The Geography of Justice: Assessing Local Justice in Colombia’s Post-Conflict Phase Introduction Over the last ten years in Colombia, there has been an increasing interest in studying the role of institutions relating to both the country’s economic development and the construction of democracy. According to substantial literature on the subject, institu- tions have a central role in democracy-build- ing. However, various parties question the dimensions of this role and the way it relates to other variables such as violence, culture, and the market (Acemoglu & Robinson 2012; González, Bolívar, & Vásquez 2003; Gutiérrez 2010; López 2010; Rodríguez & Portes 2012; Romero 2003). This article analyzes the role of state insti- tutions at the local level in Colombia. The focus is not on exploring national security or political regime issues (although these top- ics are mentioned), but rather on the impli- cations stemming from state incapacity or even deficiency in terms of the population’s rights. The argument goes as follows: popula- tions who live in the peripheral areas of Colombia, where state institutions are weak or non-existent, are permanently vulnerable to violations because they are not served by institutions that uphold their rights. These are not exceptional cases of communities lost in the national geography, but rather are part of a widespread rights violation. According to calculations presented in The Right to a State (Garcia Villegas and Espinosa 2013), the State is not present in 60 per cent García-Villegas, M and Espinosa, J R 2015 The Geography of Justice: Assessing Local Justice in Colombia’s Post-Conflict Phase. Stability: International Journal of Security & Development, 4(1): 41, pp. 1–21, DOI: http://dx.doi.org/10.5334/sta.gc RESEARCH ARTICLE The Geography of Justice: Assessing Local Justice in Colombia’s Post-Conflict Phase Mauricio García-Villegas* and Jose Rafael Espinosa† This article combines descriptive empirical research with theoretical reflections to offer policy guidelines on what the role of local justice institutions in Colombia’s post-conflict phase should be. The article is divided into two parts. In the first, we present empirical evidence to illustrate the ways in which justice operates differently across the territory. In addition to illustrating these disparities, we also demonstrate the connection between these disparities and some phenomena relevant to understanding the Colombian conflict. Based on these findings, the second part of this article defines the state-building challenge confronted by the Colombian State during the post-conflict phase. Following this part, we propose a solution to this state-building challenge: the State must adopt a combination of efficacy and justice, and we provide guidelines on how a post-conflict justice system can operate to achieve that combination. * Professor at Universidad Nacional in Bogotá, Colombia; Fellow at Institute for Legal Studies at the University of Wisconsin-Madison; Founding Member of the Center for Law, Justice and Society-Dejusticia, Colombia mvillegas@wisc.edu † Associate Researcher at Center for Law, Justice and Society-Dejusticia; Graduate student at Harris School of Public Policy – University of Chicago, United States jespinosa@dejusticia.org stability http://dx.doi.org/10.5334/sta.gc mailto:mvillegas@wisc.edu mailto:jespinosa@dejusticia.org García-Villegas and Espinosa: The Geography of Justice Art. 41, page 2 of 21 of the national territory, an area that is inhab- ited by more than six million people. These people lack what Hannah Arendt called ‘the right to have rights’ – the right to live in a political community in which their rights are recognized and protected.1 The central state owes a historical debt to these people which must be paid and, from a constitutional point of view, represents a systematic and flagrant violation of citizen rights similar to (or possibly worse than) violations that the Constitutional Court’s jurisprudence defines as an ‘unconstitutional state of affairs.’ In The Right to a State we used the expres- sion ‘institutional apartheid’ to describe what is currently happening in broad areas of the national territory in which the State is unstable or non-existent. As a result, the population is affected by the fact that their rights are not recognized or protected. Although the situation does not correspond exactly to the origin of the word ‘apartheid,’ which is closely linked to the South African context, the book defends the idea that even though the situation in Colombia is differ- ent, the discriminatory effects are similar. Institutional abandonment of large portions of the country results in segregation of the people who live there. While in South Africa segregation was based on the prevalence of one race over another, in Colombia it is founded on the prevalence of some regions over others. In this article we answer the following question: given the current peace process between the Colombian Government and the guerrilla group, and given that eventual agreements from that process will effect an enormous institutional challenge in the peripheral regions of Colombia, how can the local state be strengthened in order to address that institutional apartheid? In addi- tion, what role does justice play in such an institutional strengthening plan? This article is divided into two parts. In the first, we present empirical evidence to illustrate the ways in which justice operates differently at the regional level. In addition to illustrating geographic disparities in that operation, we also demonstrate the con- nection between the disparity and some phenomena relevant to understanding the Colombian conflict. Based on these findings, the second part defines the state-building challenge confronted by the Colombian State. We defend the idea that in order to confront this challenge, the State must adopt a combination of efficacy and justice, and we provide guidelines on how a post-conflict justice system should operate. The performance of (formal) local justice In this first part, we present the results of an empirical investigation that demonstrates the disparate functioning of (formal) justice in the country. We constructed an indicator on local justice performance that evaluates both presence and efficacy of formal justice institutions. Presence This first section evaluates the presence of justice representatives in the Colombian regions. It is a formal measurement that takes into account the presence of justice representatives in the territory and gives an idea of the geographic distribution of judges in Colombia, but it does not judge the effi- cacy or efficiency of their presence. The following Map 1 shows the distribu- tion of judges in the Colombian territory in July 2012, taking into account the popula- tion and size of the municipality.2 In order to address the dispersion of the data, we grouped municipalities based on their differ- ences from the national average.3 In this case, the municipalities ‘without data’ are non- municipalized and their jurisdiction is the responsibility of neighboring municipalities. As this Map 1 shows, the presence of judges is concentrated in the center of the country, plus two small niches: in the north around Cartagena and Barranquilla, and in the south in Pasto and the surrounding area. The presence of justice representatives should not only be analyzed in terms of quantity, but also in terms of quality. The García-Villegas and Espinosa: The Geography of Justice Art. 41, page 3 of 21 Map 1: Presence of Judges. Source: Superior Judicial Council (Consejo Superior de la Judi- catura – CSJ), National Administrative Statistics Department (Departamento Administrativo Nacional de Estadistica - DANE), Agustin Codazzi Geography Institute (Instituto Geografico Agustin Codazzi– IGAC) (2012). García-Villegas and Espinosa: The Geography of Justice Art. 41, page 4 of 21 Colombian State created the justice career track to guarantee the quality of justice rep- resentatives. In theory, only the best judges occupy public positions and rarely is (e.g. in the case of temporary licenses) the office occupied by provisional judges. The justice career track mechanism also serves to guar- antee the judges’ independence with respect to political representatives and other judges. This political independence is due to the fact that provisional representatives are named directly by the magistrates of their respective hierarchical superior tribunal.4 The judicial career track – similar to many public policies – has had a different impact in the various regions. We constructed a database of the provisional nature of judicial offices in the country based on information collected from administrative courts of the sectional judicial councils.5 The following Map 2 illustrates how pro- visionality is distributed across the country. This map demonstrates two notewor- thy aspects of the way in which the justice career track operates in Colombia. First, the justice career track does not function in a great part of the territory. According to information provided by the sectional judicial councils, temporary judges occupy 31 per cent of justice posts.6 In regional terms, the percentage of all provisional judges is equal to or greater than 80 per cent in 425 municipalities (from a total of 1103). Second, the map shows that provi- sionality is not randomly distributed across the country. As in previous maps, provision- ality – associated as much with a lack of merit as with the risks of clientelism – tends to concentrate in peripheral municipalities. Efficacy Second, we measure the efficacy of the jus- tice system at the regional level, understood as the system’s capacity to reach expected objectives. We limit ourselves to the efficacy of the criminal system for two reasons: first, the efficacy of criminal justice (as opposed to civil justice, for example) is a good indicator of state capacity to confront other social actors who compete for the monopoly on the use of force (Garcia-Villegas 2008); sec- ond, the criminal system is the only process- ing system in Colombia to offer a modicum of trustworthy information about its own management.7 In addition, we are focusing on the justice system’s efficacy in sanctioning homicides that occurred in each municipality. There are three reasons why we focus exclusively on homicides. First, they are the most seri- ous crimes and as such should be prioritized in judicial processing. Second, homicides are the most registered of crimes (as it is more difficult for a homicide to go unnoticed than a scam, for example). Finally, by limiting our- selves to homicide, we dodge an objection that could be made about our indicator, that is, whether the desirable result of processing is always a sentence. For some crimes, such as stealing a carton of milk, this is a point of discussion in judicial and criminal policy debates, as well as in the philosophy of law. In the case of homicides, however, the attor- ney cannot avoid formulating a charge, nor the judge a sentence, because of the serious- ness of the crime. To measure the performance of justice, we have therefore constructed an ‘efficacy rate.’ This indicator measures the percentage of cases entered into the system for which a sen- tence is formulated. In addition, when evalu- ating the number of sentences compared to the total entries, it takes into account the magnitude of the justice demand in the municipality.8 It is necessary to make three methodo- logical clarifications before moving on to the results. First, of the total municipalities, those for which there is a lack of complete information (number of entries, formulation of charges, and formulation of sentences), or those whose information suggests that there was an error (e.g. more sentences than charges or entries) were excluded. This reduced the number of municipalities from 1103 to 848. García-Villegas and Espinosa: The Geography of Justice Art. 41, page 5 of 21 Map 2: Provisional Nature of Justice Offices. Source: Sectional judicial councils (Consejos sec- cionales de la judicature) (2013). Authors’ own calculations. García-Villegas and Espinosa: The Geography of Justice Art. 41, page 6 of 21 Second, due to the availability and quality of information, we limited ourselves exclu- sively to criminal processes under the accu- satory penal system (Law 906 of 2004). For each municipality, we added existing infor- mation from the system’s period of validity, and based the efficacy analysis on informa- tion for each municipality. 9 Third, we measured the level of efficacy in the sanction of homicides that took place in a determined municipality. This does not nec- essarily mean that the criminal process took place in that municipality. Instead, it means that we are not measuring justice that physi- cally operates there, but rather justice that is effectively applied there. This does not repre- sent a problem for the research. We are inter- ested more in measuring the extent to which the system fulfills its purpose in the territory than in physical institutional presence. Map 3 shows the spatial distribution of the efficacy indicator.10 As the map shows, ‘high’ and ‘very high’ efficacy tends to concentrate in the depart- ments of Huila, Casanare, Tolima, the cof- fee region, to a lesser extent in Antioquia, Cundinamarca, Boyacá, Santander, and a few municipalities in Bolivar and Atlántico. Meanwhile, ‘low’ and ‘very low’ efficacy is concentrated in Nariño, Putumayo, Casanare, Meta, Vichada, Chocó, southern Córdoba, Catatumbo, southern Bolívar, and La Guajira. The majority of municipalities with ‘low’ and ‘very low’ efficacy are in corridors that share various features including: i) presence of armed groups; ii) areas of illegal crops; iii) areas of input transport for the produc- tion of cocaine; iv) presence of illegal mining; and v) low quality of institutions. The most noteworthy trait of the map is that these fac- tors operate in a context of local-level judi- cial inefficacy, which allows these activities to continue at low criminal cost. These maps do not always allow for defini- tive conclusions to be made about the insti- tutional capacity in a municipality. There are situations in which the justice and munici- pal government indicators are very good, but a more detailed analysis of the context, the history, and the internal power relations of the municipality show that institutional power only appears strong. For example, the department of Casanare – an example of institutional seizure or ‘coopted reconfigu- ration of the State’ – seems to have strong justice indicators, but in reality, they are not as strong as might be expected (Garay, De León, & Salcedo, 2010; García Villegas & Revelo Rebolledo, 2010; Hellman, Jones, & Kaufmann, 2000). The performance of local justice Until now, this paper has examined how the justice system’s presence and efficacy in the prosecution of homicides are distributed. Based on this analysis, we constructed a global indicator of local justice performance, taking these two dimensions into account. The indicator is composed of two variables: presence of judges and an efficacy rate for each municipality. The indicator includes a scale from 0 to 100, where 100 corresponds to the highest score and 0 to the lowest. This score was obtained by weighting the qualifications from the two components of the index,11 assigning a weight of 60 per cent to efficacy and 40 per cent to presence. We made this methodological decision because we believe that even though presence is an important dimension of justice, in many municipalities this presence is more nominal than real – just because there is a judge in a munici- pality does not mean that he/she imparts justice. This was one of the conclusions of Judges without a State (García-Villegas 2008). In this sense, a municipality with fewer judges who are more effective would be preferable to one with more judges who are less effective. The different possible scores of the index are grouped into five categories, which cor- respond to the five levels of local justice per- formance (see Annex 1). The Table 1 shows how the municipalities at different levels of the index are distributed. García-Villegas and Espinosa: The Geography of Justice Art. 41, page 7 of 21 Map 3: Efficacy. Source: AGO (Fiscalía General de la Nación). Authors’ own calculations. García-Villegas and Espinosa: The Geography of Justice Art. 41, page 8 of 21 As Table 1 shows, the majority of munici- palities (356) are distributed in the medium level of local justice performance.12 The low- est category is comprised of 136 munici- palities, a number greater than that of the two highest categories put together: 97 municipalities are in the ‘high’ category while 20 are in the ‘very high’ category. The table also shows a noteworthy result: with respect to the national average, more than a third of the municipalities of the country (375 of 848) have ‘low’ or ‘very low’ levels of performance. The following Map 4 shows these munici- palities spatially distributed by category. This map highlights patterns seen in previous maps. The municipalities with ‘medium,’ ‘high,’ and ‘very high’ levels of performance are found mainly in the center of the country, and to a lesser extent in the Caribbean Coast, Valle del Cauca, and the departments of Meta and Casanare. On the other hand, the municipalities with ‘low’ or ‘very low’ levels of justice perfor- mance tend to be located in the peripher- ies, especially in the departments of Nariño, Chocó, Putumayo, Caquetá, and Vichada. In Córdoba, Magdalena, Atlántico, and Sucre the ‘low’ and ‘very low’ performance of jus- tice coincides with the most southern areas of the departments, which are characterized by difficult access, floods and challenging transport channels which isolate the munic- ipalities from regional centers. Different expressions of local justice performance in the territory Justice does not operate in isolation in a municipality. On the contrary, its perfor- mance is related to the context in which the everyday workings of the justice system take place. This section analyzes the relation- ships between the local justice performance indicator and various socio-economic and violence-related variables that we thought relevant. We are interested in identifying differences when comparing the justice indicator with these other variables in dif- ferent categories of the municipality. We can thereby observe that municipalities with a ‘low’ or ‘very low’ level of local justice per- formance present similar behavior in other institutional dimensions. The disparate per- formances of justice are related to the exist- ence of coca crops, forced displacement, presence of illegal armed groups, percentage of Afro-Colombians, and percentage of indig- enous people.13 Graph 1 illustrates the relationship between the average density of coca crops in a group of municipalities and local jus- tice performance.14 The difference in crops between the five categories of local justice performance (very high, high, etc.) is strik- ing. Municipalities catalogued with ‘low’ and ‘very low’ performances in terms of local justice have an average of twelve per cent and thirteen per cent crop density respec- tively (in other words, on average twelve to thirteen per cent of each square kilometer in these municipalities is dedicated to coca crops), while in other municipalities the density does not go above two per cent. The performance of justice is weak where the density of coca crops is greater. This does not necessarily mean that in municipalities with a very low justice per- formance there is more drug trafficking than in other municipalities. There are sim- ply more crops. The differences in forced displacement15 between different categories of justice are quite large, as seen in Graph 2. The Level Number of municipalities % Municipalities Very low 136 16% Low 239 28% Medium 356 42% High 97 11 % Very high 20 2 % Total 848 100 % Table 1: Municipalities by levels of local jus- tice performance. García-Villegas and Espinosa: The Geography of Justice Art. 41, page 9 of 21 Map 4: Local justice performance. Source: SJC/AGO, own calculations. García-Villegas and Espinosa: The Geography of Justice Art. 41, page 10 of 21 municipalities with very low local justice per- formances have more than double the rate of displacement when compared to those clas- sified in the other four categories. The statis- tical results allow us to conclude that a lower level of local justice performance implies a greater number of displaced people. The results found in relation to the coca crop dynamics and to displacement are convincing and reveal two facets of the Colombian armed conflict. On the one hand, the results allow for the identification of one of the illegal armed groups’ greatest eco- nomic sources, and on the other, they reflect the vulnerability of rights generated by the conflict. Tragically, local justice is weaker in areas with more coca crops and where dis- placement rates are greater. These trends gave notable results. In gen- eral, municipalities with ‘high’ and ‘very high’ levels of justice performance had a lesser presence of illegal armed groups. In Graph 1: Density of crops and levels of local justice performance. Source: SIMCI 2012; Authors’calculations. Graph 2: Average displacement rate by categories of local justice performance. Source: Vice President’s Office 2005–2011. Authors’ calculations. García-Villegas and Espinosa: The Geography of Justice Art. 41, page 11 of 21 municipalities with very high justice per- formance rates there was no paramilitary or BACRIM (bandas criminales, former paralim- itary groups) presence from 2000 to 2012. The guerrilla presence comprised almost a quarter of the overall presence of illegal armed groups in municipalities of ‘medium’ justice performance rates and almost a fifth of the presence of illegal armed groups with ‘low’ or ‘very low’ performance rates. In terms of guerrilla presence, there is a clear negative relationship: the greater the local justice performance, the lower the num- ber of years of guerrilla presence. In terms of paramilitary and BACRIM presence, the result is different, because the municipali- ties with weaker justice rates are different from those municipalities who experienced the presence of these groups for a greater number of years. Except in municipalities with high levels of justice performance, the paramilitaries and the BACRIM were present no matter the institutional justice capac- ity. In general terms, whereas the guerrillas aim to fight against the State, the paramili- taries appropriate or strategically use the State (García Villegas & Revelo Rebolledo 2010; López 2010; Romero 2003; 2007). It is also worth noting that the BACRIM pres- ence is greater than that registered by the paramilitaries, which suggests that the demobilization process did not reduce, but rather increased, the presence levels of these groups in the municipalities, as illustrated in Table 2. These two results allow us to argue that the presence of the guerrilla tends to congre- gate in municipalities with weaker justice, while the paramilitaries or the BACRIM have a more homogenous presence in municipali- ties, regardless of their justice performance levels. This coincides with results shown in previous research (García-Villegas 2008). The bias in guerrilla presence and relative homogeneity in paramilitary and BACRIM presence allows us to address each group’s process of capturing institutions (or not). The guerrillas have skirted to the periphery of the country, where institutions are weaker, while the paramilitaries have managed to stay in important urban areas where they have used justice to achieve their military and eco- nomic objectives. In terms of ethnic composition, there are important differences between the five categories of the local justice index. Municipalities with ‘very low’ justice perfor- mance levels have greater levels of indig- enous and Afro-Colombian people. Graph 3 shows that municipalities located in the low- est justice category have an average of 16 per cent and thirteen per cent of Afro-Colombian and indigenous populations respectively, while in other justice categories the popula- tion does not rise above nine per cent and six per cent respectively. The municipalities with a very low level of justice performance have greater numbers of displaced people, greater hectare den- sity of coca crops, and a greater number of Category of justice Number of years with illegal groups presence 2000–2012 Guerrilla Paramilitaries (2000–2006) BACRIM (2007–2012) Very high 1.3 0 0 High 2.9 1.03 1.8 Medium 4.2 1.68 2.69 Low 5.3 1.66 2.3 Very low 5.1 1.4 2.3 Table 2: Presence of illegal armed groups broken down in years between 2000 and 2012. García-Villegas and Espinosa: The Geography of Justice Art. 41, page 12 of 21 indigenous and Afro-Colombian people in their population. A similar phenomenon occurs in terms of the presence of the guer- rilla groups, which is found less often in municipalities with ‘high’ and ‘very high’ lev- els of justice performance. From Institutional Apartheid to the Constitutional State: Local Justice in the Post-Conflict Phase Political institutions in Latin America are embedded in a social and institutional context much different from the modern European state model, which the Latin American countries have inherited. This has, in part, resulted in the common dispar- ity between norms and social realities that has characterized Colombian institutional life since time immemorial. The existence of the modern state can be claimed when the fulfillment of three conditions or basic characteristics occurs: efficacy, legality, and legitimacy (Bobbio 2005). The State in Latin America has deficiencies in each one of these three characteristics. For example, state effi- cacy is not always attained legitimately and, on occasion, the legitimate state is incapable of presenting itself to powerful individuals and groups, be they legal or illegal. These situations are particularly common in certain spaces, generally (but not exclusively) located in the periphery of the national territory. Based on a constructivist, tiered, and prag- matic vision of known state institutions (García-Villegas 2014; Migdal 2011; Portes & Smith 2010; Rodríguez & Portes 2012), we will now try to respond to the following question: how can public policy address the enormous challenge of incorporating un- institutionalized areas of the country into the social rule of law? What path must be taken to respond to this challenge?16 This question is even more relevant today as Colombia faces a post-conflict phase in which opportunities to carry out historic reforms will present themselves, as a result of the eventual peace agreements. In effect, the draft agreements published until recently imply large-scale social, economic, and politi- cal development initiatives in the periphery of Colombia. In concrete terms, the agree- ments foresee large-scale rural reforms, an increase in political participation, and the elimination of drug trafficking. To apply these eventual agreements, however, there must exist local institutions with the capacity Graph 3: Percentage of Afro-Colombians and indigenous people by municipality by category of local justice performance. Source: Dane 2012. Authors’ calculations. García-Villegas and Espinosa: The Geography of Justice Art. 41, page 13 of 21 to implement the public policy necessary to enforce the objectives of the agreements. The design of these institutions and public policies is still to be defined and will be the subject of intense debate in years to come. Discussions on the post-conflict phase often assume that these peripheral areas already have the institutional capacity to implement the agreements, which is not true. Until now, discussions on transitional jus- tice in Colombia have focused mainly on the dilemma between justice and peace. That is, how to resolve the tension between the need for accountability by those ultimately responsible for crimes, and the need to com- promise with these actors so that they aban- don weapons and enter politics. This debate, though fundamental, has been restricted to a limited definition of justice. It under- stands justice exclusively as the prosecution of those responsible or the guarantee of the rights of the victims, and ignores the need for justice dedicated to resolving daily con- flicts that present themselves in transitional areas. This is problematic because it ignores the objective of social peace, a post-conflict goal closely related to the success of the tran- sition. For the post-conflict phase, we have to think not only about the kind of justice we implement to deal with those respon- sible from all sides of the conflict, but also about which institutions will be required to consolidate peace in the regions through the resolution (formal and/or informal) of social conflicts. In order to respond to these questions, we first look at three features of the modern state that were mentioned at the beginning of The Right to the State: efficacy, legality, and legitimacy. Which path should we take, in institutional terms, to arrive at an efficient, legitimate, and legal state? Given that the legitimacy of the consti- tutional state is attained through legality and protection of citizen rights, here we propose the concept of ‘lega-timacy’ which involves and coordinates both purposes (legality and legitimacy). We assume that the constitutional state or, more precisely, the social rule of law, pursues two fundamen- tal or interdependent objectives: 1) efficacy, and 2) ‘lega-timacy.’17 Each of the two axes in the following graph (horizontal and vertical) represent the two objectives mentioned in a continuum from low to high. Based on these two axes, it is possible to compare and contrast the differ- ent scenarios featuring state institutions, as shown in the following graph. Types of Institutions The graph above demonstrates the wide range of possible scenarios, from extreme institutional weakness to extreme state authoritarianism, passing through the social rule of law (ROL), which possesses a type of contained strength. This graph also illus- trates how the vulnerability of many sectors of the population can originate not only in the excess of state power but also in its defi- cit. The constitutional state is therefore both an effective and a contained power. The ideal scenario is represented in the upper right, where there is a maximum of both variables. ROL is found there, with a greater degree of efficacy and legitimacy.18 At the other end of the spectrum – the lower left – we find the opposite of ROL – extremely weak institutions with little ‘lega-timacy.’ This area represents institu- tional apartheid (among other situations), García-Villegas and Espinosa: The Geography of Justice Art. 41, page 14 of 21 a situation characterized by vulnerability and discrimination against the population. Not all of what falls under this area of extreme instability can be characterized as institu- tional apartheid, due to the fact that not all institutional weakness implies a severe degree of discrimination. For example, in some isolated areas inhabited by indigenous communities, institutional incapacity is sup- planted by strong community cohesion. The two intermediate cases are explained in the following way: the first, in the upper left, combines a maximum of ‘lega-timacy’ with a minimum of efficacy. This situation is more theoretical than real, given that it is dif- ficult to conceive of a society under a nominal state (that has no power) and in which events occur in a legal and legitimate way (except perhaps something close to anarchism). In the far lower right we find authoritarian institutions in which a maximum exercise of power is combined with minimum ‘lega- timacy.’ This type of state maintains peace and imposes order within the territory, and is able to guarantee certain rights. However, it does so without legitimacy or legality. Many attempts to bring the State to the Colombian periphery have been reduced to the creation of this type of state. It should also be added that these are extreme cases and that, as the graph shows, it is possible to imagine intermediate cases (in fact, the Latin American and urban ver- sions of ROL, in general terms, could be cat- egorized as in the upper right square, but at a point close to the center of the graph). It is also important to acknowledge that it is not always easy to spatially identify each case. Here, we have said that institutional apart- heid is a common phenomenon in the geo- graphic periphery of the country. However, this phenomenon exists in the cities, in mar- ginalized neighborhoods and rural areas that we could call internal peripheries, where the social rule of law is relatively strong. In addition, it is important to note that institutional apartheid is not always apparent and easy to detect. There are cases in which legality, and some legitimacy, are found in regions where institutions have been cap- tured or coopted by armed or illegal actors. This is the case of the mafia-state, which not only adopts the appearance of legality, but also defends and operates with the law while distorting its meaning and using legality to favor illegality. It is even possible that in these cases, the population is overtaken. This is an example of a state outside of constitu- tional order, which nevertheless follows its judicial forms and routines. It also conceals institutional capture which results in domi- nation and oppression, perhaps worse than that of the authoritarian state. We therefore return to the question with which we began this final section. How can we transition from an institutional apartheid state to ROL? We have identified five paths, each one of them represented by a line that unites extreme institutional weakness to ROL. The first of them, path A (the gray continuous line) is defended by military regimes in Colombia and Latin America, and conceives of the con- stitutional challenge as a task that starts and ends with pacification of the territory, as the imposition of order without legitimacy. This path is not only unacceptable from an ethical and legal point of view (as it sacrifices the rights of one part of the population to improve the conditions of the rest), but also from an empirical point of view. Experience shows that there is no guarantee that this option will be successful. Indeed, the imposi- tion of force without human rights, democ- racy, or control instead seems to reproduce violence and institutional weakness in the Latin American context. The second path, B (in dotted gray), pro- poses to reach both objectives of ROL but based on sequential logic in which efficacy comes first. This path seeks to pacify the territory first (expelling or defeating the enemy) and then, only when this objec- tive has been achieved, concerns itself with the legality and legitimacy of the State. The National Territorial Consolidation Plan García-Villegas and Espinosa: The Geography of Justice Art. 41, page 15 of 21 adopted by the national government seems to have adopted this model (García Villegas & Espinosa Restrepo 2012; Palou 2011). Once pacification is achieved, a transition from military to civilian government is realized. The protection of the citizen is of the utmost importance in this transition. Finally, when this protection is consolidated, the third phase of stabilization begins and economic, social, and institutional development can be consolidated. Although path B is attractive from a mili- tary standpoint and in terms of creating order, it has been demonstrated as non- conducive – and even counterproductive – to the creation of ROL. In many cases, this first stage of military domination translates to serious human rights violations. The third path, C, originates in a logic contrary to the others. Instead of prioritiz- ing efficacy, this path emphasizes legitimacy. This path consists of introducing institu- tional reforms with evident legitimacy in social environments characterized by institu- tional crisis. The expectation is that popular support for the new institutions will result from these reforms, and combined with their intrinsic legitimacy, will serve to strengthen state institutions. This perspective inspired some of the institutional designs defined by the 1991 Constitution, in particular, those related to decentralization and citizen par- ticipation mechanisms (Gutiérrez 2010; Sánchez & Chacón 2006). This path is a type of ‘democratism’ according to which demo- cratic routines are enough to achieve social rule of law. This assumes that democratic legitimacy could remedy weaknesses in insti- tutional capacity through the passage of time and with the help of popular mobiliza- tion. However, after more than twenty years of implementing the Constitution, the limits and even dangers of this view have become evident. It is not only probable that ‘lega- timacy’ did not bring efficacy in these areas but that even worse, new democratic institu- tions were captured and corrupted by crimi- nal organizations (García Villegas & Revelo Rebolledo 2010; Gutiérrez 1996, 2010b; López 2010). It is worth noting the affirma- tion of Albert Hirschman (1977) according to whom, ‘not all good things come together.’ The fourth path, D (see the continuous line), supposes that it is possible to simulta- neously construct all three levels of the ROL building. This is the best path from the per- spective of constitutional axiology (García Villegas & Uprimny 1999). In practice, how- ever, the situation poses many difficulties. This path is feasible in territories in which the State is more nominal than effective and in which there is no competition from another armed actor. However, what hap- pens when the State finds itself in a territory occupied by an armed actor that imposes its law on inhabitants? In these cases, the first task of the State consists of gaining military control and monopoly on the use of force in the area. This is what the international litera- ture widely recommends: in the construction of the State, security must be prioritized, but with legality (Fukuyama 2004; Jensen 2008; Ottaway 2002; Rubin 2008). Opening mech- anisms of broadened political participation without consolidating a minimum of efficacy has translated to an increase in violence. The literature, however, falls short of indicating how to achieve this ideal combination of security and legality. How can this minimum of security be achieved within the constitu- tional framework? Perhaps the response to this question would be the creation of a fifth path, denoted as E. Path E would recognize the practical need for sequencing between the two objectives pursued (efficacy and legality / legitimacy), but would condition its success on the effec- tive presence of the justice system in this transition. We have stated previously that in Colombia, we must try to achieve three features of ROL at the same time. This is a normative ideal that should be maintained. However, in prac- tice, absolute concomitance is frequently a difficult goal to achieve. This is due to the fact that there is a conditionality on these García-Villegas and Espinosa: The Geography of Justice Art. 41, page 16 of 21 factors.19 Path E endeavors to respond to this difficulty. It demands that military action tar- geting the reduction of armed actors in the territory is accompanied by strong judicial control directed at the protection of funda- mental rights of those who live in the terri- tory in which the military operations take place. Some aspects of the other objective (‘lega-timacy’), such as broadened citizen participation, social investment, or the exer- cise of representative democracy, could be temporally different but always under the vigilance and control of the justice system. In this way, path E paves the way to a type of spatially and temporally exceptional state. In the literature on state-building in transi- tional contexts, state formation is always pre- sented as a process whose objective consists of creating an effective state that imposes itself on powerful actors who question its power. This means that in general, it collides with the objective of achieving an institu- tional framework characterized by legitimacy and legality. The National Consolidation Plan, formed some years ago in Colombia with the idea of bringing the State to periph- eral territories affected by the armed con- flict, seems to obey this supposition (García Villegas & Espinosa Restrepo 2012; Palou 2011). However, the collision between these two objectives is not inevitable. Our hypoth- esis is that the strengthening of local justice can serve as a bridge for the concomitant achievement of both purposes. Our hypothesis is supported by the fact that justice has the ability to ‘domesticate’ the force necessary for the State to impose itself on armed actors who operate in an area. In this process of the domestication of force, justice provides legitimacy to the State imperium. It is improbable that in this early stage, all expressions of the social rule of law are present. The defense of social rights or participatory democracy, for exam- ple, requires a degree of institutional devel- opment that it is not possible to achieve in those initial moments. However, nothing impedes justice from accompanying this first state intervention. This is not only possible but also limits the use of force and submits the coercive power of the State to the consti- tution and the law. Justice has strong links to the two sides of the tension that we have been analyzing – efficacy and ‘lega-timacy.’ On the one hand, it moderates the force of efficacy by sub- mitting the State’s need to impose itself on enemies to legal requirements. On the other hand, by avoiding the violation of human rights through the use of force, justice has a legitimizing effect on state action. This effect comes not only from the legality imposed by justice, but also from the democratic char- acter of judges and from the fact that they are defenders of the constitution and of law, which are essentially democratic normative parameters. In addition, the judge not only provides legitimacy to the state-building process, but also offers efficacy by contributing to the overcoming of social conflicts. When a con- flict must be mitigated, state intervention should target the pacification of social rela- tionships. In that sense, a justice system that manages to resolve daily conflicts (like neigh- borly issues, or property claims, etc.) also helps to construct the social fabric necessary for the State to operate normally. In a post- conflict phase, one of the first steps should be to establish successful conflict resolution mechanisms. To conclude, we would like to note (albeit tentatively, as this topic requires greater research and reflection) that local justice in the post-conflict phase needs at least four conditions to achieve its proposed objectives. First, justice-related public policy should adopt a regional focus. Colombia is a socially, culturally, and geographically complex coun- try, which requires a similarly area-specific approach to policy. Justice-related public policy should take local justice demands into account, as well as institutional capac- ity, cultural traditions, the legacy of con- flict, and geographic space. Regional studies must be conducted to identify these justice García-Villegas and Espinosa: The Geography of Justice Art. 41, page 17 of 21 demands (frequent conflicts, motivation to adopt different models, etc.) and, depending on these needs and contextual limitations (institutional, cultural, economic, etc.), spe- cific policy should be adopted in response to the demand. The ‘one size fits all’ approach should be replaced by a focus on multiple sizes and models that integrate Colombia’s regional diversity. Path E is therefore not an immobile path. By combining efficacy with ‘lega-timacy,’ the path can in some cases be closer to the axis of efficacy while in others be closer to that of ‘lega-timacy.’ Second, a regional focus on justice requires modifications to the center of jus- tice administration – the central state – in Bogota. Colombian justice is too inbred, and lacks technical competency and account- ability. The Judicial Council should be more plural, open, and managerial in order to be more democratic, accountable, and thereby more legitimate in the eyes of the citizenry. In addition, the evaluation and design of pol- icy should have its technical foundations in quantitative and qualitative data. Third, the abilities of judges who work in peripheral areas must be strengthened. Justice administration has traditionally sent judges with the least knowledge and expe- rience to the peripheral areas of the coun- try, particularly in conflict-ridden areas. However, the opposite should occur, as it is precisely those locations that need the most able judges who can contend with difficult problems. In addition to the creation of new offices and positions, a new incentive system must be established to stimulate qualified people to work in these regions. In addition, the justice system should employ managers, psychologists, sociologists, social workers, and others who are essential in identifying and addressing conflicts in an efficient and effective way. Fourthly, the term ‘justice system’ should be understood broadly. This is a corollary of the points above. The system’s institutional capacity must be strengthened (conditions 2 and 3) but with a regional focus (condition 1). This implies that it is possible to conceive of local justice institutions that are not exactly the same as those that operate in the capi- tal of the country. The debate between for- mal and informal justice could thereby be overcome by a harmonious collaboration between the two. The justice system should allow for justice standards and informal local procedures (giving equal responsibility to local leaders or conciliators) that are both democratic and efficient, to continue operat- ing and resolving conflicts. In these cases, the task will consist of coordinating standards and procedures with other state entities.20 We believe that a justice system that adopts these four principles can usefully serve as a hinge between two objectives that can be contradictory in conflict contexts: building peace and building the State. The challenge consists in creating innovative policies that brings the four principles together. We believe this to be possible only when a constructivist vision of the problem is adopted, and when there exists a fair medium between favoring all contexts and realities (and saying that the law should adapt to reality) and imposing an ideal vision of justice without reservations. Competing Interests The author declares that they have no com- peting interests. Funding Statement Part of this research was funded by The Kingdom of the Netherlands. Notes 1 Arendt had v ictims of totalitarian states in mind (1982). However, the vulner- ability that characterizes the absence of the right to have rights also occurs when there is no institutional presence or when institutions are captured by ille- gal actors. Dignity and other rights are at risk when the State is too strong or too weak – when it acquires totalitarian traits or when it loses its capacity to uphold rights. García-Villegas and Espinosa: The Geography of Justice Art. 41, page 18 of 21 2 To control for differences in population and extension of territory, we calcu- lated an indicator of presence by mak- ing the municipalities equal, assuming that all have 100,000 inhabitants and are 100km.2 We also focused on low- hierarchy judges who should be present in the entire territory no matter the size of the municipality. 3 We built categories based on the average, and on distances with respect to the aver- age in standard deviation units. 4 It is important to take into account the variable of institutional design. The mag- istrate election system favors the forma- tion of ‘roscas’ or chains of sycophantism in the judicial branch, in addition to a lack of accountability. Tribunal magis- trates are elected by magistrates in the high courts based on lists compiled by the Administrative Court of the Superior Council (whose members are elected, in turn, by the high courts). This ‘rosco- grama’ or sycophantic network is repro- duced at the local level due in part to the provisional nature of some offices. Given that tribunal magistrates directly name provisional judges, there is a higher risk of reproduction of the network in areas with high rates of provisionality. 5 We understand ‘provisional’ to be any judicial office that – for a variety of reasons – is not occupied by a repre- sentative assigned to the post by virtue of the judicial career track. We do not include decongestion offices because by definition, they tend to be occupied by provisional representatives and do not exist in all parts of the country. 6 If the number of decongestion offices is included, the statistic is even more alarm- ing, because the percentage of provi- sional justice posts at the national level increases to 39 per cent. 7 The first stage in this research consisted of compiling all statistical information available, and conducting the first inves- tigation of how different entities in the judicial branch collect and systematize information. Unfortunately, at the end of this first stage, we concluded that we could not use information from the Superior Judicial Council about entries and exits for different types of processes (admin- istrative, ordinary, criminal, disciplinary) because the data collection methods were not very trustworthy. We therefore chose to use information from the SPOA, the Attorney General’s Office’s information system which, although it has some weak- nesses, is the least problematic of all. 8 The formula we used was as follows: ( )i i i i Total sentences Total entries Total sentences Total entries - - . This equation provides a cluster of val- ues on a scale of −1 to 1 as results, in which −1 represents the municipalities where there was not even one sentence passed for all of the cases entered. 1 rep- resents the municipalities in which there were sentences for all of the entries, and 0 represents municipalities in which sen- tences were made for half of the entries. For more information about the con- struction of the indicator, see the work of Leopoldo Fergusson, Juan Fernando Vargas, and Mauricio Vela, from which we took the idea about formulation (Fergusson, Vargas & Vela 2013). 9 This allows us to control the fact that the criminal accusatory system has become valid gradually over the territory. 10 It is important to note that the categories we used to group the municipalities were constructed based on the results that the municipalities had in the indicator, not on the maximum results possible. In no municipality of Colombia were more than 40 per cent of entered cases sen- tenced between 2005 and 2011. 11 The formula we used was the following: ( ) ( ) 0.6 100,000 0.4 i i i i i i i Number of judges Total population i Total sentences Total entries Total sentences LJI Total entries Extension æ ö- - ÷ç ÷= ç ÷ç ÷çè ø æ ö´ ÷ç ÷ç ÷+ ç ÷ç ÷ç ÷÷çè ø García-Villegas and Espinosa: The Geography of Justice Art. 41, page 19 of 21 We normalized and redefined the pres- ence indicator on a scale from 0 to 100. The corrected efficacy indicator, which previously went from −1 to 1, was also recoded on a scale from 0 to 100. 12 This is of course due to the way we con- structed the indicator, which groups municipalities according to how different they are from the average. 13 The variables of coca crops and forced displacement show the municipal aver- age of each variable between 2005 and 2011, with the goal of making a homo- geneous comparison with the justice indicator. The source on coca crops is the registry conducted annually by the UNODC through the Integral Illegal Crop Monitoring System. The displacement data come from the journal of the Vice President’s Office and correspond to the number of displacements according to the municipality from which the person was expulsed. The presence of illegal groups comes from a database created by the CERAC (Restrepo, Spagat & Vargas 2006). The period of the variable is 2000 to 2012. For the presence of paramilitar- ies and BACRIM, two different periods were analyzed: for the paramilitaries, from 2000 to 2006 and for the BACRIM, from 2007 to 2012. The illegal mining data are the product of research by the People’s Ombudsman’s Office three years ago (2010). The data on indigenous and Afro-Colombian people come from the 2005 census created by the DANE. 14 To calculate the quantity of illegal crops in a municipality, values were taken from 2005 to 2011 and the average was found. Density corresponds to the relationship between cultivated hectares and coca per kilometer squared. A crop density value of 12 hectares per square kilometer means that in this area 12 per cent of each kilo- meter is used for coca cultivation. 15 The rate of displacement per 100,000 inhabitants in the period 2005–2011 was taken as a reference. 16 Balakrishnan Rajagopal (2008) has stron- gly criticized the discourse on ‘rule of law’ as a strategy for the construction of insti- tutions in the post-conflict phase. In gen- eral terms, he sustains that the discourse on ‘rule of law’ is seen as a technical, judicial, apolitical discourse that ends up being useful to escape the political dis- cussion. When people appeal to this dis- course, they hide contradictions between different public policy agendas, such as development and human rights agendas, or security and human rights, which can- not be resolved simply by invoking the ‘rule of law’ as a mantra. The discourse on the ‘rule of law’ hides commitments that should be made between these agendas in order to reach these valid goals. We believe that this criticism is true but does not necessarily imply that the discourse should be abandoned. Rather, it should be conceptualized and examined for specific tensions that could be found in its interior. We accept Rajagopal’s invi- tation to break down what we under- stand by ‘rule of law’ and accept – even if not completely – the tensions between efficacy, legitimacy, and legality in the Colombian case. 17 It is also worth noting that, analytically, legality and legitimacy are two different things. See Bobbio (2005) and Ost and van de Kerchove (2001). 18 Some Scandinavian countries are close to this case (Munck 2009). 19 It is possible to have an effective state (here we adopt a minimal, almost Hobbesian, concept of efficacy and as a result, we are not referring to administra- tive or governmental efficacy) that is nei- ther legitimate nor legal. There can also be a legal state with deficiencies in terms of legitimacy, but it is not possible to find a situation in which democracy and legality exist without an effective state. There can of course exist a lack of all ele- ments, as in areas dominated by armed actors who challenge the State, where García-Villegas and Espinosa: The Geography of Justice Art. 41, page 20 of 21 there is no democratization or protec- tion of rights. Someone could state that armed actors, when they impose order and security, are guaranteeing rights. However, we dismiss this possibility and sustain that law cannot be guaranteed by illegal means. 20 A recent report indicates that the FARC have begun to cede their guerrilla justice administration system to the community action boards (JAC by its Spanish acro- nym) of the areas where they have influ- ence (2015). It would be an error for the State to impose upon the JAC justice sys- tem without first evaluating its strengths and weaknesses, and considering ways to coordinate it with more formal justice procedures. References Acemoglu, D and Robinson, J 2012 Why Nations Fail. The Origins of Power, Prosper- ity and Poverty. New York: Crown Business. Bobbio, N 2005 Teoría general de la política. Madrid: Trotta. Fergusson, L, Vargas, J F and Vela, M 2013 Sunlight Disinfects? Free Media in Weak Democracies. Bogotá: Uniandes. Fukuyama, F 2004 State-building: Govern- ance and World Order in the 21st Century. Ithaca, NY: Cornell University Press. Garay, L J, De León, I and Salcedo, E 2010 Captura y Reconfiguración Cooptada del Estado en Guatemala, México y Colombia. Análisis conceptual de las memorias de la Primera Discusión Internacional sobre Captura y Reconfiguración Cooptada del Estado. Bogotá: Grupo Método. García-Villegas, M 2008 Jueces sin Estado. Bogotá: Siglo del Hombre – Dejusticia. García-Villegas, M 2014 La eficacia sim- bólica del derecho. Sociología política del campo jurídico en América Latina. Bogotá: Random House. García Villegas, M and Espinosa Restrepo, J R 2012 Incorporación institucional de la periferia en Colombia: descentralización, regalías y el Plan Nacional de Consoli- dación. Bogotá: Dejusticia. García Villegas, M and Revelo Rebolledo, J 2010 Estado alterado. Clientelismo, mafias y debilidad institucional en Colombia. Bogotá: Dejusticia. García Villegas, M and Uprimny, R 1999 El nudo gordiano de la justicia y de la guerra en Colombia. In Camacho, A & Buitrago, F L (Eds.), Armar la paz es desarmar la guerra. Bogota: IEPRI, CEREC, FESCOL, pp. 33–70. González, F, Bolívar, I and Vásquez, T 2003 Violencia política en Colombia. De la nación fragmentada a la construcción del Estado. Bogotá: CINEP. Gutiérrez, F 1996 Dilemas y paradojas de la transición participativa. Análisis Político, 29: 43–64. Gutiérrez, F 2010 Instituciones y territorio: la descentralización en Colombia. In 25 años de la descentralización en Colom- bia. Bogotá: Konrad Adenauer Stiftung, pp. 11–54. Hellman, J, Jones, G and Kaufmann, D 2000 Seize the State, Seize the Day: State Capture, Corruption and Influ- ence in Transition by Joel Hellman, Geraint Jones, Daniel Kaufmann. Wash- ington D.C.: The World Bank. Available at http://papers.ssrn.com/sol3/papers. cfm?abstract_id=240555 DOI: http:// dx.doi.org/10.1596/1813-9450-2444 Jensen, E G 2008 Justice and the Rule of Law. In: Call, C T & Wyeth, V (Eds.) Build- ing States to Build Peace. Boulder, CO: Lynne Rienner Publishers, pp. 119–142. López, C (ed.) 2010 Y refundaron la patria . . . De cómo mafiosos y políticos reconfiguraron el Estado colombiano. Bogotá: Corporación Nuevo Arcoiris, Congreso Visible, Dejusticia, Grupo Método y MOE. Migdal, J S 2011 El Estado en la sociedad. Una nueva definición del Estado. (Para superar el estrecho mundo del rigor). In Estados Fuertes, Estados débiles. México: FCE, pp. 15–66 Munck, G 2009 Mesuring Democracy. Bal- timore: The Johns Hopkins University Press. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=240555 http://papers.ssrn.com/sol3/papers.cfm?abstract_id=240555 http://dx.doi.org/10.1596/1813-9450-2444 http://dx.doi.org/10.1596/1813-9450-2444 García-Villegas and Espinosa: The Geography of Justice Art. 41, page 21 of 21 Ost, F and Van de Kerchove, M 2001 Ele- mentos para una teoría crítica del derecho. Bogotá: Universidad Nacional-Facultad de Derecho, Ciencias Politicas y Sociales. Ottaway, M 2002 Nation Building. Foreign Policy, 132: 16–24. DOI: http://dx.doi. org/10.2307/3183443 Palou, J C 2011 Balance de la Política de Consolidación a partir del estudio de 4 regiones intervenidas. Presented at the La Política de Consolidación Territorial del Estado Colombiano: evolución, desafíos, obstáculos y potencialidades, Bogotá. Portes, A and Smith, L 2010 Institutions and National Development in Latin America: A comparative Study. Socio- Economic Review, 1: 1–37. DOI: http:// dx.doi.org/10.1093/ser/mwq018 Rajagopal, B 2008 Invoking the Rule of Law in Post-Conflict Rebuilding: A Criti- cal Examination. William & Mary Law Review, 49: 1347–1376. Restrepo, J, Spagat, M and Vargas, J F 2006 The Severity of the Colombian Conflict: Cross-Country Datasets versus New Micro-Data. Journal of Peace Research, 43(1): 99–115. DOI: http://dx.doi.org/ 10.1177/0022343306059924 Rodríguez, C and Portes, A 2012 Las institu- ciones en Colombia; un análisis soci- ológico. Bogotá: Ediciones Uniandes. Romero, M 2003 Paramilitares y autodefen- sas 1982–2003. Bogotá: IEPRI. Romero, M (ed.). 2007 La ruta de la expansión paramilitar y los acuerdos políticos. Bogotá: Cerec – Corporación Nuevo Arco Iris. Rubin, B R 2008 The Politics of Security in Postconflict Statebuilding. In: Call, C T & Wyeth, V (Eds.), Building States to Build Peace. Boulder, Co: Lynne Rienner Pub- lishers, pp. 25–48. Sánchez, F and Chacón, M 2006 Conflicto, Estado y descentralización: del progreso social a la disputa armada por el control local, 1974–2002. In Gutiérrez, F & Sánchez, G (Eds.), Nuestra guerra sin nombre. Trans- formaciones del conflicto en Colombia. Bogotá: Norma – IEPRI, pp. 347–404. How to cite this article: García-Villegas, M and Espinosa, J R 2015 The Geography of Justice: Assessing Local Justice in Colombia’s Post-Conflict Phase. Stability: International Journal of Security & Development, 4(1): 41, pp. 1–21, DOI: http://dx.doi.org/10.5334/sta.gc Published: 24 July 2015 Copyright: © 2015 The Author(s). This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 Unported License (CC-BY 3.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. See http://creativecommons.org/licenses/by/3.0/. OPEN ACCESS Stability: International Journal of Security & Development is a peer-reviewed open access journal published by Ubiquity Press. http://dx.doi.org/10.2307/3183443 http://dx.doi.org/10.2307/3183443 http://dx.doi.org/10.1093/ser/mwq018 http://dx.doi.org/10.1093/ser/mwq018 http://dx.doi.org/10.1177/0022343306059924 http://dx.doi.org/10.1177/0022343306059924 http://dx.doi.org/10.5334/sta.gc http://creativecommons.org/licenses/by/3.0/ _GoBack