《社会保障概论》课程教学资源（课后阅读资料）APSR_Volume112. Issue 4.Nov 2018_Endogenous_Taxation_in_Ongoing_Internal_Conflict_The_Case_of_Colombia

American Political Science Review (2018)112.4.996-1015 doi:10.1017/S0003055418000333 American Political Science Association 2018 Endogenous Taxation in Ongoing Internal Conflict:The Case of Colombia RAFAEL CH New York University JACOB SHAPIRO Princeton University ABBEY STEELE University of Amsterdam JUAN F.VARGAS Universidad del Rosario ecent empirical evidence suggests an ambiguous relationship between internal conflicts,state ca- pacity,and tax performance.In theory,internal conflict should create strong incentives for gov- ernments to develop the fiscal capacity necessary to defeat rivals.We argue that one reason that this does not occur is because internal conflict enables groups with de facto power to capture local fiscal and property rights institutions.We test this mechanism in Colombia using data on tax performance and property rights institutions at the municipal level.Municipalities affected by internal conflict have tax institutions consistent with the preferences of the parties dominating local violence.Those suffering more right-wing violence feature more land formalization and higher property tax revenues.Municipalities with substantial lef-wing guerrilla violence collect less tax revenue and witness less land formalization.Our findings provide systematic evidence that internal armed conflict helps interest groups capture municipal institutions for their own private benefit,impeding state-building. 4号 INTRODUCTION gions outside of Europe (Centeno 2003;Vu 2010)and for more recent time periods(Scheve and Stasavage istorically,conflicts have motivated states to de- 2012).1 velop their fiscal capacity.According to Tilly Yet the effect of internal wars on state-building and 1992).the formation of nation-states in early modern Europe emerged in part by the fiscal demands fiscal capacity is more ambiguous.While an extension of Tilly's logic to internal wars implies that the state created by expansionary external wars.Those demands should try harder to state-build when challenged,that incentivized leaders to develop institutions that could monitor local populations and levy taxes.The result- clearly does not always happen.This raises an im- portant question:What prevents would-be contempo- ing tax revenue was used to finance armies that would rary state-builders from acting on these incentives?Of protect their kingdoms from external threats or ex- course,when two parties are seeking to build compet- 是 pand their territories.Subsequent work provides em- ing tax institutions within the same territory (e.g.,state pirical support for this theory at the cross-country and rebels),then the net effect is ambiguous.Centeno level(Besley and Persson 2008)and refines it for re- (2003)also argues that the prevalence of civil wars rather than mass wars in Latin America accounts for 2135.501g90 Rafael Ch is a Graduate Student,New York University,Wilf Family the weaker states in the region because they fostered Department of Politics,19 West 4th St.,Room 230,New York,NY elite divisions,physical destruction,and military rather 10009 (rafael.ch@nyu.edu). than mass mobilization. Jacob Shapiro is a Professor of Politics and International Affairs, Princeton University,Woodrow Wilson School of Public and Inter- Empirically,Besley and Persson(2008)and Carde- national Affairs,Princeton,NJ 08544 (jns@princeton.edu). nas,Eslava,and Ramirez(2014)provide evidence that Abbey Steele is an Assistant Professor,University of Amsterdam. internal conflict negatively affects various measures Department of Political Science.Nieuwe Achtergracht 166,1018 WV of state capacity.In Southeast Asia,however,Slater Amsterdam(REC B building)(A.A.Steele@uva.nl). Juan F.Vargas is a Professor of Economics,Universidad del (2010)finds that communal urban unrest and endemic Rosario,School of Economics,Calle 12C No 4-69,Of.315,Bogota violence are both likely to lead elites to tax themselves (juan.vargas@urosario.edu.co). And in Latin America.Soifer (2015)finds evidence We benefited tremendously from excellent comments by Fer- that the prevalence of conflict was crucial for extending nando Alvarez,Mario Chacon,Ursula Daxecker,Leopoldo Fergus- the state during the nineteenth century,and Rodriguez- son,Patrick Kuhn,Livia Schubiger,Pablo Sanguinetti,and seminar Franco (2016)finds that Colombia's elites in Bogota participants at CAF's RED 2015 Workshop in Buenos Aires,the 2015 Empirical Studies of Conflict(ESOC)Annual Meeting,the Up. began to support state-building through new taxes psala University Research Seminar,the conflict seminar at Konstanz in the early 2000s after several decades of internal University,the Amsterdam Conflict Club,and the 2017 Subnational conflict. State in Latin America Conference at the University of Amster- The literature has so far focused on national-level fis- dam.We acknowledge generous funding from CAF-Development Bank of Latin America.ONR through Grant No.N000141310097 cal capacity,but contemporary states feature multiple and AFOSR through Grant No.FA9550-09-1-0314.Dario Sal- administrative layers,and a critical ingredient for the cedo provided superb research assistance.Replication material is consolidation of the state is the introduction of sound available on the American Political Science Review Dataverse: https://doi.org/10.7910/DVN/PCM202. Despite broad support for the theory,some have challenged its Received:August 4,2017;revised:February 1,2018;accepted:May validity for early modern Europe (e.g.,Spruyt 1996;Ertman 1997; 30,2018.First published online:August 3,2018. Gorski 1993). 996

American Political Science Review (2018) 112, 4, 996–1015 doi:10.1017/S0003055418000333 © American Political Science Association 2018 Endogenous Taxation in Ongoing Internal Conflict: The Case of Colombia RAFAEL CH New York University JACOB SHAPIRO Princeton University ABBEY STEELE University of Amsterdam JUAN F. VARGAS Universidad del Rosario Recent empirical evidence suggests an ambiguous relationship between internal conflicts, state ca￾pacity, and tax performance. In theory, internal conflict should create strong incentives for gov￾ernments to develop the fiscal capacity necessary to defeat rivals. We argue that one reason that this does not occur is because internal conflict enables groups with de facto power to capture local fiscal and property rights institutions. We test this mechanism in Colombia using data on tax performance and property rights institutions at the municipal level. Municipalities affected by internal conflict have tax institutions consistent with the preferences of the parties dominating local violence. Those suffering more right-wing violence feature more land formalization and higher property tax revenues.Municipalities with substantial left-wing guerrilla violence collect less tax revenue and witness less land formalization. Our findings provide systematic evidence that internal armed conflict helps interest groups capture municipal institutions for their own private benefit, impeding state-building. INTRODUCTION Historically, conflicts have motivated states to de￾velop their fiscal capacity. According to Tilly (1992), the formation of nation-states in early modern Europe emerged in part by the fiscal demands created by expansionary external wars. Those demands incentivized leaders to develop institutions that could monitor local populations and levy taxes. The result￾ing tax revenue was used to finance armies that would protect their kingdoms from external threats or ex￾pand their territories. Subsequent work provides em￾pirical support for this theory at the cross-country level (Besley and Persson 2008) and refines it for re￾Rafael Ch is a Graduate Student, New York University, Wilf Family Department of Politics, 19 West 4th St., Room 230, New York, NY 10009 (rafael.ch@nyu.edu). Jacob Shapiro is a Professor of Politics and International Affairs, Princeton University, Woodrow Wilson School of Public and Inter￾national Affairs, Princeton, NJ 08544 (jns@princeton.edu). Abbey Steele is an Assistant Professor, University of Amsterdam, Department of Political Science. Nieuwe Achtergracht 166, 1018 WV Amsterdam (REC B building) (A.A.Steele@uva.nl). Juan F. Vargas is a Professor of Economics, Universidad del Rosario, School of Economics, Calle 12C No 4 - 69, Of. 315, Bogotá (juan.vargas@urosario.edu.co). We benefited tremendously from excellent comments by Fer￾nando Álvarez, Mario Chacón, Ursula Daxecker, Leopoldo Fergus￾son, Patrick Kuhn, Livia Schubiger, Pablo Sanguinetti, and seminar participants at CAF’s RED 2015 Workshop in Buenos Aires, the 2015 Empirical Studies of Conflict (ESOC) Annual Meeting, the Up￾psala University Research Seminar, the conflict seminar at Konstanz University, the Amsterdam Conflict Club, and the 2017 Subnational State in Latin America Conference at the University of Amster￾dam. We acknowledge generous funding from CAF-Development Bank of Latin America, ONR through Grant No. N000141310097, and AFOSR through Grant No. FA9550-09-1-0314. Dario Sal￾cedo provided superb research assistance. Replication material is available on the American Political Science Review Dataverse: https://doi.org/10.7910/DVN/PCM2O2. Received: August 4, 2017; revised: February 1, 2018; accepted: May 30, 2018. First published online: August 3, 2018. gions outside of Europe (Centeno 2003; Vu 2010) and for more recent time periods (Scheve and Stasavage 2012).1 Yet the effect of internal wars on state-building and fiscal capacity is more ambiguous. While an extension of Tilly’s logic to internal wars implies that the state should try harder to state-build when challenged, that clearly does not always happen. This raises an im￾portant question: What prevents would-be contempo￾rary state-builders from acting on these incentives? Of course, when two parties are seeking to build compet￾ing tax institutions within the same territory (e.g., state and rebels), then the net effect is ambiguous. Centeno (2003) also argues that the prevalence of civil wars rather than mass wars in Latin America accounts for the weaker states in the region because they fostered elite divisions, physical destruction, and military rather than mass mobilization. Empirically, Besley and Persson (2008) and Cárde￾nas, Eslava, and Ramírez (2014) provide evidence that internal conflict negatively affects various measures of state capacity. In Southeast Asia, however, Slater (2010) finds that communal urban unrest and endemic violence are both likely to lead elites to tax themselves. And in Latin America, Soifer (2015) finds evidence that the prevalence of conflict was crucial for extending the state during the nineteenth century, and Rodríguez￾Franco (2016) finds that Colombia’s elites in Bogotá began to support state-building through new taxes in the early 2000s after several decades of internal conflict. The literature has so far focused on national-level fis￾cal capacity, but contemporary states feature multiple administrative layers, and a critical ingredient for the consolidation of the state is the introduction of sound 1 Despite broad support for the theory, some have challenged its validity for early modern Europe (e.g., Spruyt 1996; Ertman 1997; Gorski 1993). 996 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000333

Endogenous Taxation in Ongoing Internal Conflict and efficient tax systems at the local level.Moreover. the share of tax revenues generated locally (de Mello security provision,property rights protections,and a Jr.2000).3 It also ranks in the middle of large devel- variety of development projects at the local level could, oping countries in terms of local revenue mobilization, in theory,facilitate state-building by shaping citizens above Indonesia at 38%but below India at approxi- preferences and behaviors in ways that favor the state mately 62%(Bird 2012).Third,local government au- over its competitors. thorities have substantial freedom to shape property Several potential mechanisms have been proposed and tax institutions.They can choose tax rates,order to explain the relationship between internal conflict or impede updates on the land value,select revenue and state capacity that are relevant for thinking about collection methods,and alter penalties and incentives. the relationship between conflict and tax performance In theory,this autonomy could facilitate efficient local at the local level.To mention a few:(1)By destroying tax administration by allowing the system to be tailored physical capital,inducing forced migration,and reduc- to the needs of each municipality.In practice,in the ing the market value of private property in affected ar- context of uneven state presence and varying degrees eas,conflict deteriorates the tax base.(2)Conflict gen of contestation by illegal armed groups,local institu- erates negative reciprocity of tax payers toward a state tions can be captured by private groups with vested that they judge to have failed to protect them(Car- interests.4 denas,Eslava,and Ramirez 2014).(3)A conflict envi- Fourth,and most importantly,the main combatants ronment reduces the return of productive activities and had clear preferences regarding state-backed property raise that of illegal businesses,which do not pay taxes to rights throughout much of the war.which had impli- the state(Besley and Persson 2008).(4)Armed groups cations for the local tax base and performance.Right- create their own governance systems that divert civil- wing paramilitaries favored land owners and promoted ians'resources away from the state and crowd out state the accumulation of large estates,which served as institutions (Mampilly 2011;Arjona 2016;Sanchez de means to launder illegally acquired capital (through la Sierra 2017).(5)Conflict facilitates the emergence the drug trade,for example)or realize economies of 4r元 of interest groups with de facto power(for instance be- scale in agricultural production and cattle ranching cause of their access to weapons)that can capture local (Reyes Posada 2007 [1991]).5 On the left-wing side, political and economic institutions(Eaton 2006;Lopez both FARC and ELN guerrilla groups aimed to re- 2010;Mampilly 2011).Such capture prevents state in- place what they characterized as an unjust state and stitutions from eliminating competitors,and from cre- claimed to be acting on behalf of peasants and work- ating a durable and credible relationship with citizens ers.The guerrillas backed land invasions of state and in these regions. private property in many areas (e.g.,Steele 2017).The We focus on the latter political economy mechanism. FARC also viewed state-recognized private property In the context of internal conflicts,local tax institu- as illegitimate or unnecessary:in its VIII Conference tions (both formal and informal)are shaped by differ- in 1993,it still backed collectivized property (Bernal ent types of vested interests.We explore the extent to Morales 2014).(During peace talks with the Santos which armed groups shape existing property and tax administration in 2012,the FARC supported land for- institutions differently depending on their preferences. malization for the first time.)In areas that they influ- To the best of our knowledge,this mechanism has not enced,the guerrillas regulated property in parallel to been studied in the previous literature,either theoret- the state.For instance,they claimed the right to re- ically or empirically.2 Our argument is that local tax distribute unused land or land from narcotraffickers and property rights institutions are shaped by illegal 8 (Bernal Morales 2014).5 armed actors who influence state institutions to further These positions on property rights stem from their interests and those of the civilian groups they fa- ideological commitments and make sense from the vor.Specifically,we expect that right-wing paramilitary perspective of combatants trying to maximize their groups will favor establishing formal property rights control over territory,and to mobilize supporters.His- for land owners,while left-wing insurgents will do the torically,land titling and formalization has favored opposite. We test this argument in Colombia,which has four traits that make it an ideal setting for studying the re- The share of tax revenue collected locally is the standard measure lationship between civil conflict,institutional capture of revenue decentralization in the literature,but is only available for limited subset of countries. and local tax performance.First,the dynamics of the 4 This insight follows naturally from the theory in Persson and eys internal conflict vary substantially across the country's Tabellini(2002),among others,in which heterogeneous agents take 1,122 municipalities.Moreover,within each municipal- advantage of the discretion granted them. ity,violence also varies across different periods of the Paramilitaries displaced people to take their land and titling in ar- war.Second,Colombia sits roughly in the middle of eas with paramilitary presence was not always aboveboard.For ex- the distribution of non-OECD countries in terms of ample.public notaries,the officials charged with validating the prove- nance of land,were infiltrated by paramilitaries (e.g.,Duran 2012: Verdad Abierta 2012). 6 Internal divisions with the FARC existed,but it was a centralized 2 See Eaton(2006)and Lopez(2010)on other forms of capture by organization that adopted policies in its conferences that were rel- illegal armed groups in Colombia.Mampilly (2011)analyzes rebel atively well enforced throughout.The paramilitaries were more de- 士 governance more broadly and Arjona(2014)finds that armed groups centralized but none challenged the basic legitimacy of state institu operated alongside state institutions in roughly one-quarter of the tions (even if they argued that they could be run more effectively) Colombian communities in her sample. (Gutierrez Sanin 2008;Romero 2003). 997

Endogenous Taxation in Ongoing Internal Conflict and efficient tax systems at the local level. Moreover, security provision, property rights protections, and a variety of development projects at the local level could, in theory, facilitate state-building by shaping citizens’ preferences and behaviors in ways that favor the state over its competitors. Several potential mechanisms have been proposed to explain the relationship between internal conflict and state capacity that are relevant for thinking about the relationship between conflict and tax performance at the local level. To mention a few: (1) By destroying physical capital, inducing forced migration, and reduc￾ing the market value of private property in affected ar￾eas, conflict deteriorates the tax base. (2) Conflict gen￾erates negative reciprocity of tax payers toward a state that they judge to have failed to protect them (Cár￾denas, Eslava, and Ramírez 2014). (3) A conflict envi￾ronment reduces the return of productive activities and raise that of illegal businesses,which do not pay taxes to the state (Besley and Persson 2008). (4) Armed groups create their own governance systems that divert civil￾ians’ resources away from the state and crowd out state institutions (Mampilly 2011; Arjona 2016; Sánchez de la Sierra 2017). (5) Conflict facilitates the emergence of interest groups with de facto power (for instance be￾cause of their access to weapons) that can capture local political and economic institutions (Eaton 2006; López 2010; Mampilly 2011). Such capture prevents state in￾stitutions from eliminating competitors, and from cre￾ating a durable and credible relationship with citizens in these regions. We focus on the latter political economy mechanism. In the context of internal conflicts, local tax institu￾tions (both formal and informal) are shaped by differ￾ent types of vested interests. We explore the extent to which armed groups shape existing property and tax institutions differently depending on their preferences. To the best of our knowledge, this mechanism has not been studied in the previous literature, either theoret￾ically or empirically.2 Our argument is that local tax and property rights institutions are shaped by illegal armed actors who influence state institutions to further their interests and those of the civilian groups they fa￾vor. Specifically, we expect that right-wing paramilitary groups will favor establishing formal property rights for land owners, while left-wing insurgents will do the opposite. We test this argument in Colombia, which has four traits that make it an ideal setting for studying the re￾lationship between civil conflict, institutional capture, and local tax performance. First, the dynamics of the internal conflict vary substantially across the country’s 1,122 municipalities. Moreover, within each municipal￾ity, violence also varies across different periods of the war. Second, Colombia sits roughly in the middle of the distribution of non-OECD countries in terms of 2 See Eaton (2006) and López (2010) on other forms of capture by illegal armed groups in Colombia. Mampilly (2011) analyzes rebel governance more broadly and Arjona (2014) finds that armed groups operated alongside state institutions in roughly one-quarter of the Colombian communities in her sample. the share of tax revenues generated locally (de Mello Jr. 2000).3 It also ranks in the middle of large devel￾oping countries in terms of local revenue mobilization, above Indonesia at 38% but below India at approxi￾mately 62% (Bird 2012). Third, local government au￾thorities have substantial freedom to shape property and tax institutions. They can choose tax rates, order or impede updates on the land value, select revenue collection methods, and alter penalties and incentives. In theory, this autonomy could facilitate efficient local tax administration by allowing the system to be tailored to the needs of each municipality. In practice, in the context of uneven state presence and varying degrees of contestation by illegal armed groups, local institu￾tions can be captured by private groups with vested interests.4 Fourth, and most importantly, the main combatants had clear preferences regarding state-backed property rights throughout much of the war, which had impli￾cations for the local tax base and performance. Right￾wing paramilitaries favored land owners and promoted the accumulation of large estates, which served as means to launder illegally acquired capital (through the drug trade, for example) or realize economies of scale in agricultural production and cattle ranching (Reyes Posada 2007 [1991]).5 On the left-wing side, both FARC and ELN guerrilla groups aimed to re￾place what they characterized as an unjust state and claimed to be acting on behalf of peasants and work￾ers. The guerrillas backed land invasions of state and private property in many areas (e.g., Steele 2017). The FARC also viewed state-recognized private property as illegitimate or unnecessary: in its VIII Conference in 1993, it still backed collectivized property (Bernal Morales 2014). (During peace talks with the Santos administration in 2012, the FARC supported land for￾malization for the first time.) In areas that they influ￾enced, the guerrillas regulated property in parallel to the state. For instance, they claimed the right to re￾distribute unused land or land from narcotraffickers (Bernal Morales 2014).6 These positions on property rights stem from ideological commitments and make sense from the perspective of combatants trying to maximize their control over territory, and to mobilize supporters. His￾torically, land titling and formalization has favored 3 The share of tax revenue collected locally is the standard measure of revenue decentralization in the literature, but is only available for a limited subset of countries. 4 This insight follows naturally from the theory in Persson and Tabellini (2002), among others, in which heterogeneous agents take advantage of the discretion granted them. 5 Paramilitaries displaced people to take their land and titling in ar￾eas with paramilitary presence was not always aboveboard. For ex￾ample, public notaries, the officials charged with validating the prove￾nance of land, were infiltrated by paramilitaries (e.g., Durán 2012; Verdad Abierta 2012). 6 Internal divisions with the FARC existed, but it was a centralized organization that adopted policies in its conferences that were rel￾atively well enforced throughout. The paramilitaries were more de￾centralized but none challenged the basic legitimacy of state institu￾tions (even if they argued that they could be run more effectively) (Gutiérrez Sanín 2008; Romero 2003). 997 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000333

Rafael Ch et al. the wealthy and powerful in Colombia(Albertus and due to violence hurting the economy and by exten- Kaplan 2013;Flores 2014),and the FARC issued its sion,property value.The last of these is unlikely be- own tax rules.By obstructing the formal property rights cause it implies symmetric effects of violence by the regime,the FARC could meet its ideological commit- two parties on revenue,which we do not see.Turn- ments,impede the state in territory where it had influ- ing to (1)and(2),though we cannot directly measure ence,and avoid double taxing residents.Paramilitaries, intimidation and pressure,we can measure electoral on the other hand,were supported by groups such as outcomes.Municipalities with more paramilitary vio- large land owners who benefit from formal titles,in- lence do have a greater probability of electing can- cluding to secure land acquired through violence and didates from former President Uribe's right-wing po- coercion.The fact that formal titles create an obliga- litical party coalition,while the probability decreases tion to pay local property taxes would be a relatively in municipalities with high guerrilla violence.Causal small price to pay to secure new property. mediation analysis,however,shows that little of the In Colombia,we can make these expectations con- relationship between violence and tax revenue works crete and test them using disaggregated information on through electoral outcomes:the causal mediation ef- local institutions.We constructed a novel municipal- fect is small,statistically insignificant,and close to zero. level dataset that includes information on various di- This leads us to believe that more indirect mechanisms mensions of property and tax institutions,including the of capture,such as threats and violence against may- land value recorded in the cadaster(i.e.,property reg- ors and city council representatives,account for the dif- istry),the number of cadastral updates,and the dura- ferences in tax revenue and land formalization across tion between updates.s Municipal taxes are primarily municipalities. levied on property value,which is recorded and up- Overall then,guerrilla and paramilitaries'asymmet- dated in the cadaster.Cadaster updates are supposed ric influence on tax performance and land formality are to occur every five years,but municipal administrations consistent with armed group capture at the local level. are responsible for initiating and paying for the up- The higher the level of violence by an armed group,the 4号元 date,and a high proportion do not meet this legal obli- more tax institutions'outcomes shift in the direction of gation (Departamento Nacional de Planeacion 2016). that group's preferences. These data allow us to observe property and tax in This paper contributes to our understanding of con- stitutions at a high resolution.We combine these data temporary state-building during internal wars in four with administrative information on municipal prop- ways.First,our findings show that armed groups have erty tax revenues as well as with variables that ac- the ability to capture the state's local institutions to count for the longitudinal dynamics of conflict activity shape policy outcomes in their favor,which can block in Colombia the state from developing effective institutions.Though We find that armed groups'violent activity correlates Centeno (2003)argues that a minimal administrative with differences in property formalization and taxation state is a necessary condition to generate state-building that are consistent with the groups'political positions in Tilly's framework.we show that capture is an over- Municipalities with significant insurgent violence re looked concern even when this condition is met.Sec- port less land formalization and lower tax receipts.Mu- ond,we offer evidence that armed groups'preferences nicipalities that experienced significant paramilitary vi- and civilian "constituencies"are relevant for how they S5.501g olence have more land formalization and higher tax behave (Wood 2003:Gutierrez Sanin 2003).Third.the receipts.Revenue changes are mirrored by changes in variation in how armed groups reshape local tax insti- socio-economic outcomes,including development lev- tutions in their favor implies a need for a disaggregated els(measured with nighttime light intensity)as well as approach to post-conflict reconstruction.In Colombia secondary school enrollment,which are higher in ar- for example,the state should focus on land redistribu- eas with more paramilitary violence and lower in areas tion and progressive taxation measures in areas where with more guerrilla violence(see Appendix G in the the paramilitaries were dominant.In areas where in- Online Appendix). surgents were dominant,attention to land formaliza There are three obvious mechanisms that could drive tion and tax collection should be prioritized.Finally, the relationship between violence and revenue:(1)in- there is a broader policy lesson in these results.While direct capture through intimidation and pressure on fiscal decentralization might maximize political econ- political actors to update the cadaster or not,(2)di- omy goals in stable countries,it may also engender sig- rect capture of institutions through elections of favored nificant drawbacks in those experiencing ongoing vio- candidates who then carry out the policy preferred by lence(Steele and Schubiger,forthcoming;Eaton 2006) the armed group,and (3)reductions in tax revenues To restore the state's control over local tax institutions and property rights,the central state may have to limit 7 Early paramilitary supporters included drug traffickers who pur- municipal autonomy. chased vast tracts of land(Romero 2000:Ronderos 2014) The rest of the paper is organized as follows:The 8 The cadaster and the property title registry(Registro de Instrumen- next section provides some context and discusses the L tos Publicos)form the basis for property rights.The cadaster records theoretical framework of our argument.This is fol- the physical characteristics of land plots and properties,such as the lowed by the introduction of the data sources and our size and value.The Registro records the title holder for the property. As part of the peace agreement with the FARC,a new integrated empirical strategy.The subsequent section presents the cadaster and Registro is planned to eliminate red tape and modernize main results and robustness checks,followed by the the cadaster.See Departamento Nacional de Planeacion(2016). conclusion to the paper. 998

Rafael Ch et al. the wealthy and powerful in Colombia (Albertus and Kaplan 2013; Flores 2014), and the FARC issued its own tax rules.By obstructing the formal property rights regime, the FARC could meet its ideological commit￾ments, impede the state in territory where it had influ￾ence, and avoid double taxing residents. Paramilitaries, on the other hand, were supported by groups such as large land owners who benefit from formal titles, in￾cluding to secure land acquired through violence and coercion.7 The fact that formal titles create an obliga￾tion to pay local property taxes would be a relatively small price to pay to secure new property. In Colombia, we can make these expectations con￾crete and test them using disaggregated information on local institutions. We constructed a novel municipal￾level dataset that includes information on various di￾mensions of property and tax institutions, including the land value recorded in the cadaster (i.e., property reg￾istry), the number of cadastral updates, and the dura￾tion between updates.8 Municipal taxes are primarily levied on property value, which is recorded and up￾dated in the cadaster. Cadaster updates are supposed to occur every five years, but municipal administrations are responsible for initiating and paying for the up￾date, and a high proportion do not meet this legal obli￾gation (Departamento Nacional de Planeación 2016). These data allow us to observe property and tax in￾stitutions at a high resolution. We combine these data with administrative information on municipal prop￾erty tax revenues as well as with variables that ac￾count for the longitudinal dynamics of conflict activity in Colombia. We find that armed groups’ violent activity correlates with differences in property formalization and taxation that are consistent with the groups’ political positions. Municipalities with significant insurgent violence re￾port less land formalization and lower tax receipts.Mu￾nicipalities that experienced significant paramilitary vi￾olence have more land formalization and higher tax receipts. Revenue changes are mirrored by changes in socio-economic outcomes, including development lev￾els (measured with nighttime light intensity) as well as secondary school enrollment, which are higher in ar￾eas with more paramilitary violence and lower in areas with more guerrilla violence (see Appendix G in the Online Appendix). There are three obvious mechanisms that could drive the relationship between violence and revenue: (1) in￾direct capture through intimidation and pressure on political actors to update the cadaster or not, (2) di￾rect capture of institutions through elections of favored candidates who then carry out the policy preferred by the armed group, and (3) reductions in tax revenues 7 Early paramilitary supporters included drug traffickers who pur￾chased vast tracts of land (Romero 2000; Ronderos 2014). 8 The cadaster and the property title registry (Registro de Instrumen￾tos Públicos) form the basis for property rights. The cadaster records the physical characteristics of land plots and properties, such as the size and value. The Registro records the title holder for the property. As part of the peace agreement with the FARC, a new integrated cadaster and Registro is planned to eliminate red tape and modernize the cadaster. See Departamento Nacional de Planeación (2016). due to violence hurting the economy and by exten￾sion, property value. The last of these is unlikely be￾cause it implies symmetric effects of violence by the two parties on revenue, which we do not see. Turn￾ing to (1) and (2), though we cannot directly measure intimidation and pressure, we can measure electoral outcomes. Municipalities with more paramilitary vio￾lence do have a greater probability of electing can￾didates from former President Uribe’s right-wing po￾litical party coalition, while the probability decreases in municipalities with high guerrilla violence. Causal mediation analysis, however, shows that little of the relationship between violence and tax revenue works through electoral outcomes: the causal mediation ef￾fect is small, statistically insignificant, and close to zero. This leads us to believe that more indirect mechanisms of capture, such as threats and violence against may￾ors and city council representatives, account for the dif￾ferences in tax revenue and land formalization across municipalities. Overall then, guerrilla and paramilitaries’ asymmet￾ric influence on tax performance and land formality are consistent with armed group capture at the local level. The higher the level of violence by an armed group, the more tax institutions’ outcomes shift in the direction of that group’s preferences. This paper contributes to our understanding of con￾temporary state-building during internal wars in four ways. First, our findings show that armed groups have the ability to capture the state’s local institutions to shape policy outcomes in their favor, which can block the state from developing effective institutions.Though Centeno (2003) argues that a minimal administrative state is a necessary condition to generate state-building in Tilly’s framework, we show that capture is an over￾looked concern even when this condition is met. Sec￾ond, we offer evidence that armed groups’ preferences and civilian “constituencies” are relevant for how they behave (Wood 2003; Gutiérrez Sanín 2003). Third, the variation in how armed groups reshape local tax insti￾tutions in their favor implies a need for a disaggregated approach to post-conflict reconstruction. In Colombia, for example, the state should focus on land redistribu￾tion and progressive taxation measures in areas where the paramilitaries were dominant. In areas where in￾surgents were dominant, attention to land formaliza￾tion and tax collection should be prioritized. Finally, there is a broader policy lesson in these results. While fiscal decentralization might maximize political econ￾omy goals in stable countries, it may also engender sig￾nificant drawbacks in those experiencing ongoing vio￾lence (Steele and Schubiger,forthcoming; Eaton 2006). To restore the state’s control over local tax institutions and property rights, the central state may have to limit municipal autonomy. The rest of the paper is organized as follows: The next section provides some context and discusses the theoretical framework of our argument. This is fol￾lowed by the introduction of the data sources and our empirical strategy. The subsequent section presents the main results and robustness checks, followed by the conclusion to the paper. 998 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000333

Endogenous Taxation in Ongoing Internal Conflict FIGURE 1.Tax Revenue over Total Expenditure across Colombian Municipalities period2002-2012 .4 .6 8 Tax Revenue^/Total Expenditure Source:Elaborated using fiscal data from CEDE. ^Tax Revenue tributary income,that corresponds to income by concept of tax on property,industry and commerce,and gasoline 4号元 CONTEXT body)issues the municipality's tax statute,which in- cludes the tax rates,the type of properties for which Tax Institutions and Tax Performance in each rate applies,the collection methods,and the pay- Colombia ment incentives and fines.1 Property tax rates vary sub- Municipalities in Colombia vary widely in their tax stantially between rural and urban areas,and may or performance.Figure 1 plots the ratio of tax revenue may not vary by type of property or its specific use(e.g., private housing,production,or commercial purposes). to total expenditure (averaged for the period 2000- 2012)across the country's 1,122 municipalities.The av- In addition,systems can be mixed within municipalities, erage municipality levies taxes worth 12%of their ex- with some properties and businesses taxed according to penditure,with the balance supplied by transfers from one rule (e.g.,the value of the property recorded in the municipal cadaster),and others according to another the national government based on population size and (e.g.,the socio-economic conditions of the neighbor- poverty levels,as well as royalties from natural re- source extraction.The variation is enormous,with sev- hood where the properties are located).The majority of municipalities have mixed systems that combine var- eral municipalities unable to generate practically any ious schemes revenue and a few capable of financing up to 80%of Regardless of the tax rates and system,the cadaster their expenditures with local taxes. forms the basis of property taxes.2 The cadaster in- The variation is not simply a proxy for economic ac- tivity,as Figure 2 shows.Panel (a)plots the distribution cludes the property values and the physical traits of properties.Outdated cadasters reduce the amount of of different tax revenues per capita at the municipal level,averaged from 2000 to 2012.9 Panel(b)shows the revenue a municipality collects in property taxes,be- same results normalized using nighttime lights to proxy cause recorded property values are lower than the true values(Iregui,Melo,and Ramos 2004). for economic activity (Vernon,Storeygard,and Weil 2011).The first column presents logged tributary in- come while the second shows the distribution of logged property tax income,which comprises the bulk of local 1 Asstated above.the cadaster is supposed to be updated every five revenue.10 years,but mayors are responsible for initiating the process.Then the The large variation in tax receipts reflects the free- national geographic institute,Instituto Geografico Augustin Codazzi dom that local authorities have in designing tax insti- (IGAC),is supposed to carry out the actual update with municipal resources.Although the municipal council sets tax rates,they must tutions.While the municipal mayor(the highest local- fall within a range defined by the National Congress.The current level executive authority)is in charge of updating the range for property tax is 5-16 per thousand for all types of properties. land registry,the city council (the municipal legislative Nufez (2005)reviews Colombia's local tax system. 12 Some residents pay property taxes even in the absence of formal title because it is a way to demonstrate continual presence on a parcel 9 Tax revenues are in constant Colombian pesos from 2008. of land.According to Colombian law,investments in land that make 10 Both are logged given their highly skewed distributions with a it"socially productive"can be rewarded with formal property rights long right tail of municipalities with larger tributary and property for the land in question after five consecutive years of residence (Law tax income. 200of1936). 999

Endogenous Taxation in Ongoing Internal Conflict FIGURE 1. Tax Revenue over Total Expenditure across Colombian Municipalities 0 2 4 6 8 10 Density 0 .2 .4 .6 .8 Tax Revenue^ / Total Expenditure period 2002-2012 Source: Elaborated using fiscal data from CEDE. ^Tax Revenue = tributary income, that corresponds to income by concept of tax on property, industry and commerce, and gasoline. CONTEXT Tax Institutions and Tax Performance in Colombia Municipalities in Colombia vary widely in their tax performance. Figure 1 plots the ratio of tax revenue to total expenditure (averaged for the period 2000– 2012) across the country’s 1,122 municipalities. The av￾erage municipality levies taxes worth 12% of their ex￾penditure, with the balance supplied by transfers from the national government based on population size and poverty levels, as well as royalties from natural re￾source extraction. The variation is enormous, with sev￾eral municipalities unable to generate practically any revenue and a few capable of financing up to 80% of their expenditures with local taxes. The variation is not simply a proxy for economic ac￾tivity, as Figure 2 shows. Panel (a) plots the distribution of different tax revenues per capita at the municipal level, averaged from 2000 to 2012.9 Panel (b) shows the same results normalized using nighttime lights to proxy for economic activity (Vernon, Storeygard, and Weil 2011). The first column presents logged tributary in￾come while the second shows the distribution of logged property tax income, which comprises the bulk of local revenue.10 The large variation in tax receipts reflects the free￾dom that local authorities have in designing tax insti￾tutions. While the municipal mayor (the highest local￾level executive authority) is in charge of updating the land registry, the city council (the municipal legislative 9 Tax revenues are in constant Colombian pesos from 2008. 10 Both are logged given their highly skewed distributions, with a long right tail of municipalities with larger tributary and property tax income. body) issues the municipality’s tax statute, which in￾cludes the tax rates, the type of properties for which each rate applies, the collection methods, and the pay￾ment incentives and fines.11 Property tax rates vary sub￾stantially between rural and urban areas, and may or may not vary by type of property or its specific use (e.g., private housing, production, or commercial purposes). In addition, systems can be mixed within municipalities, with some properties and businesses taxed according to one rule (e.g., the value of the property recorded in the municipal cadaster), and others according to another (e.g., the socio-economic conditions of the neighbor￾hood where the properties are located). The majority of municipalities have mixed systems that combine var￾ious schemes. Regardless of the tax rates and system, the cadaster forms the basis of property taxes.12 The cadaster in￾cludes the property values and the physical traits of properties. Outdated cadasters reduce the amount of revenue a municipality collects in property taxes, be￾cause recorded property values are lower than the true values (Iregui, Melo, and Ramos 2004). 11 As stated above, the cadaster is supposed to be updated every five years, but mayors are responsible for initiating the process. Then the national geographic institute, Instituto Geográfico Augustín Codazzi (IGAC), is supposed to carry out the actual update with municipal resources. Although the municipal council sets tax rates, they must fall within a range defined by the National Congress. The current range for property tax is 5–16 per thousand for all types of properties. Nuñez (2005) reviews Colombia’s local tax system. 12 Some residents pay property taxes even in the absence of formal title because it is a way to demonstrate continual presence on a parcel of land. According to Colombian law, investments in land that make it “socially productive” can be rewarded with formal property rights for the land in question after five consecutive years of residence (Law 200 of 1936). 999 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000333

Rafael Ch et al. FIGURE 2.Tax Revenue across Colombian Municipalities by Type average 2000-2012 Tributary income a) Property tax income c4 10 6 10 log(tributary income per capita) log(property tax income per capita) b) 10 6 5 10 15 log(tributary income) log(property tax income) normalized by nighttime lights normalized by nighttime lights Source:Elaborated using fiscal data from CEDE and nighttime light from NOAA. Tributary income:income by concept of tax on property,industry and commerce.and gasoline. The Capture of Local Tax and Property ited monopoly over the use of force."The influence of Institutions armed groups over locally elected politicians could be substantial. In Latin America.only Brazil and Venezuela have In Colombia,insurgents and right-wing paramilitary 115.5010 more decentralized systems,and in the rest of the coun- groups differed in their approaches to capture the state tries in the region,provincial or national governments (Lopez Hernandez 2010;Eaton 2006;Romero 2007). are in charge of tax legislation.In Colombia,fiscal au- Paramilitary groups frequently colluded with state tonomy was pursued because policymakers thought forces,but were independent of them.They forged that,combined with other decentralization reforms,it could help end the war (Eaton 2006,534,541).The Be- extensive ties to regional-and national-level politi- cians(Gutierrez Sanin 2010,16).Indeed,paramilitaries tancur administration(1982-1986)approved the direct were embraced,and,in some cases,even founded by election of mayors that began in 1988,and the two sub- sequent administrations under Barco and Gaviria de- the country's regional political elites (Romero 2003; Ronderos 2014:Duncan 2006). volved fiscal authority and budgetary responsibility to municipalities for several public goods such as educa- In contrast,the FARC focused on local organiza- tion,health,and road upkeep. tion,particularly through the Juntas de Accion Com- munal (JAC),committees based in the rural hamlets Unfortunately,despite those changes(and partly as within municipalities.While the FARC also engaged a result of them),the civil war intensified and spread in the 1990s (Sanchez and del Mar Palau 2006:Steele with some municipal-and national-level officials,they and Schubiger,forthcoming;Steele 2017).Municipal- did so to a far lesser extent then the paramilitaries. As Gutierrez Sanin (2010,18-9)points out,insur- ities became attractive targets for armed groups be- gents were less reliable partners for politicians,be- cause of the transfers and royalties they received and cause they were anti-state.Consistent with their leftist because the state was insufficiently equipped to pro- vide security to them.Eaton (2006,537)writes,"...the state now funds its own destabilization because armed 13 For example,in 2010 only 10 of the 277 alcaldes and municipal groups on the left and right have been able to ap- council members under investigation for ties to illegal armed groups propriate decentralized public revenues and to use were linked to the FARC:similarly.only 4%of the congress was in- vestigated for ties to the FARC,compared to 35%to paramilitary these funds to further reduce the state's already lim- groups (Lopez Hernandez 2010,33). 1000

Rafael Ch et al. FIGURE 2. Tax Revenue across Colombian Municipalities by Type 0 .1 .2 .3 .4 Density 0 5 10 15 log(tributary income per capita) Tributary income 0 .1 .2 .3 .4 0 5 10 15 log(property tax income per capita) Property tax income 0 .1 .2 .3 .4 Density 0 5 10 15 log(tributary income) normalized by nighttime lights 0 .1 .2 .3 .4 0 5 10 15 log(property tax income) normalized by nighttime lights average 2000-2012 a) b) Source: Elaborated using fiscal data from CEDE and nighttime light from NOAA. Tributary income: income by concept of tax on property, industry and commerce, and gasoline. The Capture of Local Tax and Property Institutions In Latin America, only Brazil and Venezuela have more decentralized systems, and in the rest of the coun￾tries in the region, provincial or national governments are in charge of tax legislation. In Colombia, fiscal au￾tonomy was pursued because policymakers thought that, combined with other decentralization reforms, it could help end the war (Eaton 2006, 534, 541). The Be￾tancur administration (1982–1986) approved the direct election of mayors that began in 1988, and the two sub￾sequent administrations under Barco and Gaviria de￾volved fiscal authority and budgetary responsibility to municipalities for several public goods such as educa￾tion, health, and road upkeep. Unfortunately, despite those changes (and partly as a result of them), the civil war intensified and spread in the 1990s (Sánchez and del Mar Palau 2006; Steele and Schubiger, forthcoming; Steele 2017). Municipal￾ities became attractive targets for armed groups be￾cause of the transfers and royalties they received and because the state was insufficiently equipped to pro￾vide security to them. Eaton (2006, 537) writes, “…the state now funds its own destabilization because armed groups on the left and right have been able to ap￾propriate decentralized public revenues and to use these funds to further reduce the state’s already lim￾ited monopoly over the use of force.” The influence of armed groups over locally elected politicians could be substantial. In Colombia, insurgents and right-wing paramilitary groups differed in their approaches to capture the state (López Hernández 2010; Eaton 2006; Romero 2007). Paramilitary groups frequently colluded with state forces, but were independent of them. They forged extensive ties to regional- and national-level politi￾cians (Gutiérrez Sanín 2010, 16). Indeed, paramilitaries were embraced, and, in some cases, even founded by the country’s regional political elites (Romero 2003; Ronderos 2014; Duncan 2006). In contrast, the FARC focused on local organiza￾tion, particularly through the Juntas de Acción Com￾munal (JAC), committees based in the rural hamlets within municipalities. While the FARC also engaged with some municipal- and national-level officials, they did so to a far lesser extent then the paramilitaries.13 As Gutiérrez Sanín (2010, 18–9) points out, insur￾gents were less reliable partners for politicians, be￾cause they were anti-state. Consistent with their leftist 13 For example, in 2010 only 10 of the 277 alcaldes and municipal council members under investigation for ties to illegal armed groups were linked to the FARC; similarly, only 4% of the congress was in￾vestigated for ties to the FARC, compared to 35% to paramilitary groups (López Hernández 2010, 33). 1000 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000333

Endogenous Taxation in Ongoing Internal Conflict ideology,insurgents tried to mobilize and work with Though we expect tax revenue to be higher in areas peasants and workers who were traditionally shut out with paramilitary presence and lower in areas with of the political system.The distinct "constituencies" insurgent presence,we do not argue that right-wing of insurgents and paramilitaries imply differing prefer- paramilitaries or the elites that supported them fa- ences over policies. vored increased taxation or that the left-wing insur- A particularly fraught policy area that armed groups gents opposed redistribution.Instead,armed groups' intervened in is land formalization,which has implica- preferences over tax rates and enforcement are am- tions for local tax performance.Right-wing paramili- biguous in Colombia.Paramilitaries might try to ben- taries were founded by large landowners and narco- efit supporters by pushing for lower rates and lax en- traffickers who began to purchase tracts of land. forcement of property taxes.This position would be As they expanded,they displaced peasants and claimed consistent with historical precedent in the twentieth their land (Romero 2000;Ronderos 2014;Reyes century:Sanchez-Talanquer(2018)finds that following Posada 2009).To legitimate this transfer of land, large-scale land reform attempts,local elites registered paramilitaries sought to formalize property through property in the cadaster to establish their property the state.This was consistent with what we noted rights,but manipulated property value to avoid paying above:traditionally,elites have been able to secure le- higher tax.At the same time,paramilitaries could ben- gal title to their property,while poor peasants face mul- efit from an increase in local tax revenue because they tiple barriers to doing so.Updating cadasters to reflect issued lucrative contracts to supporters to carry out newly acquired or changed properties(for instance,an municipal functions like health services(Eaton 2006; enlarged plot of land)is consistent with an interest in Verdad Abierta n.d.).16 New landowners could also securing property rights,even if the property is illegally calculate that paying taxes was a small cost to pay acquired (for instance through claiming the land of a for securing property,and employ tax resources to displaced household).14 expand their control over the local bureaucracy and Left-wing insurgents favored more equitable land bargaining power relative to higher levels of govern- 4号元 distribution but not private property rights (Bernal ment (Sanchez-Talanquer 2018;Pardelli 2018).More- Morales 2014).One way they promoted this goal was over,short-term development of specific local admin- through land invasions,where poor peasants and work- istrative capacities does not prevent the long-term hol- ers would occupy state lands(baldios)or private prop- lowing out and capture of the state (Stasavage 2014). erty.However,such invasions were typically not for- Insurgents,though they did favor redistribution,pre- malized legally.More common in insurgent areas was ferred not to work through the state's institutions be- a preference for excluding or prohibiting represen- cause doing so would legitimate the state.The FARC tatives of the state from surveying the land or plot- issued its own tax code,targeting the wealthy in areas of ropertyncessary steps the ftormalization their influence(see below).In terms of property rights, the insurgents preferred to impede the efforts of the Given the divergent preferences of and incentives state (Bernal Morales 2014)and regulate land on its for armed groups to shape local property institutions own.17 we expect differences in local property tax perfor- Before we explain how we test our hypotheses,we mance overall.We propose four hypotheses: first summarize temporal trends in capture over the 5181.10198 course of the war In areas with substantial FARC and ELN violence: Civil War Dynamics and Capture in Colombia Hia Tax revenues per capita will be lower because fewer properties will be formalized. The dynamics of the civil war in terms of territorial Hib Left-wing parties will tend to outperform in mayoral control and violence have changed significantly over and municipal council elections controlling for the pre- time.We separate the war's recent evolution and the existing partisan balance. armed groups'capture strategies into four periods and organize the statistical analysis around these four pe- In areas with substantial paramilitary violence: riods because pooling them would effectively assume that armed actors had the same ability to influence tax H2a Tax revenues per capita will be higher because more policies in every year,which strikes us as substantively property will be formalized. unrealistic.The four periods(described in detail in Ap- eys H2b Right-wing parties will tend to outperform in mayoral pendix H)can be defined as follows: and municipal council elections controlling for the pre- existing partisan balance. 6 Though ideally we could test the effects of armed groups'influence on tax rates and enforcement,we do not have the data to do so.In 14 Ibaniez and Munoz(2011)find that the concentration of land be- the end,what we do test is the net effect of their influence on local tween 2000 and 2009 increases as the result of an increase in plots property tax revenue.These data indicate that whatever is happening mber of plots purchsed by few people. with tax rates and enforcement,land formalization measures increase Corporacion Regional para el Desarrollo Sostenible del Area de tax revenue overall. Manejo Especial la Macarena(CORMACARENA)official inter- One example of the FARC's intransigence on state-backed prop- view with the authors,Vista Hermosa,January 26,2011.This official erty rights was its opposition to land restitution for the internally told us he could not conduct land surveys in the area because the displaced through the Victims Law of 2011,calling it "legal dispos- FARC would not permit it. session"(Bernal Morales 2014). 1001

Endogenous Taxation in Ongoing Internal Conflict ideology, insurgents tried to mobilize and work with peasants and workers who were traditionally shut out of the political system. The distinct “constituencies” of insurgents and paramilitaries imply differing prefer￾ences over policies. A particularly fraught policy area that armed groups intervened in is land formalization, which has implica￾tions for local tax performance. Right-wing paramili￾taries were founded by large landowners and narco￾traffickers who began to purchase tracts of land. As they expanded, they displaced peasants and claimed their land (Romero 2000; Ronderos 2014; Reyes Posada 2009). To legitimate this transfer of land, paramilitaries sought to formalize property through the state. This was consistent with what we noted above: traditionally, elites have been able to secure le￾gal title to their property, while poor peasants face mul￾tiple barriers to doing so. Updating cadasters to reflect newly acquired or changed properties (for instance, an enlarged plot of land) is consistent with an interest in securing property rights, even if the property is illegally acquired (for instance through claiming the land of a displaced household).14 Left-wing insurgents favored more equitable land distribution but not private property rights (Bernal Morales 2014). One way they promoted this goal was through land invasions, where poor peasants and work￾ers would occupy state lands (baldíos) or private prop￾erty. However, such invasions were typically not for￾malized legally. More common in insurgent areas was a preference for excluding or prohibiting represen￾tatives of the state from surveying the land or plot￾ting the property—necessary steps in the formalization process.15 Given the divergent preferences of and incentives for armed groups to shape local property institutions, we expect differences in local property tax perfor￾mance overall. We propose four hypotheses: In areas with substantial FARC and ELN violence: H1a Tax revenues per capita will be lower because fewer properties will be formalized. H1b Left-wing parties will tend to outperform in mayoral and municipal council elections controlling for the pre￾existing partisan balance. In areas with substantial paramilitary violence: H2a Tax revenues per capita will be higher because more property will be formalized. H2b Right-wing parties will tend to outperform in mayoral and municipal council elections controlling for the pre￾existing partisan balance. 14 Ibáñez and Muñoz (2011) find that the concentration of land be￾tween 2000 and 2009 increases as the result of an increase in plots and an increased number of plots purchased by few people. 15 Corporación Regional para el Desarrollo Sostenible del Area de Manejo Especial la Macarena (CORMACARENA) official inter￾view with the authors, Vista Hermosa, January 26, 2011. This official told us he could not conduct land surveys in the area because the FARC would not permit it. Though we expect tax revenue to be higher in areas with paramilitary presence and lower in areas with insurgent presence, we do not argue that right-wing paramilitaries or the elites that supported them fa￾vored increased taxation or that the left-wing insur￾gents opposed redistribution. Instead, armed groups’ preferences over tax rates and enforcement are am￾biguous in Colombia. Paramilitaries might try to ben￾efit supporters by pushing for lower rates and lax en￾forcement of property taxes. This position would be consistent with historical precedent in the twentieth century: Sánchez-Talanquer (2018) finds that following large-scale land reform attempts, local elites registered property in the cadaster to establish their property rights, but manipulated property value to avoid paying higher tax. At the same time, paramilitaries could ben￾efit from an increase in local tax revenue because they issued lucrative contracts to supporters to carry out municipal functions like health services (Eaton 2006; Verdad Abierta n.d.).16 New landowners could also calculate that paying taxes was a small cost to pay for securing property, and employ tax resources to expand their control over the local bureaucracy and bargaining power relative to higher levels of govern￾ment (Sánchez-Talanquer 2018; Pardelli 2018). More￾over, short-term development of specific local admin￾istrative capacities does not prevent the long-term hol￾lowing out and capture of the state (Stasavage 2014). Insurgents, though they did favor redistribution, pre￾ferred not to work through the state’s institutions be￾cause doing so would legitimate the state. The FARC issued its own tax code, targeting the wealthy in areas of their influence (see below). In terms of property rights, the insurgents preferred to impede the efforts of the state (Bernal Morales 2014) and regulate land on its own.17 Before we explain how we test our hypotheses, we first summarize temporal trends in capture over the course of the war. Civil War Dynamics and Capture in Colombia The dynamics of the civil war in terms of territorial control and violence have changed significantly over time. We separate the war’s recent evolution and the armed groups’ capture strategies into four periods and organize the statistical analysis around these four pe￾riods because pooling them would effectively assume that armed actors had the same ability to influence tax policies in every year, which strikes us as substantively unrealistic. The four periods (described in detail in Ap￾pendix H) can be defined as follows: 16 Though ideally we could test the effects of armed groups’influence on tax rates and enforcement, we do not have the data to do so. In the end, what we do test is the net effect of their influence on local property tax revenue.These data indicate that whatever is happening with tax rates and enforcement,land formalization measures increase tax revenue overall. 17 One example of the FARC’s intransigence on state-backed prop￾erty rights was its opposition to land restitution for the internally displaced through the Victims Law of 2011, calling it “legal dispos￾session” (Bernal Morales 2014). 1001 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000333

Rafael Ch et al. 1988-1996:FARC ascendancy.According to one widely in the correlation between tax performance and estimate,the FARC went from a presence in 173 the type and level of armed group presence;second. of the country's municipalities in 1985 to 622 by and following the historical shifts of the Colombian :000100006/L0LoL 1995(Echandia 2006.28).By the end of the period. war,this variation changes over time,a feature that the FARC and the ELN both enforced election will be exploited in the empirical specification to test boycotts in areas under their control and threat- our hypotheses and in the substantive interpretation of ened elected mayors and local council members results.19 ("FARC Prohiben Elecciones en 23 Municipios,' El Tiempo,October 23,1997).While some re- gional politicians supported paramilitaries'forma- Data tion during this period(Ronderos 2014,37),there Table 1 provides descriptive statistics of our main out- is little evidence that paramilitary groups tried to come variables,used to test both the main empir- capture political institutions directly at the local ical relationships and the mechanisms described in level during this period. our testable hypotheses(Panel A).Outcomes include 1997-2002:Paramilitary expansion.In 1997,re- logged tax revenues per capita,proxies of cadastral per- gional paramilitary groups united under the formance,the land informality rate,and electoral out- umbrella group United Self-Defense Forces of comes.Tax revenues are highly skewed,with a long Colombia (AUC)and the war spread.By 2001, right tail of municipalities with substantially greater tax the AUC was powerful enough to convene a meet- revenues per capita than others.20 We therefore run our ing with nearly 100 politicians to formulate a con- estimation of the impact of violence on property tax certed effort to win elections at all levels,and to revenues per capita on logged values.21 support Alvaro Uribe's candidacy for president in Table 1 also describes the independent variables of 2002(known as the Santa Fe de Ralito pact).Com- interest(Panel B),namely the cumulative attacks per- pared to the paramilitaries,the FARC's influence petrated by each of the two main armed groups,dur- 4号 remained indirect during this period as the group ing each one of the four periods that mark the evo- eschewed official electoral politics,preferring to lution of Colombia's civil war,as described in the threaten municipal candidates that the group did Civil War Dynamics and Capture in Colombia sec- not approve of,or acting mayors. tion.22 Measuring the influence exercised by an armed ● 2003-2006:Paramilitary demobilization.In 2003, group over a specific location is extremely challenging. the Uribe administration negotiated a ceasefire Indicators of presence and nonviolent coercion over with paramilitary groups and eventually adopted a large set of municipalities cannot be systematically the Justice and Peace law,allowing paramilitary recorded in an objective way.Violence,on the other commanders to demobilize their troops in ex- hand,while more easily observed,is only imperfectly change for lenient sentences.Paramilitary demo- correlated with territorial dominance.For instance,it bilizations from 2003 to 2005 transformed the war may be the case that municipalities with low levels of into a contest between the state and remaining in- violence or no armed contestation represent an armed surgent groups(the FARC and ELN).The conflict group stronghold,where tax policies are likely to be with the FARC continued apace during this period, influenced. with no change in their capture strategy. However,nonviolent dominance is unlikely to occur 2007-2010:State resurgence.Having pushed the without any violence inflicted in the past,either as a FARC into peripheral areas by the end of 2006, way to legitimize influence with the citizenry or to oust the Colombian military and police redeployed any contesting(legal or illegal)group.It is thus reason- to major population centers and roads,improv- able to assume that the ability to inflict localized vio- ing measures of security.The weakened FARC lence over a relatively long period could be expected agreed to peace talks following the 2010 elec- to translate into influence in different ways.Moreover, tion of Uribe's Minister of Defense,Juan Manuel as all our results are robust to controlling for violence Santos.Former paramilitary groups morphed into by the other actor,we posit that municipalities with new organizations-including the Black Eagles, greater violence are more likely to be captured by the and drug-trafficking groups such as the Urabenos perpetrating armed group.?3 and the Rastrojos,which sometimes engaged in ac- tions against the FARC,the ELN,and the civil- ian population,though at much lower rates than 19Correlations are carried out at the department level since munic- in previous times. Taking into account these time periods,Figure 3 shows dard deviations above the mean. the department-level relationship between cumulative 21 The results are substantially the same when we add to the set of controls the log of the municipal population instead of using per attacks by armed groups and average property tax rev- enues in the following time period.There are two See Appendix A for detailed descriptions of all the variables and main takeaways:first,departments in Colombia vary measures. 23 A more stringent test is to focus on the subsample of places where violence by both groups is recorded,dropping the municipalities 18 There are 32 departments in Colombia.This administration level where one group does not engage in any violence at all during the is equivalent to US states. sample period.Even with the statistical power reduction implied by 1002

Rafael Ch et al. 1988–1996: FARC ascendancy. According to one estimate, the FARC went from a presence in 173 of the country’s municipalities in 1985 to 622 by 1995 (Echandía 2006, 28). By the end of the period, the FARC and the ELN both enforced election boycotts in areas under their control and threat￾ened elected mayors and local council members (“FARC Prohíben Elecciones en 23 Municipios,” El Tiempo, October 23, 1997). While some re￾gional politicians supported paramilitaries’ forma￾tion during this period (Ronderos 2014, 37), there is little evidence that paramilitary groups tried to capture political institutions directly at the local level during this period. 1997–2002: Paramilitary expansion. In 1997, re￾gional paramilitary groups united under the umbrella group United Self-Defense Forces of Colombia (AUC) and the war spread. By 2001, the AUC was powerful enough to convene a meet￾ing with nearly 100 politicians to formulate a con￾certed effort to win elections at all levels, and to support Álvaro Uribe’s candidacy for president in 2002 (known as the Santa Fe de Ralito pact). Com￾pared to the paramilitaries, the FARC’s influence remained indirect during this period as the group eschewed official electoral politics, preferring to threaten municipal candidates that the group did not approve of, or acting mayors. 2003–2006: Paramilitary demobilization. In 2003, the Uribe administration negotiated a ceasefire with paramilitary groups and eventually adopted the Justice and Peace law, allowing paramilitary commanders to demobilize their troops in ex￾change for lenient sentences. Paramilitary demo￾bilizations from 2003 to 2005 transformed the war into a contest between the state and remaining in￾surgent groups (the FARC and ELN). The conflict with the FARC continued apace during this period, with no change in their capture strategy. 2007–2010: State resurgence. Having pushed the FARC into peripheral areas by the end of 2006, the Colombian military and police redeployed to major population centers and roads, improv￾ing measures of security. The weakened FARC agreed to peace talks following the 2010 elec￾tion of Uribe’s Minister of Defense, Juan Manuel Santos. Former paramilitary groups morphed into new organizations—including the Black Eagles, and drug-trafficking groups such as the Urabeños and the Rastrojos, which sometimes engaged in ac￾tions against the FARC, the ELN, and the civil￾ian population, though at much lower rates than in previous times. Taking into account these time periods, Figure 3 shows the department-level relationship between cumulative attacks by armed groups and average property tax rev￾enues in the following time period.18 There are two main takeaways: first, departments in Colombia vary 18 There are 32 departments in Colombia. This administration level is equivalent to US states. widely in the correlation between tax performance and the type and level of armed group presence; second, and following the historical shifts of the Colombian war, this variation changes over time, a feature that will be exploited in the empirical specification to test our hypotheses and in the substantive interpretation of results.19 Data Table 1 provides descriptive statistics of our main out￾come variables, used to test both the main empir￾ical relationships and the mechanisms described in our testable hypotheses (Panel A). Outcomes include logged tax revenues per capita, proxies of cadastral per￾formance, the land informality rate, and electoral out￾comes. Tax revenues are highly skewed, with a long right tail of municipalities with substantially greater tax revenues per capita than others.20 We therefore run our estimation of the impact of violence on property tax revenues per capita on logged values.21 Table 1 also describes the independent variables of interest (Panel B), namely the cumulative attacks per￾petrated by each of the two main armed groups, dur￾ing each one of the four periods that mark the evo￾lution of Colombia’s civil war, as described in the Civil War Dynamics and Capture in Colombia sec￾tion.22 Measuring the influence exercised by an armed group over a specific location is extremely challenging. Indicators of presence and nonviolent coercion over a large set of municipalities cannot be systematically recorded in an objective way. Violence, on the other hand, while more easily observed, is only imperfectly correlated with territorial dominance. For instance, it may be the case that municipalities with low levels of violence or no armed contestation represent an armed group stronghold, where tax policies are likely to be influenced. However, nonviolent dominance is unlikely to occur without any violence inflicted in the past, either as a way to legitimize influence with the citizenry or to oust any contesting (legal or illegal) group. It is thus reason￾able to assume that the ability to inflict localized vio￾lence over a relatively long period could be expected to translate into influence in different ways. Moreover, as all our results are robust to controlling for violence by the other actor, we posit that municipalities with greater violence are more likely to be captured by the perpetrating armed group.23 19 Correlations are carried out at the department level since munic￾ipalities are the lowest administrative unit in our data. 20 The wealthiest municipalities had tax receipts as much as ten stan￾dard deviations above the mean. 21 The results are substantially the same when we add to the set of controls the log of the municipal population instead of using per capita tax revenue. 22 See Appendix A for detailed descriptions of all the variables and measures. 23 A more stringent test is to focus on the subsample of places where violence by both groups is recorded, dropping the municipalities where one group does not engage in any violence at all during the sample period. Even with the statistical power reduction implied by 1002 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000333

Endogenous Taxation in Ongoing Internal Conflict FIGURE 3.Relationship between Property Tax Revenues and Attacks per Armed Group and Time period across Colombian Departments Guerilla Attacks:1988 to 1996: Guerilla Attacks:1997 to 200 Guerilla Attacks:2003 to 2006; Guerilla Attacks:2007 to 2010 Tax-evenue:1997 to 2002 Tax revenue:2003 to 2006 Tax revenue:2007 to 2010 Tax revenue:2011 to 2013 Paramilitary Attacks:2003 to 2006: Tax revenue:2007 to 2010 2011o2013 Correlation -0.82-0.67-0.66-0.33☐0.32-0.0☐0.010-0.33☐0.34-0.67☐0.68-1.0 Notes:Violence data from the Human Rights Observatory of the Office of the Vice President,Colombia,and fiscal data from CEDE Revenues are average property tax receipts by specified time period.Attacks are cumulative attacks per 100,000 inhabitants per armed group,by specified time period. We thus follow a growing empirical literature on the proxies of presence.The recent peace agreement Colombian conflict(see,e.g.,Acemoglu,Robinson,and signed by the Colombian government and FARC in Santos(2013),Fergusson et al.(2016)and Fergusson, September 2016 was followed by the demobilization Vargas,and Vela (2018))and use a past stock mea- of over 10,000 FARC combatants in the so-called sure of violence over a period of years as an(imper- Espacios Territoriales de Capacitacion y Reincorpo- fect)indicator of influence.Arjona and Otalora(2011) racion,ETCR(Spanish for territorial areas of training compare existing databases of civil war violence in and reincorporation).The ETCR are located in 25 Colombia to survey evidence on armed groups'pres- of the municipalities of FARC's highest historical ence (for the small subsample of municipalities for influence.The correlation between an ETCR indicator which the latter is available)and conclude that while and cumulative guerrilla attacks of the last conflict violence is likely to underestimate-by roughly the period we study (2007-2010)is 0.3 (significant at same magnitude-both guerrilla and paramilitary con- the 1%level).In contrast,the correlation of ETCR trol,there is a nonnegligible correlation between both with cumulative paramilitary attacks is 0.09.Similarly, measures.24 following a DDR process with the government of To further validate our proposed measure of armed Alvaro Uribe (see Civil War and Capture in Colombia groups'influence,we report the correlations between section),several paramilitary units adding up to over groups'cumulative attacks and other (group-specific) 35,000 ex-combatants demobilized in 36 municipalities where,arguably,their dominance was high.The cor- the limited sample,Appendix D shows that our main results are ro relation between an indicator of these municipalities bust to this exclusion. and cumulative paramilitary attacks over the period 24 The authors do not use cumulative past violence,hence the corre- 1997-2002 is 0.1,and significant at the 1%level.In lation of survey-based presence indicators with our proposed mea- sure is probably larger. 1003

Endogenous Taxation in Ongoing Internal Conflict FIGURE 3. Relationship between Property Tax Revenues and Attacks per Armed Group and Time period across Colombian Departments Guerilla Attacks: 1988 to 1996; Tax revenue: 1997 to 2002 Guerilla Attacks: 1997 to 2002; Tax revenue: 2003 to 2006 Guerilla Attacks: 2003 to 2006; Tax revenue: 2007 to 2010 Guerilla Attacks: 2007 to 2010 Tax revenue: 2011 to 2013 Paramilitary Attacks: 1988 to 1996; Tax revenue: 1997 to 2002 Paramilitary Attacks: 2007 to 2010; Tax revenue: 2011 to 2013 Paramilitary Attacks: 1997 to 2002; Tax revenue: 2003 to 2006 Paramilitary Attacks: 2003 to 2006; Tax revenue: 2007 to 2010 Correlation -0.82 - -0.67 -0.66 - -0.33 -0.32 - 0.0 0.010 - 0.33 0.34 - 0.67 0.68 - 1.0 Notes: Violence data from the Human Rights Observatory of the Office of the Vice President, Colombia, and fiscal data from CEDE. Revenues are average property tax receipts by specified time period. Attacks are cumulative attacks per 100,000 inhabitants per armed group, by specified time period. We thus follow a growing empirical literature on the Colombian conflict (see, e.g.,Acemoglu, Robinson, and Santos (2013), Fergusson et al. (2016) and Fergusson, Vargas, and Vela (2018)) and use a past stock mea￾sure of violence over a period of years as an (imper￾fect) indicator of influence. Arjona and Otálora (2011) compare existing databases of civil war violence in Colombia to survey evidence on armed groups’ pres￾ence (for the small subsample of municipalities for which the latter is available) and conclude that while violence is likely to underestimate—by roughly the same magnitude—both guerrilla and paramilitary con￾trol, there is a nonnegligible correlation between both measures.24 To further validate our proposed measure of armed groups’ influence, we report the correlations between groups’ cumulative attacks and other (group-specific) the limited sample, Appendix D shows that our main results are ro￾bust to this exclusion. 24 The authors do not use cumulative past violence, hence the corre￾lation of survey-based presence indicators with our proposed mea￾sure is probably larger. proxies of presence. The recent peace agreement signed by the Colombian government and FARC in September 2016 was followed by the demobilization of over 10,000 FARC combatants in the so-called Espacios Territoriales de Capacitación y Reincorpo￾ración, ETCR (Spanish for territorial areas of training and reincorporation). The ETCR are located in 25 of the municipalities of FARC’s highest historical influence. The correlation between an ETCR indicator and cumulative guerrilla attacks of the last conflict period we study (2007–2010) is 0.3 (significant at the 1% level). In contrast, the correlation of ETCR with cumulative paramilitary attacks is 0.09. Similarly, following a DDR process with the government of Álvaro Uribe (see Civil War and Capture in Colombia section), several paramilitary units adding up to over 35,000 ex-combatants demobilized in 36 municipalities where, arguably, their dominance was high. The cor￾relation between an indicator of these municipalities and cumulative paramilitary attacks over the period 1997–2002 is 0.1, and significant at the 1% level. In 1003 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000333

Rafael Ch et al. TABLE 1. Descriptive Statistics of Main Variables Variable Mean Std.Dev. Min. Max. N Panel A:Outcomes. Tax revenues per capita Log property tax revenues per capita 1997-2002 2.32 1.08 1108 Log property tax revenues per capita 2003-2006 2.70 1.00 0 6 1107 Log property tax revenues per capita 2007-2010 2.82 1.06 6 1111 ///.sdy Log property tax revenues per capita 2011-2013 2.94 1.11 0 1111 Cadastral performance Per capita land value 2003-2006 4.60 5.79 0 7 974 Cadastral update lag 2003-2006 6.82 3.87 50 892 Num.cadastral updates 2003-2006 1.49 0.67 0 4 979 Land ownership Land informality rate 2003-2006 0.20 0.23 0 954 Electoral outcomes Uribe+Conservative party coalition 1997 mayor election 0.28 0.45 00 986 Uribe+Conservative party coalition 2000 mayor election 0.26 0.44 955 Uribe+Conservative party coalition 1997 and 2000 mayor elections 0.32 0.46 0 1182 Uribe Conservative party coalition 2003 mayor election 0.23 0.42 0 908 Uribe Conservative party coalition 2007 mayor election 0.60 0.49 0 1106 Uribe Conservative party coalition 2003 and 2007 mayor elections 0.61 0.49 0 A 1182 Uribe+Conservative party coalition 2011 mayor election 0.43 0.50 0 1040 Panel B:Violence. Log guerrilla attacks per capita,1988-1996 -35.42 31.30 -115 0 1182 Log paramilitary attacks per capita,1988-1996 -35.87 31.53 -105 0 1182 'asn Log guerrilla attacks per capita,1997-2002 -29.38 19.80 82 0 1182 Log paramilitary attacks per capita,1997-2002 -30.02 20.04 0 1182 Log guerrilla attacks per capita,2003-2006 -18.12 12.67 53 0 1182 Log paramilitary attacks per capita,2003-2006 -19.61 14.10 53 0 1182 Log guerrilla attacks per capita,2007-2010 -13.94 13.07 48 0 1182 Log paramilitary attacks per capita,2007-2010 -16.49 15.73 5 0 1182 Log guerrilla attacks per capita,2003-2010 -32.06 23.73 -101 0 1182 Log paramilitary attacks per capita,2003-2010 -36.10 27.71 -108 0 1182 Notes:All data sources are described in Appendix A. 235.5010 contrast,the correlation with cumulative guerrilla large set of potential municipal-level confounders,in- attacks is 0.02,and not significant.25 cluding natural resource royalties and transfer pay- Similar to the case of tax revenues,attacks by guer- ments,vote share by political party for mayors'elec- rilla fronts and paramilitary blocks are highly skewed, tions,the location of military bases,a dummy of especially in the 1988-2010 period(see also Figure 4). peaceful"municipalities (those without attacks dur- The most violent municipalities had attacks over eight ing our entire sample period),the number of people standard deviations above the mean.We thus normal- displaced due to the armed conflict,coca production, ize attacks by population so as to capture violence in- and municipal geographic characteristics.We also con- tensity,and use logged values in the regressions(since trol for a resources-endowment additive index.that in normalized variables remain highly skewed).26 cludes the production of gold,silver,platinum,nickel, Appendix A provides a detailed description of both and iron.27 the main and all the additional variables(summarized in Table 1A).All of our specifications control for a Estimation 25 For all the above reasons,we believe that violence is an imper- Because the modern internal conflict in Colombia had fect but valid proxy of armed influence.But even if it was not,the four distinguishable periods with their own intensity strong relationship of violence to changes in local fiscal institutions and dynamics,as summarized in the subsection on that we document in this paper is an important fact that has not been Civil War Dynamics and Capture in Colombia,we previously documented,and that should inform our thinking on how memneam of the ecomes weaker test how violence-related capture during each specific when using a log-linear model with unlogged violence per capita. with t-statistics in the 1.3-1.6 range.though all results retain the same 27 Note that some of these variables are potentially bad controls as sign.This means that the results are surely not driven by outliers in they can respond to cumulative past violence.In the Additional Ro- attacks,as logging the per capita attack variable effectively attenu- bustness Checks subsection,we discuss this issue and show that our ates those observations'influence on the estimates. results are not affected by posttreatment bias. 1004

Rafael Ch et al. TABLE 1. Descriptive Statistics of Main Variables Variable Mean Std. Dev. Min. Max. N Panel A: Outcomes. Tax revenues per capita Log property tax revenues per capita 1997–2002 2.32 1.08 0 6 1108 Log property tax revenues per capita 2003–2006 2.70 1.00 0 6 1107 Log property tax revenues per capita 2007–2010 2.82 1.06 0 6 1111 Log property tax revenues per capita 2011–2013 2.94 1.11 0 7 1111 Cadastral performance Per capita land value 2003–2006 4.60 5.79 0 79 974 Cadastral update lag 2003–2006 6.82 3.87 0 50 892 Num. cadastral updates 2003–2006 1.49 0.67 0 4 979 Land ownership Land informality rate 2003–2006 0.20 0.23 0 1 954 Electoral outcomes Uribe + Conservative party coalition 1997 mayor election 0.28 0.45 0 1 986 Uribe + Conservative party coalition 2000 mayor election 0.26 0.44 0 1 955 Uribe + Conservative party coalition 1997 and 2000 mayor elections 0.32 0.46 0 1 1182 Uribe + Conservative party coalition 2003 mayor election 0.23 0.42 0 1 908 Uribe + Conservative party coalition 2007 mayor election 0.60 0.49 0 1 1106 Uribe + Conservative party coalition 2003 and 2007 mayor elections 0.61 0.49 0 1 1182 Uribe + Conservative party coalition 2011 mayor election 0.43 0.50 0 1 1040 Panel B: Violence. Log guerrilla attacks per capita, 1988–1996 − 35.42 31.30 −115 0 1182 Log paramilitary attacks per capita, 1988–1996 − 35.87 31.53 −105 0 1182 Log guerrilla attacks per capita, 1997–2002 − 29.38 19.80 −82 0 1182 Log paramilitary attacks per capita, 1997–2002 − 30.02 20.04 −74 0 1182 Log guerrilla attacks per capita, 2003–2006 − 18.12 12.67 −53 0 1182 Log paramilitary attacks per capita, 2003–2006 − 19.61 14.10 −53 0 1182 Log guerrilla attacks per capita, 2007–2010 − 13.94 13.07 −48 0 1182 Log paramilitary attacks per capita, 2007–2010 − 16.49 15.73 −55 0 1182 Log guerrilla attacks per capita, 2003–2010 − 32.06 23.73 −101 0 1182 Log paramilitary attacks per capita, 2003–2010 − 36.10 27.71 −108 0 1182 Notes: All data sources are described in Appendix A. contrast, the correlation with cumulative guerrilla attacks is 0.02, and not significant.25 Similar to the case of tax revenues, attacks by guer￾rilla fronts and paramilitary blocks are highly skewed, especially in the 1988–2010 period (see also Figure 4). The most violent municipalities had attacks over eight standard deviations above the mean. We thus normal￾ize attacks by population so as to capture violence in￾tensity, and use logged values in the regressions (since normalized variables remain highly skewed).26 Appendix A provides a detailed description of both the main and all the additional variables (summarized in Table 1A). All of our specifications control for a 25 For all the above reasons, we believe that violence is an imper￾fect but valid proxy of armed influence. But even if it was not, the strong relationship of violence to changes in local fiscal institutions that we document in this paper is an important fact that has not been previously documented, and that should inform our thinking on how conflict influences state-building and development. 26 The statistical significance of the results below becomes weaker when using a log-linear model with unlogged violence per capita, with t-statistics in the 1.3–1.6 range, though all results retain the same sign. This means that the results are surely not driven by outliers in attacks, as logging the per capita attack variable effectively attenu￾ates those observations’ influence on the estimates. large set of potential municipal-level confounders, in￾cluding natural resource royalties and transfer pay￾ments, vote share by political party for mayors’ elec￾tions, the location of military bases, a dummy of “peaceful” municipalities (those without attacks dur￾ing our entire sample period), the number of people displaced due to the armed conflict, coca production, and municipal geographic characteristics. We also con￾trol for a resources-endowment additive index, that in￾cludes the production of gold, silver, platinum, nickel, and iron.27 Estimation Because the modern internal conflict in Colombia had four distinguishable periods with their own intensity and dynamics, as summarized in the subsection on Civil War Dynamics and Capture in Colombia, we test how violence-related capture during each specific 27 Note that some of these variables are potentially bad controls as they can respond to cumulative past violence. In the Additional Ro￾bustness Checks subsection, we discuss this issue and show that our results are not affected by posttreatment bias. 1004 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000333

Endogenous Taxation in Ongoing Internal Conflict FIGURE 4.Guerrilla and Paramilitary Attacks across Colombian Municipalities cumulative violence 1988-2010 4500. 2700. 2000 4000 6000 Attacks^per 100,000 inhabitants Source:Elaborated using violence data from the Human Rights Observatory of the Office of the Vice President,Colombia 4号 ^Sum of guerrilla and paramilitary attacks. period correlates with local tax performance during where TaxRevenuei.is the log property tax rev- & the subsequent period.We examine lagged outcomes enue per capita in a municipality,i,in a given both because tax institutions cannot shift overnight and because studying the relationship between con- periods2 atteim acerranecksescribed violencd Cum.GuerrillaAttacksi.(-Ator-1) and flict events within a given period and tax institutions Cum.ParamilitaryAttacksi.(-Ator-1)capture logged immediately after that period minimizes simultaneity total attacks per capita that accumulate over each problems. period described in the subsection Civil War Dynamics Running a standard panel data approach with mu- and Capture in Colombia subsection,and preceding t; nicipality and time fixed-effects would entail two un- Xi.r is a vector of municipality-level control variables as tenable assumptions:(1)that the conditional impact of described in the previous subsection.29 We also include violence on taxes operates in the short term and(2) a department fixed-effect,yd,to account for any the treatment effect of influence (as proxied by vio- department-level time-invariant heterogeneity across lent incidents)is the same in every period.Both seem each of the 32 second-level administrative units.the substantively problematic.On(1),tax institutions are next level administrative unit up from the municipality. slow to change,in part because mayors and council We are thus working off within-department variation in member have a 3-or 4-year office period (depend- violence,controlling for a range of geographic factors. ing on if before or after 2003)and they update the Hence,our estimates account for the excess tax revenue cadaster and the rates once in that period,if anything. in municipalities that experience more violence by one On(2),the intensity of violence varied significantly party or the other above the department mean.We across periods,so fitting a model that assumes constant report robust standard errors throughout,clustered at marginal effects of violence on outcomes would be the department level.30 incorrect Our identifying assumption with this approach is We therefore estimate an OLS specification at the that levels of violence during each period of the inter- municipality level separately for each of the periods nal conflict were conditionally independent of future mentioned above.With tax revenue,for example,we tax revenues.Our controls are grounded in the rich estimate: literature on civil war,which cites incentives for cap- ture and the advantages (or disadvantages)of terrain Tax Revenueit =a+yd+BCum.GuerrillaAttacksi.(-A to-1) 28 We report multiple posttreatment years to explore potential cu- mulative effects. +BCum.ParamilitaryAttacksi.(-A to-1) 29 Municipal-level production of coca is available only since 1999. Our post-1999 results are robust to adding this variable in the vector +ΦXr+eit Xi.We address potential concerns regarding the inclusion of post- treatment controls explicitly below. t∈{(1997to2002),(2003to2006), 30 We believe this to be a conservative approach.If we follow Ace- moglu,Robinson,and Santos(2013),by simply using robust standard ×(2007to2010),(2011to2013)}, (1) errors,all results become somewhat stronger statistically. 1005

Endogenous Taxation in Ongoing Internal Conflict FIGURE 4. Guerrilla and Paramilitary Attacks across Colombian Municipalities 0 5.0e-04 .001 .0015 .002 .0025 Density 0 2000 4000 6000 Attacks^ per 100,000 inhabitants cumulative violence 1988-2010 Source: Elaborated using violence data from the Human Rights Observatory of the Office of the Vice President, Colombia. ^Sum of guerrilla and paramilitary attacks. period correlates with local tax performance during the subsequent period. We examine lagged outcomes both because tax institutions cannot shift overnight and because studying the relationship between con￾flict events within a given period and tax institutions immediately after that period minimizes simultaneity problems. Running a standard panel data approach with mu￾nicipality and time fixed-effects would entail two un￾tenable assumptions: (1) that the conditional impact of violence on taxes operates in the short term and (2) the treatment effect of influence (as proxied by vio￾lent incidents) is the same in every period. Both seem substantively problematic. On (1), tax institutions are slow to change, in part because mayors and council member have a 3- or 4-year office period (depend￾ing on if before or after 2003) and they update the cadaster and the rates once in that period, if anything. On (2), the intensity of violence varied significantly across periods, so fitting a model that assumes constant marginal effects of violence on outcomes would be incorrect. We therefore estimate an OLS specification at the municipality level separately for each of the periods mentioned above. With tax revenue, for example, we estimate: Tax Revenuei,t = α + γd + β1Cum.GuerrillaAttacksi,(t− to t−1) +β2Cum.ParamilitaryAttacksi,(t− to t−1) +Xi,t + i,t ∀ t ∈ {(1997 to 2002),(2003 to 2006), × (2007 to 2010),(2011 to 2013)}, (1) where TaxRevenuei, t is the log property tax rev￾enue per capita in a municipality, i, in a given period, t, after each of the described violence periods;28 Cum.GuerrillaAttacksi, (t − tot − 1) and Cum.ParamilitaryAttacksi, (t − tot − 1) capture logged total attacks per capita that accumulate over each period described in the subsection Civil War Dynamics and Capture in Colombia subsection, and preceding t; Xi,t is a vector of municipality-level control variables as described in the previous subsection.29 We also include a department fixed-effect, γ d, to account for any department-level time-invariant heterogeneity across each of the 32 second-level administrative units, the next level administrative unit up from the municipality. We are thus working off within-department variation in violence, controlling for a range of geographic factors. Hence, our estimates account for the excesstax revenue in municipalities that experience more violence by one party or the other above the department mean. We report robust standard errors throughout, clustered at the department level.30 Our identifying assumption with this approach is that levels of violence during each period of the inter￾nal conflict were conditionally independent of future tax revenues. Our controls are grounded in the rich literature on civil war, which cites incentives for cap￾ture and the advantages (or disadvantages) of terrain 28 We report multiple posttreatment years to explore potential cu￾mulative effects. 29 Municipal-level production of coca is available only since 1999. Our post-1999 results are robust to adding this variable in the vector Xi,t . We address potential concerns regarding the inclusion of post￾treatment controls explicitly below. 30 We believe this to be a conservative approach. If we follow Ace￾moglu, Robinson, and Santos (2013), by simply using robust standard errors, all results become somewhat stronger statistically. 1005 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:53:05, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000333