Comp Pol FINAL - The Watson Institute for International and Public ...

War Transitions and the New Civilian Security in Latin America

This article appeared in Comparative Politics Vol. 35, No. 1 (October 2002):1-20.

Charles T. Call, Ph.D. Assistant Professor for Research Watson Institute for International Studies Brown University, Box 1970 Providence, RI 02912 (540) 868-9161 [email protected]

Charles T. Call, "War Transitions and the New Civilian Security in Latin America." Comparative Politics, 35:1, October 2002.

During the 1970s and 1980s, Latin America and the Caribbean experienced some of the most notorious human rights violations of the late 20th century.1 From the “disappearance” of thousands in Argentina, to the arrest and torture of one-fifth of Uruguay’s adult population, to the infamous death squads of El Salvador and Guatemala, Latin America became synonymous with state violence under military rule. The sweeping onset of democracy in the hemisphere in the past two decades allows us to examine crucial questions about the transformation of repressive security forces. Does democratization automatically “democratize” internal security forces that turned on their own citizens? Can police forces which were brutal, powerful and unaccountable under authoritarian rule be supplanted by new security systems rooted in respect for citizen rights, elected civilian control, accountability, and professionalism? These questions highlight the importance of internal security forces – police forces, domestic intelligence agencies and their controlling agents – rather than conventional military forces in the quality of political democracy. They also reflect important theoretical debates about democratic transitions and their relevance for the quality of citizens’ everyday lives under democracy. Some scholars argue that democratization holds little possibility for significant changes in military and political power. Others argue the opposite, that democratization leads to a reduction in military power and to more civilian security systems. Still others believe that outcomes are pathdependent, reflecting international factors or domestic processes during the transition period. These competing views offer very different answers to the question, Does democratization change how states treat their citizens? This article adopts a “modes-of-transition” approach to democratization, arguing that how democratic transitions occur – especially how prior authoritarians leave power – significantly shapes political developments after the transition. However, I conclude that modes-of-transition scholars should focus more on the state and war. Internal security reforms are possible in any new democracy, but only certain types of transitions involve significant demilitarization of internal security with important ramifications for statecitizen relations. Where the armed forces exercise control over internal security, as was the case in almost all of pre-transition Latin America, then a weakening of the military is a necessary condition for major reform. My main argument is that significant civilianization of internal security systems in Latin America requires both (a) a transition from authoritarianism, and (b) a weakening of the armed forces. To develop the argument, I introduce two concepts. First, in the vast majority of cases, military forces are weakened when they fail to win in warfare. Therefore, I use the term “war transitions,” defined as democratic transitions where the armed forces are strategically defeated by an enemy army or forced to negotiate an end to war. Much of

1 I am grateful to Bill Stanley, Katrina Burgess, Tracy Fitzsimmons, Terry Karl, Larry Diamond, Philippe Schmitter, David Holloway, Alexander George, and Lissa Ziegler for helpful comments at various stages of this project, and to the MacArthur Foundation for a Research and Writing Grant, to the U.S. Institute of Peace, and to the Institute for the Study of World Politics.

1


my analysis focuses upon war transitions, a concept which may prove useful for thinking more broadly about post-authoritarian issues. Second, I introduce the term “new civilian security” to refer to an emergent set of features of reformed internal security systems in the region. The new civilian security consists of seven indicators – ranging from restricted military roles in internal security to civilian controlled police forces to civilianized intelligence services – which distinguish it from prior militarized forms of internal policing. The elements of the new civilian security are common in liberal industrial democracies, but relatively new and rare in Latin America. In this article, I first describe competing theoretical claims about security reforms in new democracies. I then present the main argument and demonstrate that “war transitions” account for the demilitarization of internal security in post-WWII Latin America better than other arguments. After illustrating the argument through cases of war transitions in El Salvador and Argentina and through non-war transitions in Brazil and Honduras, I address the limits of the new civilian security and explore how well the argument might travel to other regional contexts. I.

Democratization and State Institutional Reforms

Scholars of democratization generally fall into one of four categories regarding the significance of transitions to democracy for state institutional reform. Those I call “optimists” believe that democratization will inevitably or naturally open the way for institutional reforms that have a positive influence in citizens’ everyday lives. Francis Fukuyama, for example, argues that Western-based liberal democracy is the best political regime and will inevitably spread throughout the world.2 He and many Western policymakers have tended to unquestioningly embrace the notion that democratization is likely to lead to the adoption of Western models of economic, justice and security institutions. Some scholars embrace this “optimistic” view of democratization’s importance as it applies to the reform of security institutions and policies. Sam Huntington, for example, claims that most new democracies experienced improved civilian control over the military during the early 1990s, despite predictions to the contrary.3 Scholars specializing in civil-military relations in Latin America have developed this view more fully. Wendy Hunter, in a study of Brazilian civil-military relations after that country’s 1985 transition to democracy, explicitly challenges the idea that military reforms are unlikely in a military-controlled transition.4 She shows that certain military reforms 2 Fukuyama also believes that Latin America’s political culture has been a barrier to state reform after democratization. The Great Disruption (New York: Free Press, 1999), pp. 11-12. 3 Samuel P. Huntington, “Reforming Civil-Military Relations,” Journal of Democracy 6 (October 1995), 917. 4 Wendy Hunter, Eroding Military Influence in Brazil (Chapel Hill, NC: University of North Carolina Press, 1997), pp. 6-16.

2


occurred during Brazil’s transition to democracy. Her argument is that in all new democracies, electoral competition provides civilians with both the capacity and the incentive to reduce constraints on their power, including military influence. Her treatment is perhaps the best developed case for what I call the “optimistic” view of democratization’s possibilities for security reform in Latin America.

Table 1 SECURITY REFORMS UPON DEMOCRATIZATION: THEORETICAL APPROACHES Theoretical Approaches Examples of Theorists Hypotheses Regarding Security Reforms 1. “Optimists” S. Huntington, W. Hunter, Democratization will gradually lead to (Democratic Gradualists; F. Fukuyama military and police reforms, rule-of-law. Electoral Dynamics) 2. “Skeptics” J. Petras and F. Leiva, Democratization is unlikely to lead to (Marxists; Culturalists; H. Wiarda, C. Anderson, military or police reforms. Organization Theory) O. Marenin W. Robinson, M. Huggins Security reforms will reflect interests of 3. “Internationalists” dominant international actors, system. (Neo-Realists; Dependency Theory; Neo-Gramscian) No security reforms in controlled or imposed 4. “Transitionists” G. O’Donnell, S. Valenzuela, transition. Negotiated security reforms in Modes-of-Transition T.L. Karl, F. Hagopian, pacted transitions. Security reforms are P. Schmitter, F. Aguero, likely in transitions by collapse. S. Mainwaring

At the opposite extreme we find the “skeptics”. This is the second of four approaches depicted in Table 1. Skeptics tend to believe that political transitions to democracy are unlikely to lead to any meaningful institutional reforms (I here use the narrow definition of political democracy as regular, fair electoral competition with the minimal liberal guarantees for individual and collective political action). Skeptics come from diverse schools of thought. Some represent neo-Marxist approaches that view political institutions and regimes as insignificant without changes in socio-economic structures.5 Others see political culture as determinant of the extent to which democratic institutions can take root. They see the Latin, Catholic, and patronage-based political culture of Latin America as unlikely to permit formal democracy from translating into effective Weberian state institutions.6 Scholars who study organizations believe that significant organizational transformation is unlikely, and almost always lengthy, resisted,

5 James Petras and Fernando Leiva, Democracy and Poverty in Chile (Boulder: Westview Press, 1994). 6 Howard Wiarda, Politics and Social Change in Latin America: Still a Distinct Tradition? 3rd edition. Boulder, CO: Westview Press, 1992; and "Political Culture and National Development: In Search of a Model for Latin America," Latin American Research Review 13 (1), 1978; Douglas Chalmers, “The Politicized State in Latin America,” in James M. Malloy, (ed.) Authoritarianism and Corporatism in Latin America (Pittsburgh, PA: University of Pittsburgh Press, 1977); C. Anderson, “Toward a Theory of Latin American Politics,” in Howard J. Wiarda, Politics and Social Change in Latin America: Still a Distinct Tradition (Boulder, CO: Westview Press, 1992).

3


and incomplete.7 All of these theoretical approaches share one thing: skepticism that transitions to democracy will lead to meaningful reforms. A third view, reflecting diverse theories of international relations, might best be termed “internationalist.” Neo-realists believe that major powers largely dictate the nature of interstate relations, and often the very nature of less powerful states. On the other hand, post-colonial and dependency theorists have long contended that the form of states is determined mainly by the needs of the dominant international economic system, especially the drive to secure access to cheap labor and raw materials without political instability. For example, William Robinson argues that the nature of new democracies and their police forces mainly reflect U.S. hegemonic efforts not simply to export its own models, but to achieve its own interests through other states.8 Others share the view that foreign police and security institutions are heavily shaped by the policies and machinations of dominant powers and the ideas about state organization they impose on less powerful countries.9 In concrete terms, the military and police forces of less powerful countries will be organized to suppress threats to international business or geostrategic interests (e.g. communist insurgents during the Cold War; labor unrest; drug trafficking; terrorism). These views are path-dependent, as the specific form of security institutions and missions depends upon the often contradictory interests of international forces. A fourth approach also embraces path dependency, but focuses on domestic actors and their roles rather than international interests. This view, known as the “modes-oftransitions” (or “transitionist”) approach, argues that the process by which democracy comes about determines the scope of institutional reform and future political action. Among “transitionists,” different paths lead to different possibilities for political action for both elected officials and civil society. Karl, for instance, distinguishes between “imposed,” “reform,” “pacted” and “revolutionary” transitions to democracy.10 She maintains that “pacted” transitions offer the best chance for democratic consolidation. Others distinguish among regime transitions based on the degree to which the prior authoritarians –usually the military in Latin America – controlled the process.11 In controlled transitions, militaries are able to create authoritarian enclaves and keep 7 James G. March and Herbert A. Simon, Organizations (New York: John Wiley and Sons, 1958); Otwin Marenin, “United States Police Assistance to Emerging Democracies.” Policing and Society. 00:1-15, 1997; and Policing Change, Changing Police (New York: Garland Press, 1996). 8 William I. Robinson, Promoting Polyarchy: Globalization, U.S. Intervention and Hegemony (Cambridge: Cambridge University Press, 1996). 9 Martha Huggins, . Political Policing: Internationalizing Security Through U.S. Assistance to Latin American Forces (Durham: Duke University Press, 1998); Michael McClintock, Instruments of Statecraft, New York: Pantheon Books, 1993; and The American Connection, Volume One (London: Zed Books, 1985). 10 Terry Lynn Karl, "Dilemmas of Democratization in Latin America." Comparative Politics 23(1) (October 1990). 11 Scott Mainwaring, Guillermo O'Donnell, and J. Samuel Valenzuela, Issues in Democratic Consolidation: The New South American Democracies in Comparative Perspective (South Bend, IN: University of Notre Dame Press, 1992); Felipe Aguero, "The Military and the Limits to Democratization in South America," in Scott Mainwaring, Guillermo O'Donnell, and J. Samuel Valenzuela, op cit.

4


important arenas of power and policymaking out of the reach of new civilian governments. In other transitions, where authoritarian regimes collapse due to economic crisis or loss in war, democratizers face less resistance to reforms and have greater impact on the economic or political life of everyday people. In general, modes-of-transition scholars are skeptical of meaningful internal security reforms under controlled transitions, and optimistic about reforms after collapse.

II.

Why War Transitions?

Regime transitions toward democracy inherently carry high degrees of uncertainty about the rules of the political game.12 While such uncertainty opens the way for changes in the rules of governance, not all realms of governance are equally open to change. Amidst heightened fears about political fates and even physical survival, the actors who control the state try to hold onto power, often by maintaining control of state institutions which might otherwise succumb to newly elected rulers.13 Given the historical control exercised by the armed forces over policing in Latin America, the military’s influence over internal security systems is paramount in analyzing the transformation of those systems. Democratic transitions pose institutional and personal threats to the military leadership for three reasons. First, loss of control over internal security forces and investigators reduces the military command’s ability to use such forces to block prosecution of its members for human rights violations. Second, losing control of police units, especially the customs and anti-drug units that deal with smugglers, may also mean losing access to bribes and other illicit income deriving from corruption. Third, losing influence over police organizations and doctrine may undermine the military’s ability to defend the country from what it perceives are internal threats to the state. These reasons generally combine to outweigh the military’s general preference not to get involved in time-consuming policing of common crime. Consequently, the armed forces tend to strongly resist the loss of any internal security prerogatives during democratization. Ruling militaries -- where not weakened -often demand retention of their prerogatives, including internal security responsibilities, as a quid pro quo for relinquishing the reins of government.14 For converse reasons, "democratizers" (i.e., regime opponents, many of whom are not so democratic) who suffered at the hands of the prior regime will seek not only to remove the military from politics, but to remove its role in internal security as a means of ending its persecution of 12 Guillermo O’Donnell and Philippe C. Schmitter, Transitions from Authoritarian Rule: Tentative Conclusions, Volume 4 (Baltimore: Johns Hopkins University Press, 1986). I use the term "democratic transitions" to refer to transitions toward democratic regimes, even when such regimes are not fully democratic. 13 All regime transitions share these qualities, but transitions toward democratic regimes render unlikely certain types of internal security systems, namely those dominated by a single political party which mobilizes political partisanship. 14 Alfred Stepan, Rethinking Military Politics (Princeton: Princeton University Press, 1988); Aguero, op cit.

5


civilians and building the foundation for political freedoms and pluralism. Unless the military is significantly weakened, it will prevail in prohibiting significant demilitarization of internal security systems. Policing is a core element of the state. Some dislocation of military power is necessary to open the way for democratizing forces to exact changes in the security apparatus. The weakening of the armed forces may happen in many ways. However, by far the most common route to the dislocation of the military’s power are “war transitions,” where democratization occurs in conjunction with a military’s obligation to negotiate with or surrender to enemy armed forces.15 War transitions offer particular circumstances that open the way to steps toward demilitarization of diverse arenas of society and the state. A military force is above all expected to win wars, and its political support is undermined when its inability to defeat an enemy army becomes apparent. When wars and war termination processes unfold in conjunction with political democratization (and liberalization), diverse political actors are jockeying for power. Political support for the armed forces is thus not assured. In highly militarized societies such as most of Latin America, a weakening of the military is necessary to open the door to civilianization of internal security. Specifically, the two-fold argument here is that (1) dislocation of military power and democratization are necessary conditions for significant demilitarizing reforms of internal security systems in the region, and (2) a particular form of this combination – “war transitions” – makes such reforms highly likely. While fundamentally reflecting a modes-of-transition approach, my argument diverges in important ways from existing literature. First, I emphasize the role of states as well as regimes in analyzing the quality of governance. Second, this argument follows in the tradition of Tilly and others who argue that wars are important in the evolution of states and their institutions.16 Rather than focusing on the process of making war, however, I stress the process by which wars end. Finally, I reject the blanket skepticism of scholars regarding imposed transitions.17 Imposed transitions may permit significant security reforms, depending upon who controls the process. As most “transitionists” argue, if prior rulers are not weakened and 15 The term “war transition” was first used in an unpublished 1996 proposal by Terry Lynn Karl, Vincent Maphai and Rubén Zamora, “War Transitions: Ending Armed Conflict and Starting Democracy in `Uncivil' Societies.” “War termination” here refers to the process of transition from armed conflict to the absence of national-level armed conflict, where the prior regime fails to emerge victorious. War transitions as used herein include democratization in conjunction with war termination of three sorts: interstate wars between two or more states, internal armed politico-military conflicts, and international interventions. The list of internal conflicts up until 1989 is drawn from Paul R. Pillar, Negotiating Peace: War Termination as a Bargaining Process, (Princeton: Princeton University Press, 1983), and Stephen John Stedman, Peacemaking in Civil War: International Mediation in Zimbabwe, 1974-1980 (Boulder, CO: Lynne Rienner Press, 1991). Subsequent classifications are my own. 16 Charles Tilly, Coercion, Capital and European States, A.D. 900-1990 (Oxford and Cambridge: Oxford University Press, 1990); and The Formation of National States in Western Europe (Princeton: Princeton University Press, 1975). 17 Terry Lynn Karl, "Dilemmas of Democratization in Latin America, and Guillermo O’Donnell, “Transitions, Continuities and Paradoxes,” in Mainwaring et al.

6


thus able to control transitions, then the chances for reform are poor. However, if international actors intervene and occupy a country, imposing democracy, then the possibilities for reform are not necessarily bleak. The outcome of security reforms after external interventions depends upon the interaction of the occupying power and national leaders whose governance is recognized. Below I turn to the evidence. III.

The Evidence

The military character of internal security in Latin America dates from its colonial days when military forces generally controlled internal security in the absence of specialized police forces. When police forces first emerged in the late 1800s, they were often drawn from active-duty soldiers and almost always trained by military officers in military academies. Only exceptional populist or democratic regimes such as Guatemala (1944-54), Bolivia (1952-64), Argentina under Perón, Costa Rica (1948-present), and Venezuela (1945-48), where policing was generally politically partisan, escaped this pattern, and usually only briefly. In larger countries such as Brazil, Argentina and Mexico, federal police were erected alongside decentralized provincial police who were often under local military commanders. But in most countries of the region, administrative police were unified, national forces. For most of the century between 1880 and 1980, these were commonly under the Ministry of Defense or War. And even in elected civilian governments where policing came under Ministries of the Interior or Justice, police training, police doctrine and the internal structure of the police remained highly military in nature. Constitutions almost uniformly granted military forces internal security missions.18 Historical military control and influence in policing and intelligence does not mean that no steps toward civilianization occurred. The structure, arms and functions of the earliest urban police forces, which appeared in the late nineteenth century in most of the region, were highly military. However, these forces operated with significant autonomy from military units, and often were commanded by civilian elites. A surprising number of countries in Latin America separated national police forces from Ministries of War (or Defense) at the close of World War II as well. Nevertheless, the Cold War and the fear of communism led many states to remilitarize their internal security forces in the early 1950s. By the late 1970s, military authoritarian regimes and civil wars in Central America and other countries left military prerogatives as high as at any time in the region’s history. Militarization and Civilianization. In Latin America, transformation of abusive internal security systems (that is, the laws and institutions governing how the state may use coercion inside its borders) focused first on their demilitarization. What does “demilitarization” mean? In the context of militarized policing structures, many scholars and policymakers focus heavily on one aspect of reform: the shift of hierarchical 18 Brian Loveman, The Constitution of Tyranny: Regimes of Exception in Spanish America (Pittsburgh: University of Pittsburgh Press, 1993).

7


command from military commanders to civilian officials.19 Yet this concept of demilitarizing police reforms is too narrow. In Latin America, shifting ministries may neither change the fundamental nature of a militarized internal security organization nor remove the armed forces' influence over policing. The extent of militarization of internal security systems encompasses the military’s internal security responsibilities, the nature of policing, internal intelligence and oversight of internal security. Here I use seven indicators of levels of militarization/civilianization of internal security.20 First, to what extent do the military’s constitutional functions include internal security as a permanent core mission, a permanent supplemental mission, or an exceptional mission? Second, are the main administrative/preventive police forces under the defense ministry (or the military command structure); under “civilian” ministries such as Interior, Justice, or Security; or under some partial or mixed arrangement? Third, to what extent are senior police personnel active-duty military officers; predominantly not active-duty or ex-military officers; or some mixture? Table 2. INTERNAL SECURITY INDEX

Prerogative A) Const’l. Military Internal Security Role B) Main Preventive Police Forces C) Senior Police Officers

“Militarized Security” 1 Permanent Core Mission Under Defense Ministry or Military Command Active-Duty Military

D) Main Intelligence Agency E) Legislative Oversight over Main Police Forces F) Legal Jurisdiction over Police Personnel

Reports to Defense Ministry or Military None or token

G) Formal Police Doctrine

Mainly Defense of the State

Military courts normally handle police personnel

2 Permanent Supplemental Mission Mixed, with some military command Some Active-Duty (or mostly former military, or under military with separate career track) Military role in directing, but not control Can alter budget, but little voice in operations or top appointments Mixed

Defense of state and citizens

“New Civilian Security” 3 Only in Exceptional Circumstances Under civilian minister (not Defense ministry) No Active Military & Few Ex-military

Reports to Civilian with no mil. role in selection Legislative powers over budget, forces, weapons; can compel testimony Civilian courts routinely handle all human rights violations. Mainly defense of citizens & their rights

Fourth, does the main internal intelligence agency report to the defense ministry (or armed forces); to a civilian official (usually the president); or to a civilian but with significant military control? Fifth, is civilian oversight of internal security absent, 19 See Stepan, op cit, (he also includes senior personnel) and Philippe C. Schmitter, "Dangers and Dilemmas of Democracy,” in Larry Diamond and Marc Plattner, The Global Resurgence of Democracy. 2nd edition (Baltimore: Johns Hopkins University Press, 1996). 20 See similar lists of military or police prerogatives in Alfred Stepan Rethinking Military Politics, and David Bayley, “What's in a Uniform?: A Comparative View of Police-Military Relations in Latin America." Paper presented at Woodrow Wilson Center Conference on "Between Public Security and National Security,” Washington, D.C., October 1993.

8


extensive, or limited? Although a variety of oversight mechanisms may exist, I have here limited the analysis to legislative oversight. Sixth, do the personnel of the main internal security forces fall under civilian (ordinary) judicial jurisdiction, or under the military justice system for gross human rights violations such as extrajudicial executions and torture? Finally, is police doctrine (as found in constitutions or police organic laws) principally focused upon the defense of the state and public order; on the protection of citizens and their rights; or on both to roughly equal measure? These seven indicators are measured in three categories “1,” (the most militarized category) “2” (an intermediate score), or “3” (the least militarized category). Table 2 depicts the seven indicators of “militarization of internal security” and their range of variation. As an example, Brazil’s “police ministry” would receive a score of “1” in 1982 when its main administrative police, the state-level “military police” forces, were under army command, but a “2” in 1988 when they passed to the command of state governors but remained “ancillary and reserve forces of the army”.21 A maximum overall score of “21” (“3” on all seven indicators) represents the “new civilian security,” while the minimum score of “7” (“1” on all seven indicators) represents a wholly militarized system. In Table 3, every regime change toward democracy (“democratization”) in Latin America, plus Haiti, has been scored twice on each of the seven indicators, first immediately before the transition and then (separated by a comma) at the end of what I refer to as a “transition period” of usually four to five years.22 Table 3. Internal Security Changes upon Democratization, 1945-2000: Pre-Transition and Post-Transition Scores and Totals

Argentina ’83 Bolivia ’82 Bolivia ’52 Brazil ’85: Chile ‘90 Colombia ’58 Costa Rica ’48 Ecuador ’80 El Salv. ’92 El Salv. ’82 Guatemala ’96 Guatemala ’86

A Milit. Role

B C D Police Police Intel. Mnstry Officers

E Legis. Role

F G Totals Legal Police (Pre,Post) Jurisd. Doctrine

1,3 1,1 1,1 1,2 1,1 1,1 1,3 1,1 1,3 1,1 1,1 1,1

2,3 3,3 1,3 1,2 1,1 1,1 2,3 2,3 1,3 1,1 2,3 2,2

1,3 1,2 1,1 1,1 1,1 1,1 2,3 2,2 1,3 1,1 1,2 1,1

1,3 3,3 1,3 1,1 1,1 1,1 3,3 1,1 1,3 1,1 3,3 3,3

2,3 3,3 1,2 1,2 2,2 2,2 3,3 3,3 1,3 1,1 2,3 1,2

1,3 1,3 1,3 2,2 1,1 3,3 2,3 1,1 1,3 1,1 1,3 1,1

1,1 2,2 1,2 1,1 2,2 2,2 2,3 2,2 1,3 1,1 2,3 2,2

9, 19 14, 17 7, 15 8, 11 9, 9 11, 11 15, 21 12, 13 7, 21 7, 7 12, 18 11, 12

Overall Change

10 3 8 3 0 0 6 1 14 0 6 1

21 Hunter, p. 52. 22 The post-transition scores are based on the internal security characteristics either five years after the transition, or at the end of the first post-transition presidency, whichever occurs first. These scores are author and research assistant’s judgments based on secondary literature for all cases (cites available on request), including studies of civil-military relations, human rights reports, the U.S. Army’s Country Study series, plus primary documents and personal interviews conducted in Central America and Haiti.

9


Haiti ’94 Honduras ‘80 Nicaragua ’90 Panama ’89 Paraguay ’89 Peru ’80 Uruguay ’85 Venezuela ’58

1,3 1,1 1,1 1,3 1,1 1,1 1,1 1,1

1,3 1,1 2,3 1,3 3,3 3,3 2,3 2,2

1,3 1,1 2,2 1,3 2,3 3,3 1,2 1,1

2,3 1,1 2,1 1,3 2,2 1,1 1,1 2,2

1,3 1,1 1,2 1,3 1,2 1,2 2,2 2,2

1,3 1,1 1,2 1,3 2,2 1,1 2,2 2,2

1,2 1,1 1,3 1,2 2,2 2,2 2,2 1,2

8, 20 7, 7 10, 14 7, 20 13, 15 12, 13 12, 14 11, 12

12 0 4 13 2 1 2 1

10


Before considering the outcomes on Table 3, several features of the region’s internal security systems are notable. First, the great majority of the internal security systems in Latin America were highly militarized under authoritarian regimes. For example, in every single country that experienced a transition toward democracy after 1945, the armed forces enjoyed permanent constitutional or legal internal security prerogatives (score: “1”) before democratization. Parliamentary oversight and control over domestic intelligence were also minimal (average scores “1.2” and “1.4”). Second, this situation changed dramatically in several countries that achieved high scores of civilianization after democratization. Post-transition Argentina, El Salvador, Costa Rica, and Panama, some of which had highly militarized systems earlier, all scored within 2 points (“19” to “21”) of a fully civilianized system. Yet variation remains high. Post-democratization totals vary from the high scores just noted to the low end of the scale in Brazil (“11”), Chile (“9”) and Honduras (“7”). Third, the degree of change between pre-transition internal security and posttransition systems varies as well. El Salvador (1992) scored a maximum 14-point change, for instance, while several countries experienced no change. Of the pool of twenty transitions toward democracy between 1945 and 2000, seven countries experienced what can be considered significant, or major, civilianizing internal security reforms. These countries showed at least six points of change between their immediate pre-transition and their post-transition security systems. The remaining fourteen states experienced less than a six-point differential, with most of these marking very little change (zero or one point). What explains the variation in security reforms? First, the variation in outcomes belies both “optimistic” and the “skeptical” views of post-transition institutional reforms. Optimists cannot adequately account for cases such as Venezuela 1958, Peru 1980, Brazil 1985, Chile 1990, Ecuador 1980 or early, partial transitions in El Salvador (1982) and Guatemala (1986), none of which experienced significant reforms upon transition. Optimism, perhaps warranted regarding the decline of military political prerogatives after authoritarianism, clearly does not extend to internal security prerogatives. Skeptics cannot account for cases such as Haiti 1994, Argentina 1983, Panama 1989, and El Salvador 1992, all of which experienced sweeping demilitarization. Whereas reforms may not prove as sweeping or significant as democratizers might like, it is inaccurate to claim that no important security reforms have been achieved. International explanations conform to some cases of demilitarizing security reforms. Of the four most far-reaching reforms – Argentina 1983, El Salvador 1992, Panama 1989, and Haiti 1994 – international actors played direct roles in three cases, making it possible in the latter two countries for restored elected leaders to eliminate the armed forces. On the other hand, significant security reforms occurred in Argentina and

11


Costa Rica 1948 without any significant international influence on decisionmakers.23 At the same time, in transitions where the United States played an important role -Honduras 1980 and earlier partial transitions in El Salvador 1982 and Guatemala 1986 – no demilitarization of internal security occurred. Instead, with U.S. support, internal security systems remained heavily militarized during democratic transitions rooted in counterinsurgency efforts.24 Furthermore, nothing indicates when, or under what conditions, international factors lead to internal security reforms. The region’s post-1945 experience suggests that, where international actors are heavily involved in regime transitions, especially within Central America and the Caribbean, international factors may shape internal security reforms. Yet none of these views is as compelling as the war transitions argument. Table 4 below divides cases of democratization by the occurrence of war transitions, on one axis, and the occurrence of major internal security reforms on the other. Security reforms highly coincide with cases of war transitions. Significant internal security reforms occurred in seven (7) out of nine (9) war transitions (77%), but in none of eleven (11) democratic transitions which occurred absent war termination or intervention. Of the seven significant police reforms that occurred in Latin America in conjunction with democratization, all occurred during war transitions. Eighteen of the twenty cases, or ninety-five percent (95%), conform to the war transitions argument. Table 4. War Transitions and Internal Security Reforms War Transition? Yes

Yes

Major Demilitarizing Internal Security System Reform?

Argentina ’83 Guatemala ‘96 Bolivia ’52 Haiti ’94 Costa Rica ’48 Panama ’89 El Salvador ’92 (7)

Colombia ’58 Nicaragua ’90

No

(2)

No

(0)

Bolivia ‘82 Honduras ’80 Brazil ’85 Paraguay ’90 Chile ‘90 Peru ’80 Ecuador ’80 Uruguay ’85 El Salvador ’82 Venezuela ’58 Guatemala ’86 (11)

23 Alison Brysk, The Politics of Human Rights in Argentina: Protest, Change and Democratization (Stanford: Stanford University Press, 1994); J. Patrice McSherry, Incomplete Transition: Military Power and Democracy in Argentina (New York: St. Martin’s Press, 1997); Leonard Baird, Costa Rica: The Unarmed Democracy (London: Sheppard Press, 1984). 24 Washington Office on Latin America (WOLA), The Administration of Injustice: Military Accountability in Guatemala (Washington, DC: WOLA, 1989); William D. Stanley, The Protection Racket State: Elite Politics, Military Extortion and Civil War in El Salvador (Philadelphia: Temple University Press, 1996).

12


Even if we accept the imprecision of the scale and the subjectivity of the cut-off mark for “significant reforms,” a general coincidence between war transitions, on the one hand, and major internal security reforms, on the other hand, seems to exist. This relationship is complex, but important for the everyday lives of citizens who live amidst heightened fears of violent crime and an incomplete rule of law. Below I present brief case studies that illustrate and elaborate on the link between war transitions and major security reforms. IV.

Illustrating and Elaborating the Argument

Whereas quantitative data point to an association between war transitions and demilitarization of internal security, case studies help elucidate the causal link between the weakening of the armed forces, democratization, and security reforms. Four brief cases below illustrate that a weakening of the military under democratization is necessary for significant demilitarization of internal security in Latin America. However, these necessary conditions do not hold for less significant security reforms. Such reforms tend to follow a pattern: during the highly uncertain transition period (measured here as the shorter of either the first elected presidential administration or the first five years after an elected government assumes office), war transitions involve significant reforms, and cases of non-war transitions experience only piecemeal internal security reforms. After that transitional period (of roughly five years), post-war cases where military weakening opened the way to deeper reforms often experience a backlash which reverses some of the progress made toward demilitarization. Cases such as El Salvador, Argentina and Guatemala exemplify this phenomenon. Conversely, in non-war transitions such as Brazil and Chile, where no demilitarization occurred during the initial transition period, then some comparatively minor demilitarizing reforms are possible later. Below I illustrate these divergent patterns with case examples, and then suggest explanations. El Salvador. In 1992, El Salvador’s twelve-year civil war between the Marxist Farabundo Marti National Liberation Front guerrillas (FMLN) and the US-backed rightist government ended in a UN-mediated settlement. El Salvador’s three main internal security forces and intelligence agencies were highly militarized, under military command, staffed entirely by active-duty army officers, and thoroughly oriented toward defending the state in accord with counterinsurgency doctrine. During the 1980s, the military and police forces gained renown for their violations of international humanitarian law, overseeing death squads responsible for the majority of extrajudicial executions carried out during the war.

13


Comprehensive peace agreements provided for among the farthest-reaching military, police and human rights reforms in Latin American history. As a result, the military’s role in internal security was curtailed, and a single National Civilian Police (PNC) was created to supplant the prior security forces. Its doctrine emphasized human rights and citizen protection, its armament and training underscored limits on the use of force, and its officers and ranks were predominantly civilians with no military background, although ex-policemen and ex-guerrillas each comprised roughly 20% of the top ranks. The legislature gained a role in the appointment and removal of the PNC director, a new human rights ombudsman took office, and the president assumed constitutional direction of a new intelligence service. High hopes greeted the newly outfitted police. However, violent crime skyrocketed in the wake of war. Between 1993 and 1997, crime was the top-ranked concern of the population, and recorded homicides virtually tripled within two years of the peace agreement.25 Thirty-four per cent (34%) of respondents reported in January 1996 that they or an immediate family member had been the victim of a violent crime in the previous four months. The crime wave empowered critics of the negotiated civilianizing police reforms, who proposed rolling some of them back. In the end, not all of their proposals were accepted. By 1998, the government had reversed some new protections for minors in light of extensive youth involvement in gang-related crime, reversed other guarantees for suspects, and altered a progressive new criminal procedure code to retain legal recognition of some extrajudicial confessions.26 It also extended the list of eligible automatic weapons permitted the PNC. In addition, vigilante justice, including killings of susected thieves, appeared after the war. This remilitarization of internal security made only a dent in civilianization achieved during El Salvador’s war transition. Argentina. Despite the salient differences between Argentina’s “bureaucraticauthoritarianism” and El Salvador’s pre-war regime, indicators of militarization of public security were similar before the Malvinas War of the early 1980s. Despite continuing under the Interior Ministry, the federal police was commanded by active-duty military officers, as were provincial police forces.27 The military junta ultimately controlled all these forces. The federal and provincial police forces were less directly implicated in human rights violations than their military counterparts, but engaged in widespread torture and extrajudicial executions.28

25 Charles T. Call, Sustainable Development in Central America: The Challenges of Violence, Injustice and Insecurity, Report for Proyecto Centroamerica 2020 of the European Commission and US Agency for International Development, Hamburg, 2000. 26 Margaret Popkin, Peace without Justice, (University Park, PA: Penn State University Press, 2000). 27 Argentina’s federal organization (as in Venezuela, Brazil, and Mexico) complicates comparison. I focus upon the main “preventive police forces,” here the federal police and the Buenos Aires provincial police. 28 Brysk 1994.

14


In contrast to its neighbors, Argentina experienced democratization in conjunction with the termination of a war. The ill-fated loss to Great Britain in 1982 discredited the military and precipitated its exit from direct governance. As McSherry says, “The 1982 defeat…meant that the military was not able to permanently entrench guardian capabilities nor impose other demands in major foundational pacts, as in other Latin American cases.”29 Over its strenuous objections, the Argentine military leadership was subjected to criminal trials for “disappearances” and lost key internal security prerogatives. The hotly contested Defense Law of 1988 prohibited the military from planning for internal conflicts, combating internal threats, and conducting internal intelligence, except when the president deemed the police overwhelmed.30 Civilian President Raul Alfonsin replaced military police chiefs with civilians, and legislative oversight was bolstered. As in El Salvador, a backlash ensued. Argentina did not experience the same degree of post-war common crime as the civil war cases in Central America. However, it survived three military rebellions sparked by attempts to curb military power. Alfonsin and military allies in the legislature adopted measures to placate the armed forces, effectively gutting some of the civilianizing measures. Alfonsin issued Decreee 83/89 after the third military uprising in 1989, incorporating military branch chiefs and the Joint Chiefs into governmental command and planning structures for internal security.31 He established a Committee for Internal Security, permitting an intelligence role and new emergency internal powers for the military. A 1991 Internal Security Law reopened the military’s role in some forms of internal conflicts. As in El Salvador, the main outlines of civilianizing reform remained, but chipped away through newer laws and presidential decrees. Brazil. If the pattern in war transitions is significant demilitarization followed by minor steps toward remilitarization, then the pattern where military forces are not weakened is the opposite. No significant demilitarization occurs, but minor reforms may occur. In general, these reforms tend toward demilitarization. Brazil illustrates both trends. Although a small federal police and small municipal forces exist in Brazil, most policing occurs at the state level. As in Argentina, Brazil’s main preventive forces, the state military police (“policia militar”) came under military control during authoritarianism, along with planning, intelligence and other elements of internal security. According to Hunter, the military’s role in internal security proved to be “one of the most inflammatory issues of the constitutional debates” at democratization.32 By 1990, Brazil’s controlled, non-war transition to democracy left the armed forces with a clear continued role in law and order, retained the status of the state military police forces as “ancillary and reserve forces of the army,” left the main internal intelligence agency in 29 McSherry p. 2. 30 Ibid., 165-68. 31 Ibid., 167. 32 Hunter, p. 47.

15


the hands of the military, and precluded ordinary courts from handling most important human rights charges against police officers.33 In addition, the longstanding military orientation, organization and doctrine of the military police remained unaltered. At the same time, the new regime took some minor demilitarizing steps. It placed the military police under governors in normal times and enhanced civilian control over the military’s internal security powers.34 Despite the emphasis some analysts have placed on these reforms, the overall picture of Brazil’s transition is one of continuity: the armed forces retained important roles in internal security. During the 1990s, after Brazil’s transition period, various levels of Brazil’s government undertook efforts at internal security reforms. In 1996, the National Congress approved the transfer of military police accused of homicide from military tribunals to civilian courts. Torture and the possession of illegal weapons were criminalized for the first time, and human rights crimes achieved automatic federal jurisdiction.35 Police in some major cities were reorganized, and new mechanisms of accountability established. These limited and piecemeal reforms slightly affected an internal security system which remained highly militarized. What explains these subsequent demilitarizing security reforms? One prominent strand of police reform analysis in established democracies, including the United States and Great Britain, attributes police reforms to scandal.36 Where incidents of police abuse or corruption receive scandalous notoriety, then police administrators and their political overseers take action. This appears to be the source of reforms in post-transition Brazil. Here, police personnel were blamed not just for permitting dramatically increased insecurity and crime after democratization, but for contributing to violence and lawlessness. As Brazilian human rights expert Paulo Sergio Pinheiro says, Besides torture, which is almost routine, the summary execution of suspected criminals by the police is a virtual epidemic. In 1999 in the metropolis of São Paulo alone, the military police killed 380 civilians in on-duty incidents--most of which were summary executions--and 197 in off-duty incidents. Compared with other democratic countries (not including those experiencing internal warfare), Brazil has the highest rate of lethal police violence.37 Where the police and justice system are unable to protect citizens, and where state officials themselves participate in extralegal violence, public disaffection with the internal security system ensues. That disillusionment rises sharply after a particularly public case of misconduct. Nevertheless, in non-war transitions, attempted reforms 33 Ibid.; Human Rights Watch, Police Brutality in Urban Brazil (New York: Human Rights Watch, 1997). 34 Hunter, pp. 48-53. 35 Paulo Sérgio Pinheiro, “Navigating in Uncharted Waters: Human Rights Advocacy in Brazil's "New Democracy," NACLA Report on the Americas 34,1 (July/Aug. 2000):47–51. 36 Lawrence W. Sherman, Scandal and Reform: Controlling Police Corruption (Berkeley, CA: University of California Press, 1978). 37 Pinheiro, op cit., p. 48.

16


remain isolated and less sweeping than post-war transition reform processes. Furthermore, they are often truncated when public attention drifts. In contrast to Brazil’s minimal post-transition reforms, another country, Honduras, experienced in the late 1990s a rare sweeping reform. That experience offers additional insights into how major civilianizing security reforms can occur. V.

Weakening the Military Without War

The case of Honduras represents an empirical anomaly. It is the only recent instance of sweeping civilianization of an internal security system absent a war transition.38 Honduras supports my argument that democratization and a weakening of the military are necessary conditions for demilitarizing security reforms, but shows that war termination is not the only path to these outcomes. A brief examination of civilianization of security in Honduras suggests alternative ways to weaken the military without war. In the early 1980s, Honduras experienced a transition to electoral democracy, electing its first civilian president since the 1950s. This regime transition, like those in El Salvador in 1982 and Guatemala in 1985, reflected strong U.S. pressure on right-wing military regimes in Central America to elect civilian presidents and help forestall replication of the Sandinistas’ revolutionary triumph in Nicaragua. The wars in Guatemala, El Salvador, and Nicaragua continued for some years, but Honduras never experienced a full-scale internal conflict. In the early 1990s, investigations into the military’s involvement in torture and “disappearances” sparked an organized effort by civil society groups to reduce the military’s powers, especially in internal security. In 1993, a newly created human rights ombudsman’s office received detailed charges from a former agent of the National Investigations Directorate (DNI), the main criminal investigative police in the country, that his colleagues had engaged in torture, political assassination, and drug trafficking during the 1980s and 1990s.39 The charges followed closely on the heels of the trial of an army colonel for the rape of a teenaged girl. In response, President Rafael Callejas formed an Ad Hoc Commission to propose reforms to the secret police and other state security forces.40 Pursuant to the commission’s recommendations, the armed forces relinquished control of the controversial DNI to civilian command, and additional reforms unfolded over the next five years, resulting in the separation of the main preventive police force from the military. By 1998, the newly named “National Police” operated under a new civilian Security Ministry, alongside a civilian-controlled

38 Honduras’ security reform occurred in 1998, well after its 1980 transition toward democracy. Consequently, it does not appear in the pool of cases in Tables 3 & 4, which include only police reforms occurring in the immediate transition period. 39 “Oficiales y Agentes del DNI Involucrado,” El Heraldo, 20 February 1993, p. 36. 40 “Callejas Reads National Agreement on Reforms,” Voz de Honduras, 2 March 1993, in Foreign Broadcast Information Service, 2 March 1993, p. 9.

17


investigative force under the Attorney General, with a new police law and doctrine.41 Internal security policy was formulated by an appointed commission that included representatives of civil society. Although intelligence matters remained controlled by the military, the formal transformation of policing was as extensive as in the war transitions of the 1980s and 1990s. What explains Honduras’ extensive demilitarization absent a war transition? The answer lies in a weakening of the military. First, the end of wars in neighboring countries, together with the end of the Cold War, produced in Honduras some of the elements common to war transitions. The country embarked upon a process of demobilization of military troops in the early 1990s and a vast reduction in military resources. By 1996, the size of the Honduran Armed Forces dropped from roughly 20,000 to an estimated 12,500, the smallest in Central America outside of Belize, and its budget dropped from $59 million in 1990 to $35 million in 1996.42 The U.S. drastically reduced its military aid, which totaled almost $500 million during the 1980s, to less than $1 million per year from 1994 to 1997.43 These events significantly weakened the material and political base of Honduras’ armed forces. Second, although the Honduran military did not suffer a defeat or have to negotiate with insurgents, it confronted a level of civil society influence and initiative that was much higher than the rest of Central America. A women’s human rights organization, the Movimiento de Mujeres Visitacion Padilla, led a campaign that ultimately led to the abolition of forced military service in 1994. The weakening of the military permitted simmering anger at the military to emerge in public discourse. Some witnesses came forward, and the floodgates opened.44 An unprecedented alliance of nongovernmental organizations, including the private sector, lobbied for the removal of the investigative and preventive police forces from military command. A “Citizens’ Forum on Public Security” successfully influenced the content of a new Police Organic Law in May 1998 in the most extensive public debate since democratization in 1980.45 Other factors, including human rights investigations of past abuses in neighboring countries and a shift in U.S. policy, contributed as well.

41 Personal interviews with four senior Honduran police commanders, Human Rights Commissioner Leo Valladares and U.S. officials, 1997, Tegucigalpa, and with six organizers of the Citizens’ Forum, May 1998, Tegucigalpa. Leticia Salomon, Policias y Militares en Honduras (Tegucigalpa: Centro de Documentación de Honduras, 1993). 42 Adam Isacson, Altered States: Security and Demilitarization in Central America (Washington, D.C.: Center for International Policy and Arias Foundation, 1997), p. 62. 43 Figures from FY 1988-1995 from U.S. Agency for International Development, Congressional Presentation, various years, Washington, D.C. FY 1996-97 figures from Adam Isacson and Joy Olson, Just the Facts, (Washington, D.C.: Latin America Working Group and the Center for International Policy, 1998). 44 “Front Against Army Rule is Building in Honduras,” This Week in Central America: A Report on Business and Politics, XVI, 10, 15 March 1993. 45 Personal interviews, op cit., Tegucigalpa.

18


The experience of Honduras suggests that war transitions are not the only path to weakening of the armed forces. The armed forces were discredited through decisive erosion of its resource base and actions by a morally indignant civil society. These material and ideational factors can produce the same effect as a strategic defeat or negotiated end to war. Where militaries are weakened without war, the mobilization of civil society may prove to be a catalyst of internal security reform. War and intervention are the most common paths to a decisive weakening of the armed forces, but not the sole paths. VI.

The Limits of the New Civilian Security in Latin America

The new civilian security is an ideal type, but not necessarily a normative ideal. Demilitarization is a necessary, but not sufficient, condition for far-reaching changes in police practice. The new civilian security is an underspecified concept that is compatible with diverse forms of security activities, police patrol patterns, intelligence gathering practices, and incorporation of multiple social groups and individuals. The degree to which demilitarized policing responds to citizen groups, to ethnic minorities, to women, and to political imperatives may vary greatly under “new civilian security” models. Demilitarization is only one dimension of internal security. Civilianized internal security can co-exist with human rights abuses, politicized policing, ethnic exclusion, corruption, and private justice. Partisanship. Demilitarization of internal security has no necessary relationship to whether or not political parties control policing. Historically, the advent of civilian rule has often involved a politicization of policing. Examples include Peron’s Argentina and Guatemala’s Arevalo in mid-century. During transitions to democracy in the 1980s and 1990s, the region’s armed forces commonly argued that civilianization would breed renewed politicization. However, the current wave of democratization seldom generated greater partisanship in policing. Political parties from El Salvador to Guatemala to Argentina gained new influence over the priorities of policing, but mainly through the legislative process. With the exception of Haiti after 1994, overt partisanship (i.e., the use of police repressive, intelligence and investigative powers for partisan gain over other parties) has not been a feature of these current democratizing regimes. Private Security. Civilianization of public security has not meant that private forms of security are absent. Under prior militarized policing, private security elements played an important role in protecting the interests of elites and the state. Private mechanisms for security ranged from landowner-paid military or rural guard units stationed on ranches to "death squads" operated in conjunction with top military officials and financed by economic elites. Although political motives for private policing have ebbed, private security agencies have proliferated throughout Latin America under democracy. The lack of relations with the community translates into lack of faith in public justice system, aggravated by low levels of funding of public security systems. Western-style private security firms include, most visibly, stationary guards for places of business, wealthy residential neighborhoods and financial institutions such as banks and

19


armored car firms. New forms of private security also include neighborhood vigilance groups, private investigator (detective) agencies, and bodyguard services for private citizens. In the two years after Guatemala’s 1996 peace agreement, private security agencies mushroomed from 40 (1997) to 114 (1999), with some 35,000 members.46 The plethora of private security firms raised concerns about the equity between service provided for the poor and the rich, as well as the adequacy of accountability of armed private guards. Human Rights Abuses. One of the main motives and assumptions behind democratizers’ efforts to reduce military influence over internal security is to improve human rights safeguards. Most countries experiencing significant security reforms after wars and interventions witnessed a dramatic improvement in human rights performance. In Argentina, El Salvador, Panama and Guatemala, politically motivated murders, disappearances and torture declined significantly after war transitions.47 In contrast, in many non-war cases (e.g., Chile, Uruguay and Brazil), politically-motivated violations of the right to life and personal integrity had been fewer before democratization and had already declined before political transitions began.48 At the same time, human rights abuses persist in almost all of Latin America. After transitions, police torture and murder, principally against young, violent or organized suspects of common crimes, occurred alongside the routine use of excessive force by police, arbitrary and corrupt justice, and violations of due process.49 In Argentina, police continued to use electric shock and skin-burning, and Guatemalan police sporadically applied torture.50 Despite improvements in public perceptions of some police forces, common citizens, especially poor people, tend to fear and avoid police and the justice system when possible. In addition, politically motivated abuses continue in some cases, most notably Haiti.51 Whereas democratization tends to reduce political violence, even significant demilitarization does not guarantee the eradication of state torture, extralegal killings, and other violations of the law and due process. Demilitarization is a necessary but insufficient condition to ensure that citizen rights and protection are the core of the new civilian security.

46 Carmen Rosa De Leon and Claudinne Ogaldes, “Guatemala: Diagnostico de la Problematica Posconflicto,” in Elvira Cuadra, Violencia Social en Centroamerica (Managua: CRIES, 1999). 47 Americas Watch/CELS, Police Violence in Argentina (New York: Americas Watch, 1991); Nigel Rodley, “Torture and Conditions of Detention in Latin America,” in Juan Mendez, Guillermo O’Donnell, and Paulo Sergio Pinheiro, The (Un)rule of Law and the Underprivileged in Latin America (South Bend, IN: University of Notre Dame, 1999). 48 Lawrence Weschler, A Miracle, A Universe (New York, Pantheon, 1990), Stepan. 49 Paul Chevigny, Edge of the Knife: Police Violence in the Americas (New York: The New Press. 1991); Stanley 1996, Call 2000. 50 Chevigny 1995:187, Rodley 1999:29. 51 Human Rights Watch, World Report 2000 (New York: HRW, 2000).

20


VII.

Generalizing the War Transitions Approach

How well does the “war transitions” argument travel? The experience of war transitions in regions outside Latin America indicates that conflict-ending regime transitions have opened the way for significant security reforms elsewhere. Important police reforms have occurred in conjunction with war transitions in places such as South Africa (1994), Bosnia (1996), Mozambique (1992), Rwanda (1994), Eastern Slavonia (1997), Northern Ireland (1998), Kosovo (1999) and East Timor (1999). These cases represent some of the most far-reaching police reform efforts in the world, and some of the most prominent war transitions of the 1990s. This association lends some support to the proposition that war transitions provide windows of opportunity for sweeping security reforms. However, the content of these security reforms did not correspond to the emphasis on civilianization seen in Latin American cases. In South Africa, where policing under apartheid was already independent of the Defense Ministry, the most significant changes involved racial composition at the officer level and changes in doctrine. In Northern Ireland, religion has been a salient concern in adjusting the composition of post-conflict police forces. And in Rwanda, Bosnia, Eastern Slavonia and Kosovo, reforms focused upon professionalization and ethnic representation rather than demilitarization.52 These experiences support the claim that war transitions provide important opportunities for significant security reforms. However, security reforms outside the heavily militarized states of Latin America entailed the weakening of the political party or ethnic group that controlled the state under authoritarianism. They suggest that the character of internal security reforms in the wake of war will reflect the issues underlying the conflict itself. Conclusion In this article I have argued that a sharp weakening of the armed forces, which most often transpires through war transitions, is a necessary condition for the demilitarization of internal security systems in post-authoritarian Latin America. Small steps toward internal security reform may occur in any new democracy, and slight reversals of demilitarization tend to follow war transitions. However, a weakening of the armed forces is necessary for significant demilitarization of the region’s policing systems. More generally, the end of wars may provide crucial openings for a range of institutional reforms that do not exist in non-war transitions. Resources and attention, therefore, may be better spent in post-war new democracies than in democracies where a strong authoritarian presence persists within the state.

52 On Kosovo, personal interviews with UN Deputy Commissioner Don Grady and Steven Bennett, Director of the Kosovo Police School, Pristina, February 2000. On other cases, see Robert B. Oakley, Michael J. Dziedzic and Eliot M. Goldberg, Policing the New World Disorder: Peace Operations and Public Security (Washington, D.C.: National Defense University, 1998) and special issue of International Peacekeeping 6(4) (Winter 1999).

21


While militarization is central to the region’s internal security systems, security scholarship about Latin America has focused excessively upon the military institution per se, and its propensity to execute coups and limit political activity at the highest levels. The sub-field of “civil-military relations,” drawn heavily from Latin America’s experiences, has experienced something of a crisis in recent years. This analysis suggests that civil-military relations must be placed in the broader context of security studies, adopting an integrated analytical framework including policing issues, intelligence issues, internal security doctrines, state legal systems and agencies – as well as private, community-level, and informal mechanisms of security. More broadly, demilitarization of internal security remains insufficient to bring about fully accountable, non-partisan, participatory, and effective security systems. Demilitarization is a necessary condition for civilian oversight, accountability, transparency, serious human rights improvements, and genuine responsiveness to diverse public constituencies. Reforms adopted after war transitions have generally shown remarkable improvement in these areas, despite ongoing deficiencies. Yet additional research in new democracies is required on the nature of politicized intelligence, police abuses, corruption, and ineffective justice and security systems. Such research will yield more discerning and useful insights into the impact of democratization on the everyday lives of citizens.

22