shared the characteristics that their primary aims were forward lookingreduce re-offending or drug use, solve problems, restore relations among offenders, victims, and communities -and not much concerned to apportion punishment to culpability or blameworthiness (Tonry 2011a).
As the twenty-first century began, and still today, most American theorists subscribe to principled retributive ideas that had currency and influence in a former time.Many American punishment policies and practices both those that are exorbitantly severe and those that are enthusiastically individualised are inconsistent with those ideas.

II. DETERMINANTS OF PENAL POLICY
The explanations for why countries have particular penal policies and practices are general and particular.They are general in the sense that some structural features of government and society such as income inequality, trust, legitimacy, and consensual political systems seem to be associated with particular kinds of crime control policies and punishment practices.They are particular in the sense that the details of national history and culture powerfully affect both those structural features and their effects.
Because of the extraordinary 500 percent increase in American imprisonment after 1973, much of the literature has focused on the U.S. The United States is the only European or North American country that retains the death penalty.It has the world's highest imprisonment rate, exceeding 750 per 100,000 residents, and regularly imposes LWOPs, the developed world's harshest criminal punishments short of death.Unlike most European countries and Canada, it refuses to acknowledge the moral force of international human rights conventions and declarations, or to incorporate them into its laws.Fifty years ago, United States governments were world leaders in promoting human rights values, the imprisonment rate was stable and in line with those of other Western democracies, capital punishment was falling into disuse, and the U.S. Supreme Court was admired throughout the world for its extension of human rights protections to criminal suspects.
Something went terribly wrong.In the simplest sense, we know what happened and why.Between 1975 and 1995, policymakers enacted a wide range of laws meant to make punishments severer.These included threestrikes-and-you're-out laws requiring minimum 25-year sentences; 10-, 20-, and 30-year minimum sentences for violent, firearms, and drug offenses; LWOPs; laws permitting or requiring prosecutions of tens of thousands of young people each year as adults; and laws extending the reach of capital punishment.During the same period, practitioners became more punitive and risk-averse: prosecutors charged and bargained more aggressively, judges sent more people to prison and for longer, parole boards released fewer prisoners, and later, returned parolees to prison more often (e.g., Blumstein and Beck 1999).
What we don't know is why American policymakers, nearly alone among leaders of Western governments, chose to enact such harsh policies and laws, or why practitioners became so much tougher.Crime rates and trends are not the answer.Crime rate trends that were much the same throughout the Western world after 1970s -rises through the early 1990s and declines since (e.g., Tonry 2004; van Dijk, van Kesteren, and Smit 2007;Eisner 2008).Imprisonment rates and penal policies however have differed enormously.In the United States, crime rates have fallen steadily since 1991 and imprisonment rates have risen every year, more than doubling since 1991 (Tonry 2004).In the rest of the developed world, there was no common pattern: increases in some, decreases in others, and stability in most.
Nor is public opinion the answer.In the English-speaking countries at least, penal policies and imprisonment rates vary enormously but public opinion is much the same.Majorities of the public believe crime rates are rising when they are falling.Large majorities believe judges are too lenient, on the basis of mistaken beliefs about the severity of punishments.The sentences citizens say they believe are appropriate are typically less severe than the sentences judges actually impose.When citizens are asked whether they prefer more punitive policies or increased investment in rehabilitative programs, majorities usually prefer the latter (Roberts et al. 2002).
Three sets of more complicated answers have been offered.The first, by David Garland (2001) attributes toughened penal policies in England and America to conditions of "late modernity." These include increased insecurities associated with globalisation and rapid social change, increased vulnerability of privileged segments of the population to victimisation by crime, the limited capacities of governments to affect crime rates, and increasing population diversity that feeds a "criminology of the other." The result is a proliferation of "expressive" policies meant more to show that government is doing something to reassure the public, and to aid politicians' efforts to win re-election, than to reduce crime.
The insuperable difficulty for Garland's analysis is that, if it is right, it should explain why all Western countries experienced steeply rising imprisonment rates and steadily harshening penal policies.All the "causal" developments he describes -increasing diversity, globalisation, and existential angsthappened everywhere.Imprisonment rates and policy trends, however, diverged dramatically.Crime rates rose steeply in most Western countries between 1970s and the early 1990s but only in a few did policymakers adopt policies intended to make punishments harsher or imprisonment rates higher.
Imprisonment rates increased five-and six-fold in the United States and the Netherlands after 1973 and doubled in England and Spain after 1993.Through most of that period, Japanese imprisonment rates fell.Finnish policymakers chose to make the country's penal policies less harsh and to reduce its imprisonment rates.Canada, and most European countries -the rest of Scandinavia, Belgium, France, Italy, Germany, Austria -did not adopt harsh new broad-based policies and in diverse ways worked to keep their imprisonment rates stable.Only in a few Western countries -England, Portugal, Spain, New Zealand -and only after crime rates began to fall in the 1990s did imprisonment rates rise significantly, but to levels at their peaks less than one-fourth the current American rate (Tonry 2004, chap. 5;Tonry 2007;Lappi-Seppälä 2008).
Michael Cavadino and James Dignan (2006) offer a second analytical framework.A three-part typology (technically four, but one contains only a single country, Japan) links imprisonment trends and penal policy differences with systems of political economy.Neo-liberal states (e.g., England, New Zealand, the United States) are said to have the harshest policies and social democratic corporatist states (e.g., Scandinavia) the mildest, with conservative corporatist states (e.g., Germany, France) falling in-between.
This analysis made sense, sort of, early in the 21st century but in a larger picture it is not informative.There are at least three kinds of anomalies it cannot explain.Firstly, even in 2006, some countries did not fit their categories.The Netherlands then had the highest imprisonment rate in Western Europe except for England and Wales, but was not counted as "neo-liberal" while Australia, which had a rate only slightly higher than Germany's, was.Secondly, countries excluded from the analysis defied the categories.Canada by all accounts should be included among the English-speaking neo-liberals but its imprisonment rate has held steady around 100 per 100,000 population for a half century, so it should have been included among the conservative corporatists with France and Germany.Spain and Portugal, also excluded, have had among the highest imprisonment rates in Western Europe at various times since the mid-1990s and no one would classify them as neo-liberal (rates have since fallen substantially in Portugal).Third, the typology collapses in other periods.In 1970, for example, the Dutch imprisonment rate was the lowest among the developed countries and the U.S. rate was in the mainstream.The Finnish was the highest.That would have made the Finns neo-liberal, the Americans conservative corporatists, and the Dutch, quite properly, social democrats.
A third set of analysis attempts to explain penal policies in terms of such things as income inequality, citizens' perceptions of the legitimacy of governmental institutions, citizens' trust in each other and in government, the strength of the welfare state, and the structure of government.All these things seem to matter.Moderate penal policies and low imprisonment rates are associated with low levels of income inequality, high levels of trust and legitimacy, strong welfare states, professionalised as opposed to politicised criminal justice systems, and consensual rather than conflictual political cultures (Downes and Hansen 2006;Lappi-Seppälä 2007, 2008;Tonry 2007;Di Tella and Dubra 2008).
Even that analysis, however, does not explain American developments.For each of those factors, the United States falls at the wrong end, the end associated with more punitive policies and practices.We need to understand why.
The story of American penal policy since 1973 is thus not about globalisation, neo-liberalism, or conditions of late modernity any more than it is about rising crime rates or harsher public attitudes.The simplest explanation for why American penal policies are so harsh is that American voters elected candidates who said they would make them that harsh, and did.Politicians took advantage of recurring features of American political culture in order to win elections and wield power.That explanation, however, is too simple because it cannot explain why Americans believed that serious problems of crime and disorder are amenable to easy repressive solutions, and accordingly why they voted as they did.
And that explanation can be found only in cultural and political values, which evolve over time and are shaped by history and experience.In the United States, four factors -a tradition of populist extremism, Protestant fundamentalism, governmental structure, and a history of racial conflict -go a long way to explaining why penal policies evolved as they did in the final quarter of the twentieth century (Tonry 2011c).However, they point to a need for yet another level of explanation.
Why have the populist extremism and moralistic Protestant intolerance recurred again and again, and why has the influence of tortured race relations remained so powerful?Big ideas about American history, including the Puritanism and intolerance of the first settlers, ideals of individualism and libertarianism associated with the frontier, and the early slavery-based southern economy, no doubt need to be woven into the answers.
Research agendas in coming years should focus on this level of explanation.The questions in other countries are different, and better questions will enrich understanding of why countries respond to crimes and criminals in the different ways they do.In France, for example, why is it that French men and women have for centuries accepted the legitimacy of broadly based amnesties and pardons of offenders and prisoners?In Italy, a mass commutation in 2006 reduced the prison population by 40 percent (and a smaller mass commutation occurred in 2008 under a conservative Berlusconi government), and there were no mass outcries.My instinct (and guess) is that similar policy decisions in the United States, England, or the Netherlands would provoke political firestorms.Similarly, something in the Scandinavian and German histories and cultures resulted in adoption of restrained penal policies for at least half a century.Explanations can be offered in terms provided aboveconsensus political systems, high levels of legitimacy and trust, and so onbut they beg the question.What is it about the (very different) histories of those countries that produce those characteristics?

III. THE SPECIAL CASE OF EASTERN AND CENTRAL EUROPE
As specialist literatures go, the body of work attempting to explain national differences in imprisonment rates is not large, but it has grown rapidly.To my knowledge, central and eastern European countries that were Warsaw Pact members of Soviet Socialist Republics have received comparatively little attention (except, notably, Krajewski 2010).This is striking because in the 1970s and 1980s the Warsaw Pact nations and the Soviet republics had very high imprisonment rates in comparison with Western Europe.For many but not all, that pattern continues.As Krajewski observes, "Europe remains visibly split into two different 'penal climates'" (2010, p. 7).
"Why, " Krajewski asks, "do high imprisonment rates persist in the region twenty years after the fall of the Berlin wall?" (2010, p. 3).In searching for an answer, he draws on Lappi-Seppälä (2008), who identifies five factors that appear to explain high imprisonment rates and harsh policies: high levels of income inequality, weak social welfare systems, low levels of trust in government, low levels of government legitimacy in the eyes of citizens, and conflictual politics and policy processes.
Krajewski notes that eastern and central European countries are characterized by high income inequality and weak social welfare systems, but devotes most of his attention to the other three factors.Political legitimacy, he argues, may be crucial.Most countries in the region were for long periods dominated by foreign powers.This diminished trust toward state authority of any kind: "This was especially conspicuous in the case of communist authorities which lacked political legitimacy among the vast majority of the population.This was probably one reason why communist crime control policies were so punitive; they were not necessarily rational, penological tools, but socio-technical instruments" (Krajewski 2010, p. 7).
Lappi-Seppälä showed that high levels of governmental legitimacy in the eyes of citizens are strongly associated with mild crime control policies and low imprisonment rates.One source of legitimacy of criminal justice systems is that judges and prosecutors are career civil servants who are well-insulated from the political pressures that in various ways influence American and British officials.While this is formally true in Eastern and Central Europe, Krajewski observes that it may be less true than it appears and that citizens' memories of the communist period likely make them skeptical.
The criminal justice system was of special importance to the communist party governments.Political loyalty constituted a primary criterion for professional career development.This most directly affected public prosecutors.Judges were affected to a lesser extent (even at that time they nominally independent).The public prosecution system was (and often still is) organised according to Soviet principles, and constituted a rigid hierarchy in which lower-rank prosecutors had little autonomy.In such a situation prosecutors usually tried to avoid decisions that could bring them into conflict with their superiors.After 1989 there usually were not general purges of public prosecutors or judges.In consequence old habits remained, and despite the new environment, appeared to be transmitted to younger generations, although without any special ideological element (Krajewski 2010, pp. 8-9).
Comparative political scientists have long observed that governmental policies tend to be more liberal and humane in countries that have consensual-as contrasted with conflictual-political systems (Lijphart 1999).Consensual political systems are usually characterised by more than two major political parties, coalition governments, proportional representation, and multi-seat electoral districts.Major policy decisions are based on broad consultation within the coalition government and outside it.Major decisions are seldom changed abruptly, even after new governments are elected.New coalitions are likely to contain parties from the old one, and parties newly in power often were involved in development of policies of the former government.Most western European political systems are more consensual than conflictual.
Krajewski observes that most of the former communist states fit the form but not the substance of consensual political systems.There are generally more than two major political parties and proportional representation electoral systems, but that does not produce consensual policy-making processes as in most of Eastern and Central Europe: Countries in Central and Eastern Europe tend not to have orderly political climates as a consequence of lacking an established system of political parties.
In principle, all have multiparty political systems which usually require formations of political coalitions to form governments.This should in theory create pressures for a consensus-oriented political system.
But the opposite seems to be the case.Political systems in the region are permeated by intense conflicts.Conflict between post-communists and political parties which emerged from dissident and opposition movements is one of the most important.As a result, political compromise is not necessarily common."This probably fosters a political dynamic closer to the conflict political culture model" (Krajewski 2010, p. 8).
"The 'penal climate' in central and eastern Europe, " Krajewski observes, "thus seems still, despite all the 'European influences' to which the region has been subject, to be shaped by its communist inheritance… In the area of crime control, the influence of this communist past, which was so profound in many other areas of political, economic, and social life in the region, is especially resistant to change" (2010, p. 10).
Imprisonment rates and patterns vary among the former socialist states.Some like those in Hungary and Poland have trended downward toward Western European levels (and Slovenia to sub-Scandinavian levels).Others, such as in the Baltic countries, remain high and close to both former Soviet and contemporary Russian levels.Nonetheless, the former Soviet view of prisoners as class enemies and social enemies is likely for a good while yet to engender a fundamental lack of sympathy with the interests and experiences of prisoners.
IV. THE FUTURE Cultural and political values are not serendipitous.They evolve over time and are shaped by history and experience.In the United States, three factors -the Evangelical Protestant fundamentalism, the conflictual political system, and the history of race relations -go a long way to explaining why penal policies evolved as they did in the final quarter of the twentieth century.In eastern and central Europe, the lingering effects of communist rule, in concert with longer-term features of the distinctive histories and traditional cultures of individual countries, are important backdrops to contemporary policies and practices.Different historical and cultural characteristics are likely to explain humane and moderate punishment policies and attitudes in the Scandinavian and German-speaking countries.
Normative ideas interact with government policies even if they do not fully explain them and even if they sometimes, as in the contemporary United States, seem largely incompatible.Two things seem clear.The traditional division of normative idea and theories into retributive and consequentialist categories is incomplete.The ideas encompassed in retributivism and consequentialism are ultimately concerned to explain and justify what happens to convicted offenders in terms that relate to them as individual human beings and that to some considerable extent take account of their interests and autonomy.In many countries at diverse times, however, punishment policies take little account of offenders' interests.Prominent contemporary examples include the United States and to a lesser extent England and Wales, South Africa, the former socialist states of Eastern and Central Europe, and Russia.
What all those countries share is histories characterised by political cultures in which the interests of some people do or did not count -black people in the United States and South Africa, class enemies in Russia and the other former socialist states, and the working class in England and Wales.The last example is the most tentatively offered; it is hard not to see the common characterizations of young working-class offenders by English politicians and newspapers as louts, yobs, and scum as signifying a social category of lowerclass people who have not accepted their inferior statuses and the need to behave themselves as their betters would wish.All of these social categories of people are the targets of laws, policies, and patterns of punishment that suggest they are seen primarily as social threats and not as people whose interests deserve the concern and respect that traditional retributive and consequentialist theories of punishment would give them.
The second influence is contingency.The adoption of indeterminate sentencing in the United States created institutions and conventions that were especially vulnerable to a radical change in prevailing ways of thinking about crime and criminals.
These two considerations go a long way toward explaining why American punishment policies became so much more severe and imprisonment rates rose after the early 1970s.Before then, criminal justice policy was not highly politicised and imprisonment rates had been stable for a half century (Blumstein and Cohen 1973).After that, however, conservative Republicans for three decades pursued what became known as the "Republican Southern Strategy" of focusing their campaigns and strategies of governing on issues such as welfare fraud and "law and order" which everyone knew to be veiled appeals to racial resentments of white southern and working-class voters (Edsall and Edsall 1991;Tonry 1995;Gottschalk 2006).
During a time of rising crime rates, conservative politicians were in any case attracted to adoption of policies aimed at increasing the severity of punishment.Partly they sought to contrast their punitive approaches to the more social welfare-oriented and rehabilitative approaches of many Democrats.The Southern Strategy provided a more cynical additional logic for a political emphasis on law and order.Democrats eventually adopted a strategy associated with President Bill Clinton of "not letting the Republicans get to his right on crime." After that, politicians in both parties competed to show who was tougher.
Tough ideological anti-crime policies have so far been adopted in a few other countries.Examples include England and Wales during New Labour's period of governance under Tony Blair, a handful of Australian states and Canadian provinces at various times, and in the past few years the Canadian national government of Conservative Party prime minister Stephen Harper.
In all those places, a-normative punishment ideas, unconcerned about justice to or the interests of offenders took hold.The effects, however, have been much more pronounced in the United States.Partly this is because the period of dominance of a-normative ideas has lasted longer in the U.S. than elsewhere.Partly it is because of the history of American race relations.Partly, however, it is because of the historical contingency of enthusiastic adoption in the 1920s and afterwards of indeterminate sentencing and full-throated consequentialism.When these lost support in the 1970s, the apparatus of indeterminate sentencing-unlimited discretion for practitioners and a tradition of "bark-and-bite" sentencing in which sentences seemed much harsher than they really were-remained.A generation of policy-makers who were indifferent to the interests of offenders, and multiply motivated to be harsh, readily adopted the old nominal numbers.Prison sentences measured in terms of decades and lifetimes, and an imprisonment rate approaching 800 per 100,000 population, are the results.
Changing the American situation will not be easy because changes will be required both in ways of thinking and in the fundamental architecture of the American punishment system.For so long as a-normative thinking remains predominant, and routine imposition of decades-long prison sentences remains acceptable, not much is likely to change.
The situation in other parts of the world now characterized by a-normative thinking may be different.Russia, South Africa, and some Eastern European countries have reduced their imprisonment rates in the past 20 years, partly because elites believe that is a humane and civilized thing to do.The current coalition government of England and Wales is dismantling many of the worst features of its predecessors' criminal justice policies.The durability of the current Conservative government of Canada remains to be seen.For all these places, the task of reinstituting more humane policies may be less daunting than in the United States because they lack the architecture and ways of thinking associated with indeterminate sentencing.