Explanations for the Punitive Turn of Crime Policy in Bolivia, 2006-2016

In the last decade, Bolivia, as with most countries in the region, has seen an unprecedented increase of its prison population. This is often explained as the consequence of a punitive populism sweeping Latin America. Our article investigates what triggered this punitive turn in Bolivia by identifying some of the factors that impact crime policy and growing prison populations since the election of president Evo Morales in 2006. We argue that a complex array of local and international factors and shifts in crime policy to harden approaches to domestic violence led to steep increases in remand populations. Combined with other inefficiencies in the criminal justice system, this led to sustained increases in the prison population throughout most of this period. This study is based on new and previously unstudied statistical data produced by the Bolivian institutions in charge of implementing crime policy.


Introduction
After a decade of increasing economic inequality and crime rates in the 1990s, most Latin American countries have hardened crime policy during the 2000s. As a result, these countries have experienced unprecedented increases of their prison populations (Riego and Duce 2009: 10;Stippel 2013: 61-63). 1 Subsequently, criminologists have tried to explain this development, referring to it as the 'punitive turn' or 'new punitiveness' (Goldson 2002;Pratt 2007), and focusing their analyses mostly on the phenomena of 'penal populism' (Paladines 2016;Sozzo 2014) or a process of penal state building (Garland 2015;Hathazy and Müller 2016;Wacquant 2009).
In another context, the punitive turn affecting developed countries can be explained partially by the rise of far-right and populist political movements which use the punitive rhetoric as an essential element of their anti-migration and xenophobic discourse (Hogg 2016). As a result of electoral competition, the punitive rhetoric has pushed many western governments to harden their criminal policies and to move away from left realism and the rehabilitation ideal (Matthews 2014). This thesis seems heuristic for understanding the evolution of crime policies in developed countries. Nonetheless, this argument does not explain, in our opinion, the reasons for the punitive turn in Latin America, in countries where leftist 'populist' 2 governments came into power, as in Bolivia, Ecuador and Argentina in the last two decades. Accordingly, we argue that, in Bolivia, the punitive turn cannot be understood primarily as the consequence of a populist political change. However, the prison population has increased substantially since the adoption of the new constitution in 2009 and the birth of the Plurinational State of Bolivia (the country's new official name).
Evo Morales was elected President in 2006, when he was the leader of the political movement based on the alliance between rural indigenous activists and middle class urban Marxists (Errejón 2011;García Linera 2009). This movement's major objective at that time was to recover national sovereignty endangered mostly by the United States (US) anti-drugs intervention in the country. After the election of Morales, the new government adopted administrative measures based on increasing the number of civil servants-mostly police officers and prosecutors-and the adoption of a new preventive paradigm in the anti-drug policy (Stippel andSerrano Moreno 2016, 2018).
In this context, our article aims to answer to the following questions: How much did not only national but also international interests shape Bolivian crime policy? Does the incarceration trend mirror the 'populist' ideology and rhetoric of Morales' government? Could the growing police force and increasing number of prosecutors be related to the incarceration trend? How does the law as it is practiced impact upon the implementation of crime policy in Bolivia?
Our sources are informed by criminal law in force from 2006 and 2016 and by newly released and previously unstudied statistical data produced by the Bolivian crime policy institutions. We show that the traditional explanations for the punitive turn to some extent help in an understanding the Bolivian case. Nevertheless, the most important factor from our point of view is not the politicization of the Bolivian crime policy but its endemic inefficient institutions that lack professional public management and general interest in accountability of public servants.

Main factors for the increasing prison population in the region
The Argentinian criminologist, Máximo Sozzo (2014), has offered an analysis of the evolution of crime policy in his country. He argued that penal populism 3 was a main factor that determined the harshening of crime policies and thus the increase in the prison population. Sozzo observed that, up to the end of the last century, the elaboration of crime policy followed an elitist mode. It focused on the word and knowledge of the expert whose agenda was protected from intervention by the general public. Consequently, since the end of the 1990s, a wave of 'penal populism' IJCJ&SD 35 Online version via www.crimejusticejournal.com © 2018 7(4) emerged. Sozzo qualifies this first wave of populism, as 'penal populism from above' (populismo penal desde arriba) and he argued that it was generated mostly by political agendas and media. For the first time since the end of Argentine's transition to democracy, what 'people think and want' was emphasized as the main focus of new policy. Sozzo stressed that 'penal populism from above' also had an elitist component because certain privileged actors, such as politicians and journalists, claimed to speak on behalf of the public and presented their views as unambiguous and preexisting. The 'experts' were seen to be among the victims of this development, as they were side-lined by the elitists that made the decisions. Sozzo argued that 'penal populism from above' then turned out to be the forerunner of the next wave of policy changes, 'penal populism from below' (populismo penal desde abajo).
This new form of penal populism implied that the policies were pushed 'from below', following the claims of social movements and their representatives. As a result, certain sectors of society, those whose lifestyle had been adversely affected by criminal activities or those who had been a victim of crime, were able to become representatives of the abstract notion of 'people'. Using this legitimacy, they could ask for changes in the field of crime policy (Sozzo 2014(Sozzo : 116, 2016. Nevertheless, Sozzo (2014Sozzo ( : 135, 2016 observed that this development came to a standstill after the Argentinian elections in 2005. Since then, a more moderate approach to crime policy has been in place. Still, he finds it is hard to imagine that the levels of punitiveness gained in the populist period might be overcome. From another perspective, the Ecuadorian criminologist and public defense lawyer, Jorge Paladines (2016), analyzed the changes to penal policies in Ecuador. He argued that economic factors are not sufficient to explain changes in punitiveness. In spite of Ecuador's considerable progress in reducing inequality and poverty, the fact that there were fewer poor people did not mean that prison occupation rates were reduced. On the contrary, the poverty reduction programs had been accompanied by an increase in prison population numbers not previously experienced. He argued that the increase could not be explained from an economic standpoint; instead, it provided evidence of the dysfunctional character of penal institutions, because they do not curb but increase selective imprisonment (Paladines 2016: 178). Going even further, he thinks that the progress made in improving the living conditions of Ecuadorians had triggered a climate of zero tolerance towards crime (Paladines 2016: 180).
Adopting a macro-sociological perspective, Paul Hathazy and Markus-Michael Müller (2016) analyzed what they call the 'rebirth of the prison in Latin America' in order to explain the punitive turn in the region. They argued that the expansion of police powers and judicial capacities along with the passing of more punitive legislation 'unsurprisingly' led to a boost in the region's prison population, due to the weakness of what they labelled 'the impartial legal dimension of the exercise of state power' (Hathazy and Müller 2016: 120). The authors identified this flaw-which they thought was historical-in court practices and those of the prosecutors (Hathazy and Müller 2016). They also thought that the rise of the region's inmate population was mainly related to new forms of 'punishing the urban poor' (Hathazy and Müller 2016: 114).
In a broader and comparative approach, the United States law professor, Michael Tonry (2007), reviewed the literature in search of generalizations that help to explain national differences in penal policies and practices. His focus was not Latin America but Western countries but his findings are helpful for our analysis.
Tonry understood penal cultures and policies to be local; thus, the reasons for harsher penal policies can be found in distinctive cultural, historical, constitutional and political conditions. He argued that, in some places, penal policies might have become harsher but, in most places, changes are caused by adjustments in practice that moderate and sometimes nullify the policy variations, and by other policy changes that move in the opposite direction. Tonry argued that comparisons of shifts in penal policy also need to differentiate between their enactment, their implementation and their practical use. In other words, sometimes policies are enacted to send messages or to make expressive or symbolic statements with no clear expectation that they will be implemented (Torny 2017: 1, 12). He argued there are 'nonfactors' that do not explain many of the change in crime policies. Amongst those, he counts social and economic changes; politicization of crime policy; increased population diversity and intergroup conflict; the effects of the women's, gay, and civil rights movements; and increasingly global and sensationalistic media. As those developments affect all developed countries, they are just background conditions without a specific explanatory value. Whether those factors impact crime policies depends on their interaction with other factors (Torny 2017: 17).
If we relate Tonry's findings to Sozzo's (2014) arguments, the politicization of crime policy by populist means would be, on the one hand, a 'nonfactor' if, at a regional level, they could be considered to be background conditions affecting all countries. On the other hand, Tonry's arguments would match partially the findings of Paladines (2016). Both considered economic factors essentially unhelpful in explaining penal policies as they relate to incarceration rates. Regarding the expansion of police powers and judicial capacities and also more punitive legislation-factors mentioned by Hathazy and Müller (2016) and considered at an earlier time by -it is doubtful if those policy changes were implemented in practice or if they only changed the laws in the books, without much practical impact.
We now consider how far any of the before-mentioned factors and explanations help to understand the changes of crime policy in Bolivia.

Foreign interests
If we look at the major legislative efforts that have shaped Bolivian crime policy, we find that there are some particular influences not previously mentioned. Contrary to the experience described by Sozzo (2014) in the case of Argentina, Bolivian crime policy has often been criticized for responding to foreign interests and for being driven by the international cooperation agenda rather than by local imperatives. This is true, firstly, for the drug legislation.
The major drug legislation, commonly known as 'Law 1008', was published on 19 July 1988. It is a 'hybrid' law, combining in a single act provisions that regulate production, substitution measures and eradication of coca leaves, along with criminal and procedural law and rules on prevention (Stippel and Serrano Moreno 2018). Law 1008 has been criticized by different sectors of Bolivian society and from different perspectives. The National Council Against Illicit Drug Trafficking (CONALTID), 4 composed by the Ministers of the Presidency, Foreign Affairs, Defense and Health, criticized in 2007 the approval of the law, arguing that it responded mainly to external demands without offering structural solutions to the drug problem in Bolivia. They argue that, on the contrary, it contributed to the enslavement of history, identity, culture, traditions and damage to the environment in the country (CONALTID 2017: 18). The historian and former president of Bolivia, Carlos Mesa Gisbert (2017: 64), considers Law 1008 to be an instrument that bypassed the country's constitution, with special courts and almost omnipotent anti-narcotic prosecutors. 5 The government of President Evo Morales promised a review of Law 1008 and eventually drafted a general law on coca, published in March 2017. The new Coca General Law (Law 906, published on 8 March 2017) expands the areas of legal coca leaf production (from 12,000 to 22,000 hectares) and determines that the Ministry of Rural Development is in charge of supervising the commercialization and transport of coca leafs. 6 It also establishes that the coca plant is part of the cultural heritage of Bolivia, protected as a renewable natural resource and necessary part of social cohesion. 7 Nevertheless, the new legislation does not replace the penal rules of Law 1008 which are still in force. 8 IJCJ&SD 37 Online version via www.crimejusticejournal.com © 2018 7(4) We find that one of the cornerstones of Bolivian crime policy, which accounted traditionally for more than half of Bolivians prison population 9 and still makes up for about one fifth of all incarcerated persons, 10 can be understood as a product of external demands.

Political change
A major shift in Bolivia's crime policy took place after the new constitution was enacted in 2009. A wave of new penal laws was passed by parliament (Plurinational Assembly). During the period from 2009 to 2017, the Bolivian legislator adopted 34 new laws, with an impact on criminal law. 11 Concerning the legislative activity in the previous period, from the end of the dictatorship in Bolivia in 1982 to the promulgation of the new constitution in 2009, a total of 13 acts were issued that created or hardened criminal offences. 12 During the short transitional government of President Eduardo Rodríguez Veltzé (from 9 June 2005 to 22 January 2006) an act was passed that repealed in part rules increasing penalties (Law 2625) approved in the previous government.
We find that the Plurinational State of Bolivia dedicated considerable effort towards shaping its crime policy through enacting new criminal laws. Already the fourth law (known as Law 004) approved by the new Plurinational Legislative Assembly in 2010 created eight new statutory offences, all linked to the 'fight' against corruption. 13 The law was not limited to identifying offences; it also created both anti-corruption courts and special anti-corruption prosecutors. 14 Accordingly, the National Council to Combat Corruption, Illicit Enrichment and Legitimacy of Illicit Profits was established. 15 It does not seem coincidental that this first law with relevance to crime policy-Law 004formed institutions similar to those foreseen in Law 1008, including specialized courts 16 and prosecutors, 17 and an inter-ministerial council 18 dedicated to approve, among other things, a national plan to fight against corruption. 19 It appears that the Bolivian legislator wanted to signal that, from now on, the priorities of criminal policy were different.
Another similarity between Law 1008 and Law 004 concerns their deficiencies in the protection of civil rights. Although Law 004 represents progress because it no longer violates the presumption of innocence or the right to defense as did Law 1008, the new law infringes the general principle of the prohibition of retroactivity of ex post facto criminal laws recognized in many foreign democratic constitutions. Indeed, art. 123 of Constitution of the Plurinational State of Bolivia specifically recognizes the retroactive effects of criminal law concerning political and administrative corruption for 'crimes committed by public servants against the interests of the State'. The constitution attempted to allow the prosecution of politicians and public servants of the 'ancien régime' for actions dated before the enactment of the new constitution in 7 February 2009 and later qualified as crimes by the news laws adopted by the new authorities. It appears that, nowadays, it is not only international pressure and foreign interests that are shaping the crime policy of the Plurinational State of Bolivia, but that national politics is also playing a major role. Examination of available statistics may indicate whether these legislative changes have had a practical impact in terms of registered offences and incarcerations, to which we now turn.

Registered offences
If we analyze the number of offences registered by the Bolivian Police (see Table 1), we find that there was a steady decrease in the number of offences registered in the period from 2000 (47,300) to 2005 (32,150). From 2005 to 2015 the total number of registered offences steadily grew to an all-time high in 2015 (88,895) although, in 2016 (the last year for which data are available) the total number of registered offences decreased. If we look at the relative figures (the number of registered offences per 100,000 inhabitants), we find that, in 2000, the probability of becoming a victim of a crime (561 per 100,000 population) was almost identical to more than a decade later (in 2011, 567 inhabitants per 100,000 population); even so, there was an increase of more than 10,000 registered offences. The decrease in the offences registered in the first five years of the new century could be related to the social unrest that lead the former president to flee the country after an estimated 68 people were killed in riots after the declaration of martial law, many on the central square of La Paz and in El Alto in October 2003. 20 The future of the state as a whole was uncertain. People lost confidence in the state institutions including the police force, and they therefore sought to solve their problems without the help of the police. 21  The data support the hypothesis that a weak and de-legitimized state is at least one determinant for a softer approach to crime policy. Hence, in the case of Bolivia, a harsher crime policy might not be so much associated with punitive legislative measures but with the re-legitimization of state institutions. 22

Type of offences registered
The type of offences registered by the Bolivian police allows us to draw further conclusions. For our analysis, we focused on the development of the offences registered that relate to the proliferation of special laws. We looked at offences related to corruption (Ley 004) and racism (Ley 045).
Registered offences committed against the public service include corruption of public servants. Most of these laws were made tougher by Law 004 of the Plurinational State of Bolivia. The data show that the number of offences registered under this rubric had already increased in the years before Law 004 was enacted in 2009 (see Table 2). Nevertheless from 2009 to 2014, the registered offences increased almost fourfold (from 832 to 3,163). To a lesser degree but also noteworthy, the number of crimes related to corruption and mismanagement in the judiciary doubled over the same period (from 369 in 2009 to 778 in 2014). Similar increases occurred with offences related to dignity, amongst them crimes like racism and discrimination, which doubled from 2009 to 2014 (from 299 to 616). 23 The data indicate that the proliferation of special laws contributed to the increase of offences registered by the police. This does not automatically imply that the same is true for the incarceration rates. In fact, the prison population trend does not directly mirror the development of the registered offences.

Increasing incarceration
There was a decline in the total prison population in the early years of the twenty-first century (from 8,124 in 2000 to 5,669 in 2003) that might have been influenced by the enactment of the new act on criminal proceedings known as Law 1970(in May 2001. 24 Nevertheless the total numbers remained relatively stable until 2009 (8,096) and, since then, there has been a steady and steep increase until 2016 (15,056 prisoners). Figure 1 and Table 3 illustrate these trends:    The prison population grew around 75 per cent from 2000 (8,124) to 2014 (14,220). In comparison, registered offences increased by 46 per cent over the same timeframe (from 47,300 in 2000 to 68,918 in 2014). Therefore, responses to a growing number of crimes might have received answers other than prison, which could be a positive development as it could mean that penal sanctions were being diversified.
Nevertheless, if we look at the evolution in the numbers of pre-trial detainees, this possible conclusion appears to be less likely. Thus far, the absolute peak of the pre-trial detention rates was reached in 2012 with more than 85 per cent (12,163) of all prisoners being on remand detention. In the following years to 2016 (data are not available for 2015), we can see a continued decrease (down to 67.8 per cent in 2016).
Judges are legally obliged to modify the pre-trial detention order for a less invasive measure or to lift it at any time if they find that the facts that originally backed their decision changed in favor of the suspect. 25 Nevertheless, the fact that still almost 70 per cent of the prison population is made up by pre-trial detainees indicates that judges do not act this way. Thus, the system did not find other answers to crime than prison; rather, it did not find any answer at all.
In order to gain a better understanding of the practical impact of the latest changes that Bolivia's crime policy has experienced, we examined the data related to the offences for which prisoners were detained in 2016 (see Table 4). Most people were detained for robbery (3,497 or 23 per cent of the total prison population), followed by persons accused of or sentenced for rape (2,985 or 20 per cent) or for drug related offences (2,977 or 20 per cent). Persons accused of or sentenced for murder and homicide comprised another 14 per cent of the prison population (1,274 for murder; 625 for homicide; and 262 for intent of homicide).
It is surprising that, in spite of the high sanctions associated with corruption, the thousands of offences apparently related to acts of corruption by civil servants or the judiciary (there were almost 4,000 such offences registered in 2014), none appeared as detainees in the prison statistics. Hence, it seems that Law 004, the first law of the Plurinational State of Bolivia that was aimed at containing new offences and increasing sanctions for several existing ones, did not contribute to the increase in the incarceration rate.
We cannot correlate these data with other statistics. But, the apparent lower rates of incarceration for corruption could be related to several circumstances. For instance, many of those crimes associated with corruption are difficult to prove in court and might involve legally trained citizens-and/or people who have the means to bribe somebody-so they have more chances to escape prosecution than a person who, for example, commits a robbery. Crimes of the powerful such as corruption are systemically under-criminalised (Barak 2015). Thus, the statistics indicate that prosecuting corruption in Bolivia needs more a sophisticated system of detection and prosecution in order to be effective. Enacting a tougher law has not lead to the imposition of tougher sanctions.
Nevertheless, the statistics show that another special law of the Plurinational State of Bolivia-Law 348 enacted in March 2013, 26 the law meant to guarantee women a life without violenceapparently had a considerable impact on the incarceration rate. This act elevated the minimum penalty for rape from four to 15 years 27 and created offences like feminicide, 28 family or domestic violence 29 and sexual harassment. 30 IJCJ&SD 42 Online version via www.crimejusticejournal.com © 2018 7(4) The high minimum penalty for rape and the fact that judges are requested to impose the preventive detention in order to protect the women during the investigation 31 contributes towards the elevated percentage of people imprisoned under these charges.

Growing of control institutions
Alongside the increase in registered offences, the numbers of persons employed within the main institutions in charge of controlling crime steadily grew. In the last decade, the Police and the General Prosecutor's Office have benefitted from a constant increase in their staff and budget. 32 The personnel resources of the police force increased by about 50 per cent in the last decade (from 23,756 in 2004 33 to 36,567 in 2016 34 ). It is noteworthy that the most substantial increase, more than 7,000 officers, took place between 2006 (26,503) and 2007 (33,523). In his report to the nation, the President underlined that, in 2005, there was one officer per 360 inhabitants and that, during his government, this number grew to one officer per 304 inhabitants by 2013. His goal was to reach a rate of one police officer per 250 inhabitants by 2025, which would point to a total number of 45,000 officers by that year (Morales 2013: 17).
The largest increase in police force numbers took place after Morales gained power in January 2006 when about 7,000 additional police officers were appointed. population in 2004 to 348 per 100,000 population in 2005). Arguably, the appointment of additional police officers were, to some extent, aimed at stabilizing the processes of policy and institutional change anticipated after the election of a new President.
Part of the increase in registered offences from 2007 onwards could then be explained by the growth in police force numbers and thus more police activity aimed at detecting offenders. Hence a larger police force could be another factor pushing up rates of reported crimes.
Parallel to the increase in police numbers was the steady growth in the number of prosecutors. Upon full implementation of the Criminal Procedure Code on 30 May 2001, the number of prosecutors had increased from 220 to 311 (Ledezma 2005: 215). In 2011, there were 411 prosecutors; this rose to 498 in 2015 (Delgadillo Ramírez and Mayta 2015: 134).
It is interesting to note that a considerable proportion of the General Prosecutor's Office budget is donated by different international agencies. If we look at the recent development, the percentage of the budget supplied by international cooperation decreased from eight per cent in 2010 and 2011 to three per cent in 2013. 35 Many of the funds from international cooperationmainly from UNICEF, UN Women and from Denmark-go into investigating crimes committed against women (General Prosecutor of the Plurinational State of Bolivia 2015: 38). As we saw in Table 2, crimes committed against women fluctuated accordingly with the decrease of the international cooperation funds. Table 4 also shows their impact on the composition of the prison population.

Inefficient institutions
The lack of efficiency of the justice system has a direct impact on the prison population trend. If we look at the percentage of cases that are pending each year at the entrance courts (Juzgados de Instrucción), we find that more than half (69 per cent in 2013; 57 per cent in 2014; 52 per cent in 2015) of all cases are not being dealt with each year. They accumulate into lengthy delays in court appearances.  Council 2013Council : 420, 2014Council : 420, 2015 The large number of pending cases is, to a large extent, also the responsibility of the prosecutors. In 2015, prosecutors in the capital cities and El Alto formalized charges in 13,096 cases. 36 An additional 5,961 cases were presented outside the capital cities, 37 making a total of 19,057 formal charges filed on a national level. That means that, on average, each of the 498 prosecutors presented about 38 formal charges in the year, amounting to less than one per week.
The most common way prosecutors end a case is to file refusal reports. Generally when this happens, the Police Force-under the direction of the General Prosecutor's Office-has been unable to identify the possible perpetrator. 38 In practice, the case is archived and ends without anyone being held accountable for the alleged commission of the reported crime. We have to bear in mind that the maximum duration of preliminary police investigations shall not exceed 20 days. 39 By the end of this period, the police should send their report to the prosecutors, who then should analyze the content immediately. In exceptional cases, the law allows the extension of this period 40 (Judicial Council 2013: 420, 2014: 420, 2015 The data indicate that there is a serious organizational problem at the General Prosecutor's Office. In the Institutional Strategic Plan, the General Prosecutor's Office itself established that the core issue adversely affecting its internal organization is due to the limited and inefficient internal management and organization of the entity (General Prosecutor 2014: [35][36]. They attribute the problem to organizational design, incipient planning and management control systems, informal and non-automated information systems, lack of explicit processes and procedures to guide the administrative management, among others. The General Prosecutor's Office assures that the organization is undergoing a reform since the passing of the new Basic Law of July 2012 and the designation of the General Prosecutor who is in charge of dismantling 'the corrupt, perverse and inefficient structure of the General Prosecutor's Office' (General Prosecutor 2014: 22).
It is noteworthy that the internal management issues of the General Prosecutor's Office are not new or unknown. A decade ago, Ledezma (2005) declared that one of the most worrisome aspects of the transformation of the Bolivian criminal justice was the fact that the General Prosecutor's Office had not become a leading institution in the reform, and that their annual operative programs disregarded indicators that showed the (in)efficiency and (under)achievement of institutional goals (Ledezma 2005: 213). Ledezma (2005: 219) indicated that the data produced each year had no impact on the Institution because the data were not being analyzed to improve performance. While this is disturbing, the recent declarations of the General Prosecutor justify some hope for change; nevertheless, the numbers still do not.
The data on the refusals also aid the assessment of the quality of the work done by the National Police and the General Prosecutor's Office when leading the investigations. If more than 80 per cent of the cases prepared by police cannot be used to file a case at court, there is a systemic lack of quality control in the preparation of cases, and other deficiencies in the functional management of criminal investigations. 42 It is also impossible to pursue specific goals, such as increasing the percentage of cleared crimes with regard to homicide, rape or theft or any other felony, if there is no guidance nor any data about investigation refusals differentiated by type of crime. The consequence is that the police determine the 'what' and 'how' of investigations; they are the gatekeepers of the criminal justice system. One of the main goals of the criminal procedure reform fully enacted in mid 2001-to professionalize the investigation work in order to prosecute more complex crimes-is clearly failing in practice.

Conclusions
The Bolivian criminal justice policy lacks coherence. It is plagued with systemic flaws in processing and prosecuting crimes. While harsher sentences were introduced for some crimes (that is, corruption) and new crimes added to those punishable, the dysfunction in the system did not automatically lead to more practical punitivitness. If the idea of all those legal and institutional changes ever was to establish a 'rule through law' (Hathazy and Müller 2015: 116), any such intent failed in Bolivia due to deficiencies of internal organization, weak management and a lack of strategic orientation. Thus we agree with some of the main findings of  that, in order to understand the punitive turn observable in many countries both in the Global North and South, we need to focus on legal practice at a local level that can neutralize the policy changes and adopt a multi-factorial approach centered on concrete local particularities.
The Bolivian case also illustrates that some legal modifications did result in more offenders going to prison-and for longer terms-for violence against women (Law 348). The enactment of this law has been accompanied by efforts to strengthen the prosecution of these crimes, counting also on support from international cooperation. On the one hand, the high percentage of alleged offenders for pre-trial detention casts some doubt on whether this policy change will lead to more convictions or just to more people awaiting trial. On the other hand, the new anti-corruption legislation (Law 004) appears to make a symbolic statement, as it is without a real impact on practice, as the statistics in Tables 2 and 4 show.
Both examples support Tonry's thesis that, even when penal policies become harsher, what matters is whether practices change or not because, if it is the latter, policy changes can be neutralized (Tonry 2007: 1 and 12). Hence, the enactment of a law with more severe sanctions itself is not a conclusive indicator for harsher policies being implemented.
We saw that the new Plurinational State of Bolivia has enacted multiple laws establishing new crimes, often answering to claims from different interest groups. Nevertheless, we do not think that this is any proof of penal populism-from above or from below (Sozzo 2013(Sozzo , 2014Tonry 2007: 1, 12). 43 We rather think that this is evidence of a process of growing self-determination of crime policy. Bolivian officials use the term of 'nationalization' when they talk about the change affecting the anti-narcotic policies (CONALTD 2011: 16).
We argue that self-determination is a more fitting term that can be used in reference to crime policy in general, because the policies contained in laws have always been national but not necessarily self-determining. Indeed, at the beginning of the century, about 60 per cent of all prisoners were doing time for drug-related offences. Today this percentage has dropped to about 20 per cent. Nowadays, other crimes, such as robbery and rape, make up for the similar proportions (23 per cent and 20 per cent, respectively). New crimes-for example, fraud in sports, violence against women or political corruption-have been created to clearly respond to national problems and political claims. Tonry (2007: 32) argues that 'few officials in any Western country would argue that public attitudes and beliefs are inappropriate considerations in setting general policies, within certain limits'. We argue that the same applies in the case of Bolivia. We believe that the country has chosen to use public opinion rather than international pressure as its main source for legitimizing changes of crime policy that have led to the increase in the prison and remand populations. Nevertheless, we also argued that the punitive intent of some of these reforms were neutralized in practice. Part of the steep increase of the Bolivian prison population can be understood as a result of institutional deficiencies and increases in remand rather than as direct consequences of a more 'punitive' or 'populist' legal policy.