Unequal Punishment: Universal Observations and Hidden Mechanisms

That ‘the rich get richer and the poor get prison’ (Reiman 1997) seems to be a well-established criminological wisdom. In fact, prisons across the globe are manifestations of inequality. In all nations, the most marginalised groups of society are over-represented in prisons, and in all other institutions of the criminal justice system. In the United States, African Americans are imprisoned at extraordinary rates, far exceeding any other group of the population, while Australia has the worldwide highest imprisonment rate of its indigenous population. In most western European countries, immigrants are over-represented in the criminal justice and correctional system. Besides racial and ethnic minorities, prisons worldwide house disproportionately and nearly exclusively individuals who count among the least educated, most unemployed, and poorest groups of society, many of them having spent their childhood and adolescence in disadvantaged families and neighbourhoods (see, e. g., Western 2006 for the US case).

The criminal justice system obviously transmits, manifests, and solidifies social inequality as it targets the poor and disadvantaged. Criminal justice reveals itself as a major driver of ‘injustice’, squarely contradicting democratic values and rule of law principles of equality before the law. However, from an evolutionary perspective, punishment and punitive intuitions should take into account the standing and importance of both victims and perpetrators when assessing and meting out punishment for the offence (Eisner et al. 2017). Historically, criminal law in many state-organised societies graded punishments by the social status of the perpetrator and the victim, and high-ranking perpetrators could only be charged before a court of their equals or were held in special prisons (e.g. in the Bastille in the Ancien (p.23) Régime in France). According to Whitman (2003), these terms of inequality and the status of prisoners have shaped contemporary criminal justice and punishment until today, and they also account for the huge discrepancies between the USA and Europe. While in the egalitarian society of the United States, criminal punishment was standardised against the most powerless and marginalised social groups, whether slaves or newly arrived immigrants, in the hierarchical societies of Europe, punishment was standardised against the high-status political prisoners and the prison regime to which they were subjected.

However, while it is obvious that the criminal justice system targets marginalised and disadvantaged populations, the link between social inequality and inequality of punishment has been much more elusive, and the process of transmission of one type of inequality into another much more inconclusive, as will be briefly shown below. Besides the crucial question of the relationship between social inequality and crime (on which see, e.g., Nivette 2011; Western 2006) and its impact on imprisonment, two general questions are involved when dissecting the link between inequality and punishment. First, in what way is the unequal representation of social groups and the over-representation of marginalised groups within the criminal justice system related to social inequality more generally? And second, how are the manifestations of punishment–extending beyond the fact of imprisonment to include, for example, repressive justice and conditions inside prisons–shaped by inequality more broadly? While the first question addresses the processes and mechanisms through which social inequality is transmitted into unequal criminal punishment, the second question addresses the social forces of inequality that shape the ways in which justice is administered and punishment executed in a given society. While the first question aims at analysing why the ‘poor get prison’, the second question links structures and values of inequality to overall characteristics of punishment and justice. Addressing both questions simultaneously has led to a number of problems, including both over-and under-theorising the crucial links between inequality and punishment, and over-rating the concomitant change of social inequality and punishment. There has been considerable neglect of the multifaceted nature of both social inequality and punishment.

This chapter seeks to address the second question, first givings an overview of pertinent research and outlining a number of conceptual problems. It then presents analyses of the relationship between socio-economic inequality and different dimensions of criminal justice, including imprisonment rates, conditions in prisons, and repression and human rights violations at the hands of the criminal justice system.

Theorising Inequality and Punishment

Theorising the relationship between inequality and punishment starts from the basic fact that structural inequality determines the economic, social, political, and moral life of any society, and shapes the lives of its citizens. As Wilkinson and Pickett (2009; 2018) demonstrate for richer countries, higher levels of inequality (p.24) negatively affect all spheres of life, from individual health and well-being to collective well-being in communities and neighbourhoods. The uneven distribution of social and economic power translates into political power and advantages of powerful groups and elites in determining policies. In contemporary nation-states, criminal justice is a genuinely political enterprise, and as such determined by the political system and its actors. Economic and social power and related interests of social groups thus shape the criminal justice system. These are not necessarily the richest or elite social groups; rather, in contemporary societies, they are often the middle classes who act and react as voters and citizens, demanding protection against crime and punitive sanctions against offenders (Garland 2001; see also, for Latin America, Karstedt 2020; Bergman 2018).

It is for these reasons that the political economy of punishment has been an influential framework ever since Rusche and Kirchheimer’s ( 2003[1939]) path-breaking study. Presently, as inequality is rising in the advanced economies of the global north, the political economy of punishment focuses on neoliberal economic policies that took hold in these countries from the 1980s (De Giorgi 2006; 2018; Farrall & Jennings 2014; Lacey 2008; for an overview, see Lacey et al. 2018). The framework of neoliberal policies and their impact also provided a tool for understanding the links between inequality, crime, and punishment in the emerging economies of Latin America (Snodgrass Godoy 2005; 2006; Sozzo 2017). The original idea of the political economy of punishment that criminal justice institutions control economically marginalised and potentially unruly populations was equally foundational for the new political economy of punishment. In the mass incarceration context of the United States, the ‘neoliberal government of social insecurity’ disciplines and ‘punishes the poor’ through the means of criminal justice, which itself involves substantive economic interests (through the economy of rural prisons in particular) (Wacquant 2004).

The contemporary political economy of punishment cannot ignore the role of established and emerging middle classes in shaping criminal justice policies. In particular, the values and attitudes of these groups, and their reactions to socioeconomic and structural inequality, have been seen as powerful determinants of criminal justice policies and drivers of punitiveness. From the ‘moral indignation’ of the struggling middle classes, diagnosed by Sven Ranulf in 1938, to ‘penal populism’ in the first decade of the 21st^t century (Pratt 2007), the ideological correlates of socio-economic and structural inequality and sweeping socio-economic changes affecting the middle classes have been identified as a centrepiece of the political economy of punishment. It is, in particular, social and economic insecurity and the fear of downward mobility that feeds into penal populism and the demands from middle-class voters that politicians be ‘tough on crime’. Whether middle classes are identified as the new powerful groups or not, this shifts the analytical weight from the economically and politically powerful at the top to the core and centre of society (but see also Miller 2016). It also changes the nature of social groups that are perceived as a ‘social threat’, as the view from the top of the social pyramid on ‘unruly classes’ below is replaced by the perspective from an endangered large (p.25) middle class and their fears of crime and criminals. It makes ethnic minorities and immigrants particularly vulnerable targets for the ‘punitive impulses’ of the middle classes. They are deemed to be a ‘social threat’ to other social groups mostly situated in the core of contemporary societies, independent of their economic status (Baumer et al. 2003). In a comparison of US states, Campbell et al. (2015) find that race and racial inequality have an enduring presence across decades in determining imprisonment in the country (see also Campbell 2018; Schoenfeld 2018; Western 2006). In Europe, immigrants are similarly seen as threatening.

The role of the welfare state is particularly under-theorised in the political economy of punishment. However, as Garland (2001) points out, it is a decisive link between the distribution of income, goods, and services in contemporary societies and the criminal justice system itself (see also Karstedt 2011b). Access to resources and support structures constitutes an important part of the (re)distribution of income. This levels out economic and social inequality, particularly at the bottom of the social pyramid, and provides a safety net for marginalised groups. This includes access to basic and higher education, health and welfare services, both for those in need and also for the middle classes who profit from free education, health care, and pensions. The withdrawal of welfare support for the poor is identified by Wacquant (2004) as a driver of the hyperinflation of social control through criminal justice in the USA. In contrast, the more generous and equal provision of welfare in European countries is seen as a shield against rising imprisonment rates both in comparison with the USA and within Europe (Snacken & Dumortier 2012). Indeed, exclusion from welfare provision often is a consequence of imprisonment in the USA (Uggen et al. 2006), thus deepening the ensuing disadvantage and inequality.

Dynamics of Change, Contexts of Inequality

Ironically, Wacquant’s analysis was published when the US imprisonment rate started to decline without any signs of a substantive change of neoliberal policies, but actually as part of neoliberalism (Brown 2013; Karstedt et al. 2019). The introduction of neoliberal policies in the United Kingdom in the 1980s did not coincide with an increase in imprisonment, which started later (Farrall & Hay 2010). In Latin America, inequality and poverty were on the decline when imprisonment rates rose dramatically over recent decades. Rising rates of violent crime and a burgeoning middle class seem to have been the drivers of this change (Bergman 2018; Karstedt 2020). Finland drastically reduced its imprisonment rate in the second half of the 20th century without major changes to its level of inequality in terms of income or other social structure indicators. Over the past few decades, European countries have increased as well as decreased their prison populations without discernible changes in social and economic inequality, except for the growth of the proportion of immigrants in these countries (Karstedt 2015a). Penal populism in the United States, Canada, the United Kingdom, New Zealand, and Australia developed along (p.26) different trajectories within the context of a range of economic policies, yet without much change having occurred in parallel in terms of measures of socio-economic inequality (Roberts et al. 2003).

Seemingly, imprisonment rates and prison populations wax and wane while economic and, perhaps even more so, social inequality is highly immutable. If at all, social and economic inequality undergoes rather long-term changes (Piketty 2015), while criminal justice systems and incarceration rates can undergo shortterm change. This raises doubt about the concurrence of social inequality and punishment, and definitely about the existence of causal links between them (see, e. g., Phelps & Prager 2016 for the US case).

With little empirical support for the concomitant development of inequality and criminal punishment, and the hypothesis that as the gap between the rich and the others widens the poor get more prison, cross-sectional comparisons between states and societies might provide a better understanding of the links between inequality and punishment. The criminological gaze has focused on the exceptionalism of mass incarceration in the USA (Enns 2016; Reitz 2018) and simultaneously on ‘Scandinavian exceptionalism’ (Pratt & Ericksson 2013). This contrasts the extreme endpoints of the range of global imprisonment rates, but not those of social and economic inequality on a global scale (see Miethe et al. 2017). However, within the group of economically advanced (OECD) countries, the USA represents indeed extremes of both imprisonment and inequality, while the Scandinavian states are firmly situated at the lower end of both income inequality and imprisonment rates (Wilkinson & Pickett 2009). From a global perspective, the effects of socio-economic inequality on punishment are ‘always context-specific’ and mainly depend on levels of development besides regional clustering (Miethe et al. 2017: 168). Thus, high socio-economic inequality is a condition for high imprisonment only in advanced economies, but not in less developed countries, where no clear pattern of relationships between inequality and imprisonment rates is identified. Generally, across the globe, deviant cases are the rule, that is, countries with high inequality and little imprisonment (mostly poor countries) and with low inequality and high imprisonment (for instance, Australia). This suggests not only different contexts, but more generally different dimensions of social inequality and power differentials, and potentially different mechanisms through which social inequality shapes criminal punishment and accounts for the over-representation of disadvantaged groups in the correctional system.

Understanding the Multifaceted Nature of Inequality and Punishment

Social and economic inequality has many faces, not least a gendered one. It also has different meanings in different contexts, varying from advanced economies in democratic countries to highly hierarchical and authoritarian countries. Marginalised groups differ widely, from immigrants in countries in the global north to indigenous (p.27) populations in Latin American countries. The gap between those living beneath the poverty line and even the lower middle classes might be more meaningful in terms of socio-economic inequality than the gap between the top 1 per cent and the rest of the population.¹ Generally, such social and economic inequalities do not seamlessly translate into power differentials between groups, or into state power. The multifaceted nature of social and economic inequality suggests a complex set of mechanisms that link inequality and the characteristics of criminal justice systems.

First, high levels of economic and income inequality create highly vulnerable groups at the lower end of the income distribution, which have a higher risk of contact with the criminal justice system, but are also vulnerable as a consequence of imprisonment (Petersilia 2003; Western 2006). Second, differential access to resources of the welfare state is as important as the distribution of income. Exclusion from such services often is a consequence of imprisonment, particularly in the USA (Uggen et al. 2006), thus deepening the ensuing disadvantage for marginalised groups. Third, socio-economic inequality translates into power differentials, with those at the upper end of the income distribution wielding more political and institutional power. However, political power might be less important than opportunities to access law and legal institutions, to defend themselves and weather a criminal procedure. As such, socio-economic inequality translates not only into unequal treatment within the criminal justice system, but also into keeping specific social groups out of it (e.g. perpetrators of white-collar crimes). Finally, discrimination against specific groups based on race or ethnicity is a decisive dimension of inequality that partially but not fully coincides with all other dimensions of inequality. It makes these groups particularly vulnerable to being targeted by criminal justice agencies, as they are deemed to be a ‘social threat’ to other social groups, independent of their economic status (Baumer et al. 2003).

The focus on ‘penal populism’ has obscured the role of values and cultures of inequality in shaping the link between inequality and punishment. Egalitarian values are the yardstick against which most contemporary societies measure the fairness and equality of income distributions, access to resources, welfare provisions, or discrimination against minorities. As James Whitman (2003) has shown, egalitarian values such as those found in the USA are conducive to harsher punishment. For a global sample, my research on the relationship between egalitarian values and punishment found a strong and consistent relationship with prison conditions, i.e. the harshness of punishment, but no or only a weak relationship with prison populations (Karstedt 2011a; 2011b). However, those environments where egalitarian values are more meritocratic and predicated on equality of chances (e.g. the USA, Australia, or the United Kingdom) tend to have higher imprisonment rates. Di Tella and Dubra (2008) support these results, having found that harsher punishment is a characteristic of meritocratic (‘effort pays’) cultures.

(p.28) These findings suggest a different role for the middle classes and their values in influencing criminal justice, and thus different mechanisms that link inequality and punishment. Growing middle classes, as in many developing countries (Goldman Sachs 2008), drive down the level of socio-economic inequality, and as they wield more political power, the link between economic and political power weakens. As recent research on Latin America has shown (Karstedt 2020), significantly decreasing economic inequality and poverty coincided with sky-rocketing imprisonment rates, thus reversing the assumed link between inequality and punishment. The reduction of poverty enlarged the lower middle classes and the economy generated gains for the socio-economic groups in the centre of society, which transferred into increasing political power and voice for this group. Thus, more powerful rather than socially and economically squeezed middle classes seem to be the ‘penal populists’ in developing countries and drive the process of increasing punitiveness, in terms of both prison populations and harsh prison conditions (see also Garland 2001).

Like socio-economic inequality, criminal punishment is a multifaceted concept. Mostly, imprisonment, and in particular mass imprisonment, has been analysed within the framework of the political economy of punishment. However, prison populations are just one indicator of punishment. Inequality in political power has been found to be a predictor of retention of capital punishment (Miethe et al. 2005). As Whitman (2003) has shown, conditions of imprisonment and the status of prisoners are related to specific cultural meanings of egalitarianism. My own research (Karstedt 2011a; 2011b) corroborated this finding for a contemporary cross-national sample; prison conditions are worse in countries with higher levels of socio-economic inequality and less egalitarian values. As criminal justice is an expression of state power, it is here where state repression in its more exorbitant forms is executed, and where the basic human rights of those who come in contact with the system are potentially violated on a large scale. This includes extra-judicial killings, torture, disappearances, and political imprisonment at the hands of security forces and within criminal justice. Levels of such human rights violations are significantly lower in more egalitarian and equal societies (Karstedt 2011b; 2015b). In Latin American countries, such human rights violations decreased concomitantly with economic inequality and poverty. Trends in the distribution of political power and state violence moved in the same downward direction during recent decades. However, Latin American societies that retained higher levels of urban poverty also have worse records on such state violence and human rights violations. Seemingly, repressive state and criminal justice action particularly targets the urban poor and addresses conflicts emerging in urban spaces and the neighbourhoods of the poor (Karstedt 2020; see also Glenny 2015).

These results suggest that the link between socio-economic inequality and punishment is predicated on the dimension of criminal justice that is explored. Socioeconomic inequality will have a different impact on each of the three different dimensions–imprisonment, prison conditions, and human rights violations within criminal justice. The larger groups of vulnerable people in more unequal societies will account for larger prison populations. Economic and political powerlessness (p.29) will transform prisons into warehouses with deteriorating conditions. Repressive violence and human rights violations will be more prevalent under conditions of socio-economic inequality and unrestricted power of groups at the top of the social hierarchy.

Inequality and the Global Contours of Criminal Punishment

Given the context-dependence of inequality as a condition of punishment, the ensuing analysis focuses on a global sample of countries as well as a sample of advanced economies. The sample comprises between 80 and 169 countries from all regions of the globe, the precise number depending on the availability of data, and covers the years from 2010 to 2015. All measures used are mean values for this period. Three indicators of economic and political inequality are used, including two indicators of income inequality and one indicator of differential political power, the latter basically reflecting foundational arguments of the political economy framework. Income inequality was measured with reference to the Gini index (of household income) and the Twenty/Twenty index (top 20: bottom 20 ratio of household income), both of which include transfers like welfare provisions or pensions. The latter measures how much richer the top 20 per cent of a given population are compared with the bottom 20 per cent, with higher values indicating a concentration of income at the top end of the social pyramid. Compared with the Gini index, it reduces the effect of outliers at the very top (the 1 per cent richest) and the bottom, and makes the impact of inequality in the middle 60 per cent of the population visible. The third indicator measures the distribution of political power by socio-economic position and, as such, the extent to which economic inequality translates into political power differentials. The index is based on expert assessments, and ranges from zero to four, with higher values indicating a more equal distribution of power across all socio-economic groups. Here a standardised interval scale is used, ranging between–5 and +5, with zero approximately representing the mean of all country-years included.

The prison population data are based on the World Prison Lists and the World Prison Brief; the incarceration rate per 100,000 was used. Prison conditions reflect conditions in prisons on a national level. The scale is based on information retrieved from the annual Human Rights Reports issued by the US Department of State, and therefore does not include the USA; it ranges from one (prisons meet minimum international standards) to five (with life-threatening conditions like lethal violence and diseases). Human rights violations or state violence are measured by the Political Terror Scale, which includes extra-judicial killings, torture, disappearances at the hands of security forces, and political imprisonment. The measure used here combines two sources, the US State Department and Amnesty International Country Reports, achieving a sum scale ranging from one to nine. Higher values indicate higher levels of state violence (for all indicators, see Appendix).

(p.30) In the global sample, both indicators of income inequality are weakly (Gini) or not at all (Twenty/Twenty) related to imprisonment rates. Figure 2.1 shows, for the Gini index, that more egalitarian countries at the lower end widely differ in terms of their imprisonment rates. African, Asian, and Latin American countries have low imprisonment rates relative to their higher level of income inequality. Outliers among countries with more equal income distributions are eastern European post-communist countries as well as the USA, which confirms its ‘exceptionalism’ throughout all indicators of inequality. Figure 2.2 demonstrates the importance of context: in the group of countries with advanced economies (Organisation for Economic Cooperation and Development, OECD), high income inequality drives imprisonment rates up, which applies to both the Gini (as shown in the figure) and the Twenty/Twenty index (not shown). In roughly equalised conditions of economic performance and welfare state provisions, as well as political stability and democratic regimes, income inequality in fact makes a difference between low and high incarceration countries. Scandinavian countries imprison less even in relation to their low levels of income inequality. Together with post-communist countries or the United Kingdom, the USA emerges with an exceptional incarceration rate given its level of income inequality, even in this group of approximately similar countries.

The extent to which political power is defined by economic status is unrelated to the size of prison populations in the global sample, as shown in Figure 2.3. However, we find again a context-specific impact in the advanced economies of the OECD countries, indicated by the white triangles in Figure 2.3. In Europe,

Figure 2.1 Inequality and imprisonment: Gini index and imprisonment (rate, ln) (country mean 2010–15).

Sources: See Appendix; N = 80; R²= 0.082; p < 0.010.

(p.31)

Figure 2.2 Inequality and imprisonment in OECD countries: Gini index and imprisonment (rate, ln) (country mean 2010–15).

Sources: See Appendix; N = 33; R²= 0.379; p < .001.

Figure 2.3 Political power and imprisonment: power distributed by socio-economic status and imprisonment (rate, ln) (country mean 2010–15).

Sources: See Appendix; N = 151; R²= 0.0007; p = 0.753.

the Scandinavian countries, the Netherlands, France, and Germany, where political power is mostly independent of socio-economic position, have low incarceration rates. Where economic and political power go hand in hand, as in the UK, Mexico, or the USA, incarceration rates are higher; nonetheless, even in this group (p.32)

Figure 2.4 Inequality and prison conditions: Twenty/Twenty index and prison conditions (country mean 2010–15).

Sources: See Appendix; N = 107 (without South Africa and Haiti); R²= 0.235; Adj. R²= 0.220; p = 0.282.

of countries, the USA’s mass incarceration stands out relative to the country’s level of economic and political power distribution. In sum, support for a cornerstone of the political economy of punishment, the impact of economically based political power and economic inequality more generally, is mixed at best, highly context-dependent, and nested in advanced rather than developing economies and societies.

We now turn to the ways in which socio-economic and political inequality shape the conditions of prisons and the living conditions of prisoners. Figure 2.4, based on a global sample of 107 countries, shows a non-linear relationship between the Twenty/Twenty index of income inequality and prison conditions, which is not significant. However, two clusters can be identified: one cluster of countries with prison conditions that meet international standards as, e.g., stipulated in the socalled Mandela Rules, and another cluster of mainly developing countries with life-threatening prison conditions. Many countries in this latter group share a small band of levels of inequality with those that provide far better conditions in their prisons. As we move from one cluster to the next, a sea change in context and conditions rather than small changes in inequality sets these two clusters apart and seems to account for life-threatening prison conditions. Within the group of advanced economies as represented by the OECD countries (white triangles), a weak though significant relationship between income inequality and prison conditions is found.² (p.33)

Figure 2.5 Political power and prison conditions: power distributed by socio-economic status and prison conditions (country mean 2010–15).

Sources: See Appendix; N = 154; R²= 0.272; p < .001.

Countries as diverse as Japan and Portugal stand out with harsh conditions in prisons relative to their level of income inequality, while New Zealand, Germany, and the Nordic countries do better than their level of income inequality would suggest.

In contrast to prison populations, prison conditions are related to the interaction between power differentials and socio-economic stratification (Figure 2.5). In particular, prison conditions are much better in countries where power is distributed independent of socio-economic status, as compared with countries where differences in political power are shaped by economic power. This latter group includes, among others, Russia, and African, Asian, and Latin American countries. However, we find numerous outliers on either side of a linear relationship, with South Africa, Belarus, and Niger having worse conditions relative to their power distribution, and the UK, Australia, New Zealand, and Germany having much better conditions relative to the influence of economic on political power.

The extent of human rights violations committed within the criminal justice system is indicative of the repressive and violent nature into which criminal justice might morph, often in ‘wars’ declared on crime. As Figure 2.6 shows, based on a global sample (155 countries), political power differentials are distinctly related to this type of state violence, with countries at the lower end of the scale of state violence distributing political and economic power in more equal ways. Authoritarian and fragile states where economic elites wield the highest degree of power, as well as conflict and post-conflict states, have exceptionally high levels of state violence/human rights violations (e.g. Russia, Congo DR, Sudan, Nigeria, Colombia). A similar picture emerges for the advanced economies (OECD countries, not shown), where a weak relationship is found, with outliers such as Mexico, (p.34)

Figure 2.6 Political power and state violence: power distributed by socio-economic status and human rights violations (country mean 2010–15).

Sources: See Appendix; N = 155; R²= 0.328; p < .001.

Israel, and Turkey. Both indicators of income inequality are related to human rights violations; in a global sample (Gini index: 80 countries; Twenty/Twenty index: 109 countries, two outliers excluded, not shown), such human rights violations are significantly more frequent in unequal societies though the relationship is weak. In contrast, in the group of advanced economies and mostly established democracies (OECD countries, not shown), socio-economic inequality is a considerably strong determinant of human rights violations within the criminal justice system. Socioeconomic inequality unfolds its impact on state violence in the context of political stability rather than in conflict and highly insecure environments, which in themselves drive state violence to exceptionally high levels mostly independent of levels of inequality, as demonstrated by outliers such as Israel, Turkey, and Mexico.

The Idiosyncrasies of the Political Economy of Punishment

Using a few selective indicators of both socio-economic and political inequality and of the criminal justice system, our bi-variate analyses only find tenuous and variable links between inequality and criminal punishment. For both sides of the equation, scores of other indicators are available and could be tested. For inequality, income distribution and inequality as used here differ from wealth distribution, which has recently sparked a debate on rising inequality, with the increase of the share of the ‘1 per cent’ of the super-rich against the 99 per cent of all others (Piketty 2015). Besides questions of measurement (The Economist 2019), it is doubtful whether changes in wealth distribution and inequality are related to changes in the political (p.35) economy of punishment at all. Distributional inequality should include indicators of deprivation and exclusion, or poverty, rather than measures that juxtapose the super-rich with the range of all socio-economic groups. The adoption of neoliberal policies might have an impact on the small group at the top of the pyramid of wealth and income; however, we can assume that it is mainly the impact on groups in the middle range and on those at the bottom that could and should make a difference in penal policies. Similarly, criminal punishment needs to be subjected to a broader test and include multidimensional measurement (Hamilton 2014); on a global scale that would include, for example, the retention of the death penalty, usage of life imprisonment, or admission rates rather than prison populations for gauging the characteristics of criminal justice systems. However, the findings presented here contribute to a body of large-scale comparative research that stresses the highly contextual nature of the link between inequality and criminal punishment (Miethe et al. 2017). The fact that such a link exists for the conditions of advanced economies and mostly established democracies, as in the group of OECD countries (Wilkinson & Pickett 2009: 148–9), for the US states, or even smaller groups of European countries (Lappi-Seppälä 2008), does not allow for generalising to other economic and political conditions. Nor can this link be generalised across time: changes in socio-economic inequality are not linked to changes in incarceration rates in OECD countries. Between 1996 and 2015, income inequality measured as the relation between the top 20 and bottom 20 per cent of household income remained mostly stable and decreased slightly after 2008, actually implying an increase in the income share of the lowest 20 per cent. During this period of socio-economic stability, the mean incarceration rate in these countries increased from below 120 per 100,000 to above 160, and only slightly levelled off after 2008/2011. For Latin American countries, socio-economic and political inequality as well as poverty levels decreased substantially, at the same time that imprisonment rates sky-rocketed (Karstedt 2020). Consequently, this link mostly disappears in a global sample of countries with a huge variety of political economies, ranging from established democracies to narco-states or authoritarian oilstates. The results presented here are thus indicative of the highly idiosyncratic nature of the political economy of punishment. With regard to criminal justice and punishment, more equal societies not only do not always do better, but they can do worse. If the middle classes get richer, imprisonment soars (Bergman 2018). Nonetheless, some general conclusions and lessons can be teased out.

The impact of socio-economic and related political inequalities seems to unfold mostly under the conditions of advanced economies, political stability, and democratic regimes, as these are found in most OECD countries, in European countries, and in the USA. Under conditions of political instability, state fragility, and developing economies, socio-economic and political inequality become negligible forces in shaping criminal justice systems and punishment, which under these conditions are often not only dysfunctional, but also directly used for political purposes. Here the political economy of punishment follows different pathways to those identified for the rich countries mostly of the global north.

(p.36) These findings in conjunction with those for Latin American countries draw attention to the political economy of the middle classes. As inequality and poverty decreased in Latin America, middle classes increased and became politically more powerful, with the impact of powerful economic actors on the political system diminishing. However, the rising middle classes might also have been a seedbed for penal populism amid insecurity and fear, and thus partially responsible for the sky-rocketing prison populations and deteriorating prison conditions in most Latin American countries. In contrast, in the advanced economies of the global north, the middle classes lose out in terms of income, welfare provisions, and overall living standards. Here, penal populism seems to be the result of a deteriorating economic position, even if levels of inequality remain more or less stable (as in the OECD countries), rather than of middle-class empowerment. In addition, middle-class sentiment is not seamlessly translated into penal policies in a large number of these countries that under the European Human Rights regime ‘resist punitiveness’ and have the capacity to do so (Snacken & Dumortier 2012).

This does not imply that each nation embarks on its own and unique pathway of a political economy of punishment. However, it calls for identifying and cautiously dissecting different contexts in which a political economy of punishment unfolds. Simultaneously, it is a warning against over-generalisation of findings from a few countries or from a group of specific countries (like those of the OECD), and more generally against drawing sweeping conclusions from in-depth analyses of a single nation and its adoption of neoliberal policies. We need to use the results from such studies cautiously in our search for a multiplicity of political economies of punishment and a variety of links between social and economic inequality and our ways of punishing.