Latin America (Only for the_ideas)

http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 Subverting Democracy : Elite Rule and the Limits to Political Participation in Post-War El Salvador* SONJA WOLF Abstract.After decades of authoritarianism and a twelve-year civil war ended by a negotiated peace agreement in 1992, El Salvador is a markedly different country.

Despite important changes, however, public institutions have remained largely un- responsive, acute social exclusion persists and violent crime has soared. Rather than the possibly inevitable by-products of a post-conflict situation, these and other developments are the consequences of a regression from an incipient electoral democracy to electoral authoritarianism. The elite-controlledAlianza Republicana Nacionalista(ARENA) party had more to gain from the preservation of the status quo than from democratic changes and only accepted a politically inclusive system to restore the oligarchy’s dominant position through electoral politics. Uncommitted to democratic consolidation, successive ARENA administrations maintained an in- stitutional fac¸ade of democracy to reproduce authoritarian governance and defend elite interests.

Keywords :quality of democracy, electoral authoritarianism, elite rule, political parties, El Salvador Introduction After decades of authoritarianism and a twelve-year civil war brought to an end by a negotiated peace agreement in 1992, El Salvador appears in many ways transformed. Prior to this watershed in Salvadorean history, the oligarchy had held economic and political power and sponsored security forces to carry out violent repression of popular resistance against the pre- vailing social and political exclusion. 1The persistence of these conditions prompted the guerrilla forces of theFrente Farabundo Martı´ para la Liberacio´ n Nacional(FMLN) to launch an armed struggle. Concerned to prevent the erosion of their privileges, members of the elite began funding death squads Sonja Wolf is Research Fellow at the Centro de Estudios y Programas Interamericanos, Instituto Tecnolo´ gico Auto´ nomo de Me´xico, Mexico City. Email : [email protected].

* The author would like to thank the editors and the three anonymousJLASreviewers for their constructive comments on earlier drafts of this article.

1Tommie Montgomery,Revolution in El Salvador(Boulder, 1995), pp. 30–2.

J. Lat. Amer. Stud.41, 429–465fCambridge University Press 2009 429 doi:10.1017/S0022216X09990149 Printed in the United Kingdom http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 and founded their own party, theAlianza Republicana Nacionalista(ARENA). 2 Presidential elections in 1989 saw ARENA triumph, and, spurred by a war- shattered economy and a military stalemate, it initiated peace negotiations and oversaw the subsequent transition to electoral democracy. Among the most notable changes the country has since experienced are the FMLN’s conversion into a political party and its electoral participation, judicial re- forms, demilitarisation, the abolition of the old security forces, and the cre- ation of a newPolicı´ a Nacional Civil(PNC). Civil society enjoys greater freedom of expression than before, and declining poverty rates point to apparently greater prosperity.

Despite these appreciable advances El Salvador’s democracy remains weak and exhibits important continuities with past practices. Neoliberal policies intensified economic inequalities, and poverty reduction is chiefly attributable to out-migration and remittances. 3Social exclusion remains pervasive and feeds the country’s gang problem.4Deficient investigative procedures permit high levels of impunity, and PNC members have been implicated in criminal activities, human rights violations, the torture of de- tainees, and death squads. 5The homicide rate has reached such alarming levels that El Salvador now ranks among the most violent nations in Latin America. 6 Contrary to the belief that these are shortcomings in a country moving steadily towards democratic consolidation, I contend that El Salvador has relapsed into electoral authoritarianism (EA). ARENA, benefiting more from the preservation of the status quo than from democratic changes, had accepted a politically inclusive system only to end the war and restore the oligarchy’s dominant position through electoral politics. Uncommitted to democratic strengthening, rightist governments erected an institutional fa- c¸ade of democracy to reproduce authoritarian governance and continue de- fending elite interests. The objective of this article is to explain the emergence of EA in El Salvador and to trace its existence in the institutional and electoral spheres. 7The article begins by locating the EA concept in debates on the quality of democracy, before considering the elite’s 2Ibid., pp. 132–3.3Programa de las Naciones Unidas para el Desarrollo (PNUD),Informe sobre Desarrollo Humano, El Salvador 2003(San Salvador, 2003), p. 35.

4Miguel Cruz and Nelson Portillo,Solidaridad y violencia en las pandillas del gran San Salvador(San Salvador, 1998), pp. 155–62.

5Procuradurı´ a para la Defensa de los Derechos Humanos (PDDH),Violaciones a los derechos humanos por responsabilidad de la PNC(San Salvador, 2007).

6Fundacio´ n de Estudios para la Aplicacio´ n del Derecho (FESPAD),Informe anual sobre justicia penal juvenil, 2005(San Salvador, 2006), p. 21.

7EA also affects civil society, notably through the criminalisation of social protest, but an analysis of these developments exceeds the scope of this article. 430Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 transformation and its role in maintaining authoritarian practices. Sub- sequent sections highlight the politicisation of state agencies and examine the manipulations of recent electoral processes.

Electoral Authoritarianism Post-war administrations have tended to affirm that El Salvador emerged successfully from its conflictive past. External observers, though sensitive to the country’s problems, generally believe that it remains on the path towards full democracy. 8By contrast, I argue that El Salvador is neither democratic nor currently headed towards democratic consolidation. Although the early 1990s witnessed important political openings and institutional improve- ments, democracy building has since stagnated and has yielded a new and ambiguous form of authoritarianism.

The academic literature on the quality of democracy has offered three perspectives on political regimes that couple formal democratic institutions with authoritarian practices. One school of thought describes these cases as defective democracies or democracies ‘ with adjectives ’. 9For example, O’Donnell’s ‘ delegative democracy ’ is devoid of checks and balances, Zakaria’s ‘ illiberal democracy ’ fails to uphold the rule of law, and Wood’s ‘ low-quality democracy ’ lacks fundamental attributes of a liberal democracy. 10 These authors continue to characterise such regimes essentially as democ- racies, even though the extent of authoritarian regression they exhibit challenges this position. A second line of enquiry locates ‘ third-wave democ- racies ’ such as El Salvador at the centre of the political spectrum, considering them neither democratic nor authoritarian. Concepts such as ‘ hybrid re- gimes ’, ‘ semi-authoritarianism ’ and ‘ semi-democracy ’ all capture the notion of genuinely mixed regimes. 11By contrast, the third and most recent way of addressing this conceptual ambiguity is to identify these instances of non- democratic rule as new forms of authoritarianism. Analysts who propose 8Dinorah Azpuru et al. (eds.),Construyendo la democracia en sociedades posconflicto(Guatemala City, 2007) ; John Booth, Thomas Walker and Christine J. Wade,Understanding Central America : Global Forces, Rebellion and Change(Boulder, 2006) ; Frances Hagopian and Scott Mainwaring (eds.),The Third Wave of Democratization in Latin America : Advances and Setbacks (New York, 2005).

9David Collier and Steven Levitsky, ‘ Democracy with Adjectives ’,World Politics, vol. 49 (1997), pp. 430–51.

10Guillermo O’Donnell, ‘ Delegative Democracy ’,Journal of Democracy, vol. 5 (1994), pp. 55–69 ; Elisabeth Wood, ‘ Challenges to Political Democracy in El Salvador ’, in Hagopian and Mainwaring,Third Wave, pp. 179–201 ; Fareed Zakaria, ‘ The Rise of Illiberal Democracy ’,Foreign Affairs, vol. 76, no. 6 (1997), pp. 22–43.

11Larry Diamond, ‘ Thinking about Hybrid Regimes ’,Journal of Democracy, vol. 13 (2002), pp. 21–35 ; Marina Ottaway,Democracy Challenged(Washington DC, 2003) ; Peter Smith, Democracy in Latin America(Oxford, 2005). Subverting Democracy431 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 terms such as ‘ pseudo-democracy ’ and ‘ competitive authoritarianism ’ make explicit the idea that these regimes display ‘ the trappings but not the sub- stance of [democracy] ’. 12 The emergent study of EA builds on this last body of literature by re- cognising these countries to be authoritarian, albeit in ways that differ from traditional authoritarian rule. EA regimes maintain the institutional fac¸ade of democracy, including regular multiparty elections for the highest office, to conceal and reproduce authoritarian governance. 13 Schedler’s work situates these systems between electoral democracies, which lack characteristics such as checks and balances and an impartial judiciary but conduct free and fair elections, and closed or full-fledged autocracies that refrain from holding multiparty elections. 14 EA regimes establish constitutions, parliaments and courts, permit private media and limited but real political space for dissenting civil society groups, stage elections, and tolerate opposition parties.

However, although opposition parties are allowed to win votes and seats, they are denied victory. Elections are not outright fraudulent, but the elec- toral playing field is tilted sufficiently to preserve incumbents’ hold on national power. 15Thus, the shortcomings of the political opposition may not be the primary reason for its setbacks at the polls.

By permitting competitive elections EA regimes are able to retain a democratic appearance, yet, given their possible defeat at the ballot box, rulers resort to authoritarian manipulation to exert control over electoral outcomes. In an international environment in which depriving the oppo- sition of its victory would be perceived as a serious violation of the electoral process and would invite the loss of domestic and international legitimacy, stealing elections is too risky. 16The governing party therefore needs to rely on a combination of electoral manipulation and persuasion to ensure its grip on state power. 17Notwithstanding these manoeuvrings, EA regimes may be overturned and democratic consolidation restart. Strain usually arises when socio-economic crises lead voters to grow so alienated as to seek change through elections. Under these conditions regime persistence turns crucially 12 Andreas Schedler, ‘ The Logic of Electoral Authoritarianism ’, in Andreas Schedler (ed.), Electoral Authoritarianism(Boulder, 2006), pp. 4–5. On ‘ pseudo-democracy ’ and ‘ competi- tive authoritarianism ’, see Larry Diamond et al., ‘ Introduction : What Makes for Democracy ? ’, in Larry Diamond, Juan J. Linz and Seymour Martin Lipset (eds.),Politics in Developing Countries(Boulder, 1995), p. 8 ; Steven Levitsky and Lucan Way, ‘ The Rise of Competitive Authoritarianism ’,Journal of Democracy, vol. 13, no. 2 (2002), pp. 51–65. 13Schedler, ‘ Logic of EA ’, p. 1. 14Ibid.,p.5. 15Larry Diamond,Developing Democracy(Baltimore, 1999), pp. 15–16 ; Schedler, ‘ Logic of EA ’, pp. 3, 12.

16Mark Thompson and Philipp Kuntz, ‘ After Defeat : When Do Rulers Steal Elections ? ’, in Schedler (ed.),Electoral Authoritarianism, pp. 113–28.

17Schedler, ‘ Logic of EA ’, p. 15. 432Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 on the skilfulness of the manipulations. 18 When these interventions are adept, they take place in a measured way. For example, furtive institutional and procedural modifications may occur and media access may be more subtly slanted than before. Where incumbents succeed in prolonging their tenure only empirical analysis can determine whether they did so because of authoritarian controls or opposition weaknesses.

Unlike traditional typologies of authoritarian rule, which emphasise the exercise of power, EA focuses on the access to power. Despite this particular analytical concern this concept does not ignore issues of governance. Indeed, questions about how state power is accessed and how it is exercised are inextricably intertwined and illuminate the extent to which authoritarian rule erodes the freedom and fairness that democratic elections require. 19 It is important to underscore this link so that the characteristics of an EA regime can be fully appreciated. The limitations of El Salvador’s democracy, in- cluding a weak rule of law, persistent social exclusion and high levels of violent crime, might simply be attributed to the state’s incapacity to address these issues effectively. If this argument is accepted, then electoral irregu- larities might be ascribed to institutional incompetence rather than deliberate manipulations. However, although resource constraints contribute to the country’s problems, I contend that an additional and in some ways more important factor is at play here, namely the lack of political will to make public institutions perform adequately. In other words, when the authorities seem unable to respond to some of the population’s most basic needs and expectations, the reason for this may not be that the state is not functioning.

In reality the state may be working just as intended by those who control it, the objective being the protection of minority interests rather than the pro- vision of public services.

States that are considered democratic yet weak may be better con- ceptualised as authoritarian. Elites in new democracies may not be commit- ted to undertaking the necessary changes, especially if they have more to gain from the previous status quo. Instead, they may embark on a form of state building that is oriented toward the preservation of non-democratic rule. The authoritarian state that emerges is one in which public agencies are ‘ over- whelmingly subordinated to partisan political ends ’. 20 The empirical separ- ation between state and regime that is fundamental to a democracy becomes blurred, and the conduct of public officials is essentially aimed at safe- guarding the ruling party’s hold on power. Election officials may seek to 18 William Case, ‘ Manipulative Skills : How Do Rulers Control the Electoral Arena ? ’, in Schedler (ed.),Electoral Authoritarianism, pp. 95–112. 19Schedler, ‘ Logic of EA ’, p. 6.20 Lucan Way, ‘ Authoritarian Failure : How Does State Weakness Strengthen Electoral Competition ? ’, in Schedler (ed.),Electoral Authoritarianism, p. 167. Subverting Democracy433 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 manipulate electoral processes ; the media may be pressured to provide biased news coverage or be complicit in doing so ; and the police may arbi- trarily detain, and sometimes kill, political activists. 21 EA regimes can be further differentiated according to the extent to which rulers are able to exercise control over other actors and the range of issue areas they can influence. Where this control is strong and extensive in scope, leaders can rely on election officials to twist electoral outcomes, on the police to harass activists and opposition candidates, and on the media to censor criticism of the government. They will also enjoy formal or de facto power over large sections of the economy and other spheres. 22 The key characteristic distinguishing EA from electoral democracy is the quality of the electoral process. In order to determine the existence of authoritarian manipulation electoral processes must be empirically scruti- nised and judged on the basis of the available evidence. In particular, re- searchers need to consider both the pervasiveness of the manipulation and its impact on the results. Put differently, it needs to be ascertained, first, whether irregularities may be occasioned by a lack of competence or re- sources rather than fraudulent intent, and whether they are sporadic rather than systematic ; and, second, whether it was these irregularities, rather than errors in campaign message or strategy, that led to the opposition’s defeat. 23 These possibilities aside, EA regimes can differ considerably in the extent to which they resort to repression, fraud or other tools of manipulation to decrease the uncertainty of outcome that is normally associated with elec- tions. Some of these tactics, such as vote buying, voter intimidation or ballot box stuffing, may be so easily discernible that disinterested observers will likely concur in their assessments of authoritarian practices. However, other manipulative efforts are carried out behind the scenes to cover up the traces and preserve the regime’s democratic appearance. In these cases it is infinitely harder to unearth proof of dishonest conduct and reveal the existence of EA.

What criteria can be employed to identify an EA regime, and how might one address the methodological difficulty this task poses ? Electoral obser- vation is often limited to aspects such as the voting process, the vote count, preliminary results and the adjudication of disputes. However, increased scrutiny of election day practices requires incumbents to become more sophisticated in their deceptions, often well before the electoral event. 24 Scholars therefore need to focus their attention on a longer time frame and a wider range of elements. According to international benchmarks, these factors include the neutrality of the electoral authorities and their 21 Ibid., p. 169. 22Ibid., p. 179. 23 Jonathan Hartlyn and Jennifer McCoy, ‘ Observer Paradoxes : How to Assess Electoral Manipulation ’, in Schedler (ed.),Electoral Authoritarianism, pp. 48–51. 24 Ibid., p. 44. 434Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 effectiveness in enforcing the electoral law ; the use of state resources for campaigning purposes ; the process of voter registration and the accuracy and transparency of voter registration lists ; campaign finance/expenditure ; media access/coverage ; the absence of voter intimidation and vote buying (or ab- stention buying) ; and procedures for ballot creation and vote tabulation. 25 Significant failure as regards any one of these factors may be sufficient to qualify an election as undemocratic. Particularly, legal or structural biases, such as unequal media access, unregulated campaign financing and the use of state resources, constitute critical potential sources of authoritarian manipu- lation that may help undermine or eliminate the competitiveness of the opposition and render elections unfair prior to their realisation. 26 Notwith- standing the significance of these conditions, their impact is more difficult to measure than that of disenfranchised voters or stolen ballot boxes.

27Further- more, there exist as yet no guidelines on how to weigh these elements to ascertain the existence of ‘ authoritarian electioneering ’. 28 The development of such a formula exceeds the scope of this article and remains an important area for future research. Subsequent parts of this analysis will instead quali- tatively examine to what extent these criteria apply in the Salvadorean context.

In order to determine whether manipulations have occurred I offer evi- dence that permits judging the overall democratic quality of the electoral processes. The focus is on the elections that have been held since 2004 ; although irregularities arose in all post-war elections, they increased markedly in recent years. 29 Given the very nature of an EA regime, attempts to gather proof of manipulations are complicated by the politicisation of the state agencies and their failure to investigate wrongdoing. The sources I therefore draw on include theProcuradurı´ a para la Defensa de los Derechos Humanos(PDDH), the Organization of American States (OAS) and respected Salvadorean academic institutions and non-governmental organisations.

Origins and Transformation of Elite Rule El Salvador’s economic and political affairs traditionally revolved around the country’s oligarchy. Spanish colonial rule had seen the emergence of an enormously wealthy and powerful landed elite that consolidated its financial clout after independence largely on the basis of the cultivation and export of coffee. Dubbed ‘ the fourteen families ’, after the republican family groups that constituted its nucleus, the oligarchy not only established economic structures that would endure for many years to come but also began enjoying 25 Ibid., pp. 45–6. 26 Ibid., p. 49. 27Ibid., p. 49. 28 Schedler, ‘ Logic of EA ’, p. 9. 29 Universidad Centroamericana (UCA), ‘ Editorial ’,El Salvador 2009 _en la mira, no. 1 (2008), p. 1. Subverting Democracy435 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 greater political weight. 30In the early twentieth century its vast economic and social power allowed the elite to exercise substantial influence over state institutions and public policymaking. Political control was not acquired in the electoral arena but merely transferred from one faction of the ruling class to another. 31At the same time, the need to deal with periodic protest against the deepening social injustice prompted the landowners to employ private ar- mies, which upheld public order and soon acquired a reputation for brut- ality. 32After 1931 the oligarchy ceded the reins of government to the military but retained its control over economic policymaking. As part of this ‘ pro- tection racket ’ the army served as the guardian of the oligarchy and sup- pressed by force any challenge to the status quo. 33 Over the decades the elite continued to dominate commercial agriculture and extended its reach into the banking sector and the nascent industrial sector. The embryonic middle class was no challenge to the oligarchy ; 34 indeed, the latter preserved its cohesion and dominance through kinship ties and business alliances among the dominant family groups. 35 Although this situation has remained essentially unaltered, two developments triggered changes that had important political ramifications. First, the civil war occa- sioned a rapid decline in national output and massive capital flight, eventually ushering in a sectoral shift : the profitability of the agro-export sector de- creased sharply, whereas the commercial and service sectors surged, prompting a shift in elite economic interests. 36 Second, neoliberal reforms, which stimulated a new growth pattern based on non-traditional exports, services and commerce, prompted some busi- nesses to expand their activities within Central America. 37 Since 1990 both the globalised economic power groups (EPGs) that emerged from this pro- cess and the transnational corporations (TNCs) that were drawn to the isthmus by trade liberalisation and privatisations have pushed for even stronger regional economic integration. Over the years the TNCs have 30 James Dunkerley,The Long War(London, 1982), p. 7 ; James Dunkerley,Power in the Isthmus (London, 1988), pp. 344–9. 31Dunkerley,Long War, p. 10 ; James Mahoney,The Legacies of Liberalism(Baltimore, 2001), p. 36. 32Montgomery,Revolution, pp. 30–2. 33 William Stanley,The Protection Racket State(Philadelphia, 1996), pp. 6–7. 34 Dunkerley,Long War, p. 53 ; Mario Lungo,El Salvador in the Eighties(Philadelphia, 1996), pp. 11, 117–19. 35 Dolores Albiac, ‘ Los ricos ma´s ricos de El Salvador ’,Estudios Centroamericanos(ECA), vol.

612 (1999), pp. 841–64 ; Carlos Paniagua, ‘ El bloque empresarial hegemo´ nico salvadoren ˜o’, ECA, vol. 645–646 (2002), pp. 609–93. 36 Elisabeth Wood, ‘ An Insurgent Path to Democracy : Popular Mobilization, Economic Interests, and Regime Transition in South Africa and El Salvador ’,Comparative Political Studies, vol. 34 (2001), pp. 870–1. 37 Segovia, ‘ Integracio´ n real y grupos centroamericanos de poder econo´ mico ’,ECA, vol.

691–692 (2006), pp. 533–4. 436Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 penetrated the sectors formerly controlled by national elites and become economically powerful actors in their own right. 38 Nonetheless, the EPGs have themselves concentrated greater wealth and economic power in their hands, and while this has reinforced the polarisation within the private sec- tor, it has also shaped the ways in which the elite exercises its influence over the state and public policymaking. 39 Why and how this pressure is applied cannot be understood without reference to El Salvador’s immediate pre-war history.

In October 1979, concerned by the mounting revolutionary threat and the Nicaraguan Sandinista victory a few months earlier, a progressive faction of the Salvadorean army staged a coup. The officers installed a military-civilian junta and promised to end repression, create a democratic political system and initiate a series of pro-poor policies, including agrarian reform. These measures would have nationalised the elite-controlled banking system and affected many of the oligarchy’s coffee estates, but the Salvadorean right ultimately deterred these changes. 40 Significantly, the event broke the long- standing military–oligarchy alliance. Its wealth and privileges at stake, the right sought to regain control of the government, first through two unsuc- cessful coup attempts and subsequently by other means. 41One of these was the creation of death squads, built by ex-Major Roberto D’Aubuisson with the financial support of wealthy businessmen, including the founder of El Salvador’s second-largest newspaper,El Diario de Hoy. 42 The second instrument employed by the oligarchy to restore the status quo was the founding of its own political party. Established in 1981, ARENA espoused an anti-communist ideology that perceived any reform efforts as an assault on the existing order. 43 The formation of this openly partisan political vehicle signalled a major transformation of the way in which the elite would henceforth exercise its power. By the time ARENA won the 1989 presidential contest it had incorporated a broader range of businesspeople and relative moderates who were more tolerant of democratic norms than the traditional agro-export elite had been. The Cristiani government (1989–1994) swiftly relaunched the faltering peace process, and the warring sides eventually agreed on a negotiated end to the war.

The gamut of factors that facilitated this settlement include a change in post-Cold War US foreign policy, the erosion of the FMLN’s external support and a softening of its own ideological stance, a military stalemate, international outrage at the Jesuit killings, the impact of the regional peace 38 Ibid., p. 543. 39Ibid., pp. 539, 545. 40 Jeffery Paige,Coffee and Power(Cambridge MA, 1997), p. 34. 41William LeoGrande,Our Own Backyard(Chapel Hill, 1998), p. 58.42 Montgomery,Revolution, pp. 132–3. 43 Leigh Binford,The El Mozote Massacre(Tucson, 1996), p. 28. Subverting Democracy437 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 process, the socio-economic costs of the war and pro-peace public opinion. 44 However, a prime factor was a shift in elite economic interests and an as- sociated preference for ending the war and managing a transition to electoral democracy. The structural transformations of the previous decades had seen the rise of an agro-industrial faction of the elite which moved its capital into the growing service and commercial sectors. Unlike the undiversified agrar- ian faction, this modernising group was no longer reliant on repressive la- bour relations and adopted a politically more moderate outlook. 45After 1979 the right closed ranks behind the army and ARENA, but as the war became seemingly interminable and prevented the oligarchy from reviving the economy and restoring its way of life, the moderate group accepted the need for a negotiated solution : the military’s failure to defeat the FMLN made the army decreasingly useful, and ARENA’s electoral success convinced parts of the elite that a more pluralistic political system was a desirable concession in return for a cessation of the conflict. 46 Thus, the Cristiani government’s acceptance of democracy reflected a purely pragmatic decision aimed at re- viving the elite’s dominant position in the country, rather than a commitment to a more inclusive regime as such.

This is all the more troubling given that the right’s conceptions of democracy were limited and entailed little more than elections and free speech. 47Interviews conducted with elite members at the turn of the decade revealed that although the modernising elements among them had developed support for democratic norms, they remained fiercely opposed to structural reforms and attributed the war to a small group of terrorists who lacked popular backing and would never be a significant electoral force. 48 Thus, in the early 1990s parts of the elite had come to accept the arrival of democracy but not the notion that it might be used to redress the imbalance of social and economic power. Since ARENA, and the groups it represents, had more to gain from the preservation of the status quo than from the widely antici- pated socio-economic and political change, resistance to the transition pro- cess could be expected. The delays and obstructions were numerous and reflected elite opposition to the attainment of institutional strengthening, respect for human rights and socio-economic transformations that the dem- ocratisation of El Salvador required. 49 Throughout the post-war period, 44 James Dunkerley,The Pacification of Central America(London, 1993), pp. 25, 62–3 ; Montgomery,Revolution, pp. 215–16. 45 Paige,Coffee, pp. 88, 92 ; Wood, ‘ Insurgent Path ’, p. 871. 46 Tricia Juhn,Negotiating Peace in El Salvador(Basingstoke, 1998), pp. 9–11, 124–5 ; Lungo, El Salvador, p. 26. 47Paige,Coffee, p. 204. 48 Ibid., pp. 202–17. 49 Centro de Informacio´ n, Documentacio´ n y Apoyo a la Investigacio´ n (CIDAI), ‘ Los acuerdos de paz, diez an ˜os despue´s ’,ECA, vol. 641–642 (2002), pp. 220–5 ; Gino Costa, La Policı´ a Nacional Civil de El Salvador(San Salvador, 1999). 438Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 however, the elite has been wielding influence chiefly through three chan- nels : ARENA, informal lobbying mechanisms and the media.

One of the ways in which this affluent minority, particularly the EPGs, can shape government policies is through its control of ARENA. Since this party is the prime vehicle for elite participation in politics, the government’s economic agenda tended to coincide with that of the EPGs. 50Indeed, many businessmen have held important posts in past administrations, affording them involvement in strategic decisions. In a recent survey among Salvadorean entrepreneurs most respondents stated that this created a con- flict of interest and tended to result in policies that favoured only certain groups. 51 To ensure that the state would remain in the hands of market-friendly political forces, the main EPGs began playing a central role in financing political campaigns, generally supporting only pro-private-sector parties. For example, large business groups in El Salvador, Guatemala and Honduras co- sponsored Antonio Saca’s presidential campaign. 52In practice this has led to an overfunding of parties that the EPGs prefer, because El Salvador lacks legislation regulating campaign finance or expenditure. This omission is problematic partly because money may derive from illicit activities or patrons of electoral campaigns may later be rewarded with government posts or policies. Equally, however, a glaring resource disparity tilts the playing field among the contenders in the election process. In short, elite tactics and a biased legal framework help make electoral campaigns undemocratic and hamper an alternation in power. 53 A second and more alarming way in which the EPGs promote their in- terests is through a variety of informal lobbying mechanisms, including think-tanks, business schools and periodic meetings. 54 The most powerful family groups form a kind of ‘ big boys’ club ’ (locally known aslos grandotes andlos torogozones) and have ‘ latch rights ’ that permit them frequent and direct access to the highest authorities, including the President of the Republic, ministers, lawmakers and judges. 55 The EPGs are not always suc- cessful in advocating some measures and vetoing others, but analysts con- sider their influence generally very effective.

56 According to the previously cited poll, most interviewees concurred that these big businesses determined 50 Segovia, ‘ Integracio´ n real ’, p. 550. 51Instituto Universitario de Opinio´n Pu ´blica (IUDOP), ‘ La transparencia en el Estado sal- vadoren ˜o’,Boletı´ n de prensa, vol. 1 (San Salvador, 2005), p. 8.

52 Segovia, ‘ Integracio´ n real ’, p. 547. 53 On democratic electoral campaigns, see International Institute for Democracy and Electoral Assistance (IDEA),International Electoral Standards(Stockholm, 2002), p. 56. 54 Segovia, ‘ Integracio´ n real ’, pp. 555–6. 55 Ibid., p. 546. 56Ibid., p. 549. Subverting Democracy439 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 economic policy and were the only ones to have benefited from the policies of the last four ARENA governments. 57 The same study concluded that EPG influence over the decision-making process was so profound as to amount to a ‘ capture of the state ’. 58 The elite’s preparedness to use its un- paralleled wealth and power to impose its agenda has thus encouraged a highly undemocratic and exclusive policy process that reduces popular pol- itical participation to elections and fails to meet the needs of all Salvadoreans.

The country’s mass media are a third means for the protection of elite interests. After decades of state censorship and repression that had suffo- cated all critical journalism, the political transition engendered important transformations in the Salvadorean media sector, including greater freedom of expression, the emergence of alternative media, greater journalistic pro- fessionalism and more pluralistic coverage by pro-regime outlets. 59However, even these limited changes suffered a reversal, notably in the mainstream press. The papers’ old political alignments persist and are perhaps most apparent during electoral cycles, when the media become platforms for in- tense power struggles rather than reasoned public debate. Overall, there is concern that the advances of the first post-war decade have halted and have resulted in the renewed silencing of critical voices and a qualitative deterio- ration of news content. Patterns of the past, when the press helped preserve elite dominance, are re-emerging. 60 Today El Salvador appears to be blessed with an extensive range of print and electronic media, including five national newspapers, some 180 radio stations and ten TV channels. This picture of diversity, however, is decep- tive. Each market niche is dominated by a handful of advertising-rich, audi- ence-strong outlets that fail to offer a critique of the dominant political and socio-economic order. WhileLa Prensa Gra´ ficaandEl Diario de Hoyremain the leading newspapers, the television market is controlled by the Telecorporacio´ n Salvadoren˜ a(TCS). Owned by media mogul Boris Esersky, a long-standing supporter of El Salvador’s conservative causes, the TCS is the country’s dominant and most established broadcaster. Its stations cater to the agenda of ARENA and control 90 per cent of the nation’s viewing audience. 61 The contemporary media landscape exhibits three key features all of which shape the production of news content. First, media power remains in 57IUDOP, ‘ Transparencia ’, pp. 7–8. 58Ibid.,p.8. 59 Sonja Wolf, ‘ The Politics of Gang Control ’, unpubl. PhD diss., Aberystwyth University, 2008, pp. 108–16. 60 Interview with Carlos Dada, journalist,El Faro, San Salvador, 24 May 2006 ; interview with Wilfredo Herna´ndez, journalist,El Diario de Hoy, San Salvador, 16 June 2006. 61 Rick Rockwell and Noreene Janus,Media Power in Central America(Urbana/Chicago, 2003), pp. 44–9. 440Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 the hands of three family businesses with ties to the elite and a commitment to the political and economic interests of the ruling party and the private sector. WhereasLa Prensa Gra´ fica’s newsroom enjoys relative independence and encounters limits only when the owners consider their social and econ- omic status to be threatened,El Diario de Hoyexercises internal censorship and the owner himself pens the editorial. 62A similar case is that of the TCS, where stories are filtered and source blacklists are maintained. Esersky’s stations provide free airtime to ARENA members and function as the party’s propaganda arm. 63In what Hughes has termed an authoritarian news model, reporting reflects the overlapping interests of media owners, government and the private sector. 64 The Salvadorean mass media’s political alignments are unmistakable during electoral cycles but do generally yield coverage that is supportive of the government’s agenda. 65 Oligopolistic ownership struc- tures thus contribute to the narrowing of voices that can be heard and to the creation of homogeneous, uncritical and biased coverage. Insofar as such messages favour a small segment of society they ultimately undermine the consolidation of democracy.

Second, economic control mechanisms, notably the use of advertising to reward compliance and punish dissent, constrain journalistic work and serve to moderate media content or strangle alternative news organisations.

Advertising constitutes a critical source of media income, and the threat to an organisation’s profitability, if not survival, that is implied by the loss of ad- vertising may result in uncontroversial reporting to avoid offending import- ant sponsors. 66 The primary patrons are among the economic elite, but the state maintains the largest media budget. 67Ever since ARENA first came to power, official advertising has been allocated to the media according to their editorial line. Traditionally, the government has rewarded news organisations that offer favourable coverage with a greater share of its budget and punished those that fail to do so with the withdrawal of advertising investment. 68 Thus, the media are no longer exposed to attacks or open censorship but are being strangled by advertising boycotts, skilfully used by government and business to stifle dissenting voices.

Third, both the political and economic interests of advertisers or employers and job security can propel media personnel to engage in self-censorship, 62 Interview with Carlos Dada ; interview with He´ctor Vides, Director, Asociacio´ n de Radios y Programas Participativos de El Salvador (ARPAS), San Salvador, 14 July 2005. 63 Rockwell and Janus,Media Power, pp. 45, 135. 64 Sallie Hughes,Newsrooms in Conflict(Pittsburgh, 2006), p. 50. 65 Guillermo Mejı´ a and Rau ´l Gutie´rrez,Medios y democracia(San Salvador, 2005), p. 9. 66 David Croteau and William Hoynes,Media Society(London, 2003), p. 71. 67 Interview with Mauricio Funes, TV journalist, San Salvador, 20 June 2006. 68 Ibid. Subverting Democracy441 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 a practice that works to distort news content and ultimately curbs free ex- pression before it can begin. 69In short, although the overall climate is more conducive to freedom of expression than in the past, the media remain under elite control. The ideological affinity between ARENA, the private sector and the press makes state censorship largely unnecessary and allows the elite both to promote its own agenda and to marginalise voices that criticise the general direction of social and economic policy. 70 These circumstances also have important consequences for electoral politics. Elections are meant to translate the general will of the electorate into representative government. If this is to occur, all parties and candidates need to be able to present their manifestos freely to the population during the electoral campaign. However, this objective is difficult to achieve if media access is unregulated and some outlets act as party mouthpieces rather than providers of rigorous news coverage. Therefore, a country’s legal framework needs to ensure that all political parties and candidates are afforded equitable media access/treatment. This would allow the public to be informed of the political platforms of all contenders in a fair and unbiased manner, which is what the right to vote implies. 71 However, El Salvador’s Electoral Code contains no provisions on either media access during the electoral process or the need for equitable and objective news coverage of candidates, parties, their activities and their proposals. As a result, all post-war elections have included an important source of structural bias that undermines inter- nationally accepted standards for democratic elections.

Setting the Rules of the Game The Peace Accords stipulated a series of institutional reforms aimed at the construction of a democratic society. Many of the initial changes have since been reversed, and the denaturalisation of some state agencies is such that they fail to fulfil their mandate. Today the principal hallmarks of El Salvador’s public institutions are their politicisation and subordination to minority interests. It is beyond the scope of this article to provide more than a brief assessment, limited to a consideration of presidential–legislative re- lations, the judicial sector and the electoral system. However, available re- search suggests that the trends identified below are indicative of a more general, purposeful neglect of the task of equipping the country’s institutions with the capacity and resources they require. 72 In order to appreciate how democracy has been subverted it is necessary to go beyond a reading of the 69 Rockwell and Janus,Media Power,p.7. 70Segovia, ‘ Integracio´ n real ’, p. 547. 71 IDEA,Electoral Standards, pp. 61–3. 72Ricardo Co´ rdova et al., ‘ La contribucio´ n del proceso de paz a la construccio´ n de la de- mocracia en El Salvador ’, in Azpuru et al. (eds.),Construyendo la democracia, pp. 53–290. 442Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 state that focuses on formal institutions. Once the latter are seen as mani- festations of social processes they can be understood as being created and recreated by social relations and the power structures that shape them. 73 El Salvador has a multiparty presidential system, designed to ensure strong horizontal accountability between the various branches of government. 74 However, presidential systems differ significantly in the interaction between the president’s constitutional powers to affect legislation, his partisan powers, and the extent to which the president’s party retains a reliable majority in the legislature. 75 Although ARENA has generally not held a parliamentary ma- jority, it has formed alliances with parties of ideological proximity. The smaller rightist parties, thePartido Demo´ crata Cristiano(PDC) and especially thePartido de Conciliacio´ n Nacional(PCN), have played an important role in sustaining the policy status quo. Furthermore, in Salvadorean political cult- ure lawmakers commonly vote the party line. 76Party discipline affords a level of predictability and facilitates executive–legislative relations. In El Salvador, however, it has served to neutralise the checks and balances in theAsamblea Legislativa(Legislative Assembly) and has permitted ARENA unified control over the government.

To supplement routine procedures the party has repeatedly bought loyalties in return for partisan legislative support. 77 Additional tactics have included the unauthorised alteration of debated bills and the enactment of ‘ midnight laws ’ or bills that are passed without the required procedures, often in plenary sessions that extend until the early morning hours. ARENA has used its dominance in the Legislative Assembly to achieve the enactment of laws that, though controversial if not actually unconstitutional, favour its agenda.

Examples include the 2004 pension reforms, the Central American Free Trade Agreement (CAFTA), and anti-gang legislation. 78Similarly, theCo´ digo Electoral(Electoral Code) underwent almost 200 reforms, most of which modified the legal framework to suit partisan interests, not to democratise the electoral system. 79Overall, the President of the Republic has exaggerated and discretionary powers that are not tempered by adequate checks and balances. 73 Jonathan Barton and Laura Tedesco,The State of Democracy in Latin America : Post-Transitional Conflicts in Argentina and Chile(Oxford, 2004), p. 2. 74UCA, ‘ El desafio´ de la alternancia en el poder ’,ECA, vol. 656 (2003), p. 526.75 Scott Mainwaring and Matthew Soberg Shugart (eds.),Presidentialism and Democracy in Latin America(Cambridge, 1997), pp. 13–14, 51. 76UCA, ‘ La perversio´ n de la institucionalidad : El caso de la Asamblea Legislativa ’,ECA, vol.

597–598 (1998), p. 568.

77 Luis Gonza´lez, ‘ El Salvador : Partidos polı´ ticos y democracia ’, in Rodolfo Cardenal et al.

(eds.),El Salvador : la transicio´ n y sus problemas(San Salvador, 2002), p. 275. 78 Las Dignas et al. (eds.),El Salvador por dentro(San Salvador, 2005), p. 97 ; Ce´sar Villalona,La privatizacio´ n de las pensiones(San Salvador, 2005), p. 18. 79 A´ lvaro Artiga,Las reformas a la legislacio´ n electoral salvadoren˜ a(San Salvador, 2008), p. 1. Subverting Democracy443 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 Executive–legislative relations, and exclusionary lawmaking that often over- rides democratic deliberative mechanisms, speak to an authoritarian decision- making style that pre-empts public influence on policy and accountability.

The fair and independent administration of justice constitutes a key el- ement of the rule of law. Judges are expected not only to apply the law fairly and impartially but also to act as a counterweight to the other branches of government. Given the past malfunctioning of the Salvadorean justice sys- tem and its complicity in human rights violations, the Peace Accords man- dated both a reorganisation of theCorte Suprema de Justicia(CSJ) and an overhaul of the judicial appointment system. An independentConsejo Nacional de la Judicatura(CNJ) was tasked with the oversight and removal of judges, improving the selection of magistrates and ensuring judicial inde- pendence. However, like other parts of the state the judicial sector remains politicised and ineffective. For example, CNJ procedures have been modified to influence the election process of CNJ members and CSJ magistrates. 80 The CSJ itself displays a lack of independence that is evidenced by theCaso Probidad: in 2005 the Court’s Probity Section requested two domestic banks to supply information about the accounts of several former officials, in- cluding ex-President Francisco Flores. The banks, however, refused to comply with the request and asked the CSJ president to strip the Probity Section of these powers. In a plenary session the Court decided to meet this demand, thus preventing the officials in question from being investigated for illicit enrichment. 81 TheFiscalı´ a General de la Repu´ blica(FGR) is tasked with undertaking legal actions and direct criminal investigations in collaboration with the PNC. The appointment of the Attorney-General is made by parliamentary majority vote and generally meets ARENA preferences, with the result that the party has retained control over the FGR. In practice, therefore, this institution regularly fails to investigate members of the political and economic class and enquire into police abuse. 82 The PNC itself arose from the Peace Accords and replaced the militarised security forces, which had maintained order through intimidation and lacked specialised investigative skills, relying in- stead on torture to extract confessions. 83From the very beginning, however, the right sabotaged the restructuring of the public security apparatus partly because it had agreed to it as a way of achieving the disarmament of the FMLN and lacked the commitment to build a democratic institution. 84More 80 FESPAD,Estado de la seguridad pu´ blica y la justicia penal(San Salvador, 2005), pp. 85–90. 81 Ibid., pp. 95–7. 82 Ibid., p. 28 : Luis Gonza´lez, ‘ Acerca de la transicio´ n a la democracia ’,ECA, vol. 573–574 (1996), pp. 631–7. 83 Margaret Popkin,Peace Without Justice(Philadelphia, 2000), pp. 26–7. 84 Costa,PNC, p. 189. 444Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 importantly, the government appears to have been concerned about losing its influence over a key instrument of political control and social order maintenance. 85 As a result, various attempts were made to perpetuate auth- oritarian policing structures and to prevent the emergence of a professional and democratic PNC. The government resisted the dissolution of the old security forces and, most significantly, sought to ensure the dominance of conservative elements within the PNC. For example, a disproportionate number of former military staff were placed in the command structure from where they could impose an authoritarian stamp on the entire police force. 86 The politicisation of the police, particularly through the selection of its most senior personnel, has persisted over the years. The Director-General’s post has been staffed either by a civilian who identified more or less explicitly with ARENA (Mauricio Sandoval and Rodrigo A´ vila) or a member of the old security forces (Ricardo Menesses). Similarly, the command structure remains dominated by former military officers who have converted the police into a vertical and authoritarian force. 87 Since its early days the agency has lacked meaningful investigative capacities and internal control mechanisms and has witnessed growing police abuse.

88Corruption has been a persistent problem within the PNC and has been actively encouraged by some of its directors. 89 Perhaps the most serious problem concerns recurrent human rights vio- lations that are compounded by the absence of effective control mechan- isms. 90 The PDDH has repeatedly condemned practices such as the excessive use of force, the beating of detainees and the use of torture to extract confessions (sometimes in the presence of police chiefs). 91However, recommended corrections have been largely ignored, and the frequency of abuses intensified during the 2000s. 92In addition, the 2004–06 period saw an increasing number of murders by police-run social cleansing groups. Inves- tigations into these extralegal executions were not undertaken, allowing these acts to remain in impunity. 93The PNC’s institutional deterioration indicates perhaps most clearly the official indifference to human rights and democracy, suggesting that successive administrations have sought to maintain a pliant and brutal police.

At the heart of the electoral system lies theTribunal Supremo Electoral(TSE), the maximum authority on electoral matters with complete jurisdictional, 85 Miguel Cruz, ‘ Violencia, inseguridad ciudadana y las maniobras de las e´lites ’, in Lucı´ a Dammert and John Bailey (eds.),Seguridad y reforma policial en las Ame´ ricas(Mexico City, 2005), p. 264. 86 Costa,PNC, pp. 204–11. 87PDDH,Violaciones, p. 11. 88 Edgardo Amaya, ‘ Las polı´ ticas de seguridad en El Salvador ’, in Dammert and Bailey (eds.), Seguridad y reforma policial, p. 222. 89 Anonymous interview, San Salvador, 25 July 2006. 90 Co´ rdova et al., ‘ Contribucio´ n del proceso de paz ’, p. 230. 91 PDDH,Violaciones, pp. 23, 38. 92Ibid., pp. 9, 23. 93Ibid., pp. 60–1. Subverting Democracy445 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 administrative and financial autonomy. It is composed of five magistrates, three of whom are proposed by the top three vote winners in the last presidential elections while two are selected from lists drawn up by the CSJ.

The incorporation of party representatives into an electoral tribunal is likely to undermine its independence and impartiality and, ultimately, to have an impact on the integrity of electoral processes. The PDDH has repeatedly noted that the TSE lacks independence, as is evidenced by its reluctance to sanction violations of the electoral law, and has recommended that its members be citizens without party affiliation. 94However, these observations were not followed up ; if anything, the tribunal’s partisan functioning has worsened over the years.

Two decisions that illustrate this trend concern the illegal appointment of one TSE magistrate and the cancellation process of three political parties. After the 2004 presidential contest the parties entitled to put forward a TSE magistrate were ARENA, the FMLN, and the PDC-CDU (Centro Democra´ tico Unido) coalition. In clear violation of electoral and constitutional norms the Legislative Assembly appointed Julio Moreno Nin ˜os from the PCN, thus ensuring a rightist majority on the tribunal. The nomination was subsequently appealed against, but the CSJ ruled that the appointment had been constitutional. 95 In 2004 the outgoing TSE initiated the cancellation process of the PCN, the PDC and the CDU, all of which had failed to reach the 3 per cent election threshold and had to be dissolved. However, both the PCN and the PDC submitted a constitutional appeal, which the CSJ’s rightist-controlled Constitutional Chamber admitted, thereby infringing the TSE’s jurisdictional independence. 96 The CSJ held that both parties could remain registered, and in January 2005 the TSE decided not to proceed with their cancellation. That same month the Assembly passed a decree declaring the relevant legal pro- visions inapplicable. A subsequent appeal against this decision was rejected by the CSJ. 97 These developments not only highlight the erosion of the country’s democratic institutions and the rule of law but also demonstrate how the state was turned into an appendix of the incumbent regime and the power groups behind it. 94 PDDH, ‘ Informe especial sobre las elecciones presidenciales de 2004 ’, in PDDH, Recopilacio´ n de resoluciones e informes especiales(San Salvador, 2004), pp. 307–15. 95 FESPAD,Otro fraude constitucional : Nombramiento de magistrado Moreno Nin˜ os‘es apegado a la Constitucio´ n’ (San Salvador, 2008) ; PDDH,Informe especial sobre las elecciones legislativas y mu- nicipales de 2006(San Salvador, 2006), pp. 54–5. 96 UCA, ‘ Vuelvan las cosas al estado que se encontraban antes _’,El Salvador 2009 _en la mira, vol. 8 (2008), p. 2. 97 FESPAD,Elecciones separadas : ¿a favor de quie´ n?(San Salvador, 2007). 446Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 Electoral Politics Post-war electoral preferences have largely revolved around ARENA and the FMLN. Smaller parties have struggled to survive in the political system and to constitute a real alternative to the principal contenders. ARENA success- fully defended the presidency from 1989, but from 1997 experienced an important reduction of its vote share at both the municipal and legislative levels. By contrast, the FMLN began with a modest performance in 1994 but steadily increased its number of parliamentary seats and won an important number of mayoralties. In 2000 the leftist party further raised the number of local governments it controlled and overtook ARENA in the Legislative Assembly by two seats. The 2003 electoral outcome enabled ARENA to maintain its numerical strength among the municipalities it governed. The FMLN, however, capitalising on growing disenchantment with the country’s economic direction, not only came to rule the most important and most densely populated urban areas but also surpassed ARENA in the Assembly by four seats. 98 Given these results, it was widely felt that an alternation in power was now more realistic than ever before. 99ARENA thus had to adapt its strategies and, if necessary, resort to skilful manipulations if it was to prevent a leftist victory and with it the start of a political project that might erode elite interests and privileges.

The 2004 Presidential Elections The campaign for the 2004 presidential contest unfolded in ‘ conditions of radical unfairness ’ 10 0 and, in the case of ARENA, centred on two key ele- ments, theMano Duraanti-gang policy and a politics of fear. Although the street gangs were not a new phenomenon in El Salvador no coherent anti- gang policy had ever existed. 101 In July 2003, just eight months before the presidential elections, the Flores administration launched its Mano Dura plan. With it came theLey Antimaras, which permitted the arrest and criminal prosecution of suspected gang members on the basis of their physical ap- pearance alone. At the time officials held the gangs responsible for the majority of homicides in the country, the implication being that tough 98 CIDAI, ‘ Las elecciones municipales y legislativas del 16 de marzo de 2003 ’,ECA, vol.

653–654 (2003), pp. 171–96. 99 UCA, ‘ Juicio polı´ tico sobre las elecciones 2003 ’,ECA, vol. 653–654 (2003), p. 169. 10 0 Andreas Schedler, ‘ The Menu of Manipulation ’,Journal of Democracy, vol. 13, no. 2 (2002), p. 43. 101 Jeannette Aguilar and Lissette Miranda, ‘ Entre la articulacio´ n y la competencia : Las re- spuestas de la sociedad civil organizada a las pandillas en El Salvador ’, in Miguel Cruz (ed.),Maras y pandillas en Centroame´ rica : Las respuestas de la sociedad civil organizada (San Salvador, 2006), p. 58. Subverting Democracy447 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 measures against these groups would reduce the murder rate. 102 This proved ultimately not to be the case, but given the very real concerns about public insecurity the dispatch of joint army/police patrols and PNC sweeps of gang areas were widely applauded. 10 3 Both the timing and its punitive focus suggested that Mano Dura con- stituted a populist penal policy whose primary purpose was not to curb street gang activity but to improve ARENA’s electoral advantage in the run-up to the 2004 presidential elections. 104 In a country where citizens had tired of permanently high levels of violent crime and authoritarian attitudes remain pervasive, the plan helped reverse the recent swing of political sympathies towards the FMLN. Throughout 2003 El Salvador’s weak economic per- formance had dominated public opinion and was considered the nation’s principal problem. However, in an October survey, almost half the popu- lation identified insecurity as the main challenge and for the first time as many as 21 per cent of respondents specifically identified the street gangs as the key problem. 105 Within a few months, government and media discourses on the issue had thus sidelined economic woes, instead successfully depicting gang violence as the top policy priority, which ARENA proposed to tackle with Mano Dura.

Previous studies indicated that Salvadoreans’ views of the main national problems have implications for their political preferences. 106 Unsurprisingly perhaps, the same October poll showed ARENA to lead in voter preferences whereas the FMLN struggled to maintain its recently won popularity. 107 The leftist party, generally seen to offer more attractive responses to the country’s socio-economic concerns, had evidently been marginalised in favour of ARENA, which was perceived to deal most ably with crime. Given widespread popular support for Mano Dura, ARENA’s presidential candidate, Antonio Saca, pledged to continue his predecessor’s policy with theSu´ per Mano Dura plan.

In an unprecedented publicity campaign ARENA presented itself as a cohesive and efficient party and highlighted infrastructure projects and CAFTA negotiations as its achievements. At the same time any blame for the war and its destruction was laid squarely on the FMLN. 108 An alternation of power, it was argued, would jeopardise the country’s development and 102 La Prensa Gra´ fica, ‘ Mano Dura contra mareros ’, 24 July 2003, www.laprensagrafica.com. 103 IUDOP, ‘ Las preferencias polı´ ticas en octubre de 2003 : La mano dura de ARENA ’, ECA, vol. 660 (2003), pp. 1074–5. 104 On penal populism, see Julian Roberts et al.,Penal Populism and Public Opinion(Oxford, 2003), pp. 5, 66. 105 IUDOP, ‘ Preferencias polı´ ticas ’, p. 1075. 10 6 Ibid., p. 1075. 107 Ibid., p. 1072. 108 CIDAI, ‘ Una derecha inconsecuente ’,Proceso, vol. 1087 (2004), www.uca.edu.sv/publica/ proceso/proc1087.html. 448Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 political stability. 10 9 The ARENA campaign was aided in several important ways. State resources were spent on TV spots, and a number of public officials participated in campaign activities even though they were consti- tutionally barred from doing so. 110 The mass media, too, played a key role in promoting ARENA’s cause. The TCS, for example, openly sided with Antonio Saca, whose own broadcasting company is linked to Esersky’s corporation. 111 The leading newspapers, on the other hand, published highly favourable reports about the rightist candidate while speaking of the re- maining contenders in disapproving tones. 112 Both editorials and news pages exhibited a clear bias towards the former and did not provide the analytical, balanced coverage that would have permitted readers to make an informed choice between the different party candidates and platforms. 113 Furthermore, the PDDH received information that private and public sector employees were coerced into voting for ARENA. Maquila workers, for example, were threatened with dismissal, and in some cases individuals’ personal details were recorded or their ID cards withheld. Despite the criminal nature of these incidents and the violation of people’s political rights that they implied, the TSE and the FGR did not initiate investigations. 114 Mostly, however, ARENA’s campaign revolved around a politics of fear that focused on a vilification of the FMLN’s candidate and the alleged implica- tions of an opposition victory. In order to understand why and how ARENA could target its main rival, it is necessary to outline the FMLN’s internal situation at the time and to explore briefly the weaknesses of the leftist party’s manifesto and candidate.

Throughout the post-war period the FMLN has unsuccessfully wrestled with its traditional divisions between theortodoxos, the historical leaders who remain firmly wedded to their revolutionary dreams, and the more moderate renovadores, who argue for an overhaul of the party’s political project. 115 Over the years these disagreements have sparked recurrent conflicts between these twotendenciaswhich have regularly ended with the expulsion of outspoken renovadores. The ortodoxos, meanwhile, have maintained a tight grip over the party’s leadership. 116 These internal squabbles, and an incapacity or 10 9 Miguel Cruz, ‘ Las elecciones presidenciales desde el comportamiento de la opinio´n pu ´b- lica ’,ECA, vol. 665–666 (2004), p. 259. 110 PDDH,Informe preliminar sobre las elecciones presidenciales de 2004(San Salvador, 2004), p. 9 ; PDDH, ‘ Informe especial sobre las elecciones legislativas y municipales de 2004 ’ in PDDH,Recopilacio´ n, p. 320. 111 Laurence Whitehead et al.,Perfil de gobernabilidad de El Salvador(Madrid, 2005), p. 56. 112 Nathaly Guzma´n, ‘ La prensa escrita y la cobertura noticiosa de la campan ˜a electoral ’, ECA, vol. 663–664 (2004), p. 142. 113 PDDH,Informe especial 2004, p. 327. 114 Ibid., p. 331. 115 UCA, ‘ FMLN : difı´ cil aprendizaje democra´tico ’,ECA, vol. 597–598 (1998), pp. 660–1. 116 Rube´n Zamora,La izquierda partidaria salvadoren˜ a(San Salvador, 2003), p. 60. Subverting Democracy449 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 unwillingness to address them appropriately, have not only been eagerly seized on by the gossip-hungry press but also served to erode the FMLN’s credibility. By the time the 2004 campaign commenced the party had en- gaged in little debate about its identity and the policy ideas that it might realistically offer to voters. These limitations could not but impact on the FMLN’s electoral strategies in important ways.

First, the party seemed unable to present a platform that not only sug- gested coherent and viable responses to the challenges facing El Salvador but could also reduce the doubts and uncertainties that a possible FMLN victory has traditionally created in some sectors. Its proposals, often vague, centred on socio-economic reforms aimed at reducing crime and inequality, over- coming high unemployment rates and slow economic growth, reactivating the agricultural sector and – implausibly – reintroducing the country’s orig- inal currency. 117 Overall, the FMLN appeared to assume that the deterior- ating standard of living and popular discontent with ARENA’s policies would automatically translate into an opposition victory. 118 Second, while ARENA had opted for a candidate who was not associated with the party’s past, the FMLN reaffirmed its historical identity with the selection of Schafik Handal. A former guerrilla commander and secretary- general of the now extinct Communist Party, Handal was an emblematic figure of the left and known for his integrity. However, not only did his revolutionary trajectory make him an easy target of the right, but he also failed to attract the sympathies of much of the population. An FMLN- commissioned survey on pre-presidential candidates showed that Handal was not popular, yet the party leadership insisted on his nomination. 119 In the internal selection process he defeated the renovador O´ scar Ortiz. However, subsequent polls on voter preferences indicated that 44.4 per cent of Salvadoreans would never vote for Handal, who trailed in third place after the ARENA and centrist coalition candidates. 12 0 The leftist candidate was evidently a political liability for his party, and the right understood how to shatter Handal’s image as a social justice advocate and to inspire fear in a population that retains vivid memories of the war. The mass media were of critical importance in conducting this dirty campaign and repeatedly warned their audiences about the dangers of a communist 117 CIDAI, ‘ Una (pre)campan ˜a a la manera de ARENA ’,Proceso, vol. 1066 (2003), www.

uca.edu.sv/publica/proceso/proc1066.html ; CIDAI, ‘ El cambio del FMLN ’,Proceso, vol. 1055 (2005), www.uca.edu.sv/publica/proceso/proc1055.html. 118 CIDAI, ‘ Balance polı´ tico ’,Proceso, vol. 1079 (2003), www.uca.edu.sv/publica/proceso/ proc1079.html. 119 CIDAI, ‘ El debate por las candidaturas ’,Proceso, vol. 1050 (2003), www.uca.edu.sv/ publica/proceso/proc1050.html ; William Grigsby, ‘ Did the FMLN Lose or Did Fear Win the Day ? ’,Envı´ o, vol. 273 (2004), www.envio.org.ni/articulo/2183. 12 0 CIDAI, ‘ (Pre)campan ˜a’. 450Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 government. 121 In clear violation of the electoral law, which limits electoral propaganda to the political parties, a number of front organisations also supported this campaign and sponsored dozens of defamatory TV and radio spots and paid ads in the main newspapers. 122 The images portrayed Handal as a hot-tempered and despotic individual who had masterminded kidnap- pings and was determined to turn El Salvador into a Cuban-style dictator- ship. 123 These messages were particularly irresponsible in suggesting that an FMLN victory would jeopardise US–Salvadorean relations, ultimately ter- minating the Temporary Protected Status (TPS) granted to Salvadoreans in the United States and ending remittances, which help more than 20 per cent of Salvadorean families meet their basic needs. 12 4 In short, the electoral campaign not only lacked serious proposals but was also marred by il- legalities, the use of fear and lies, and relentless attacks against the dignity and honour of the main opposition candidate. 125 The FMLN could not, or did not, convincingly counter the right’s argu- ments. Given its own financial limitations and media complicity in ARENA’s cause the FMLN was unable to win the propaganda war. Instead, the party sought to improve Handal’s image by highlighting his integrity and experi- ence and to publicise its platform through house-to-house visits. 12 6 By doing so the FMLN implicitly accepted the idea that what mattered were the can- didates’ personality and character rather than their competence and policy ideas. The challenge, however, was to shift the electoral debate towards a discussion of substantive issues and their possible solutions and to transform people’s perceptions of the country’s reality and the actors capable of changing it. The TSE, for its part, was largely inactive when it came to investigating and sanctioning campaign abuses and electoral violence.

127 For example, the tribunal initially ordered the withdrawal of some of the privately sponsored publicity spots but later remained silent on the issue. 128 121 CIDAI, ‘ La campan ˜a subterra´nea de ARENA ’,Proceso, vol. 1085 (2004), www.uca.edu.sv/ publica/proceso/proc1085.html ; CIDAI, ‘ El cuarto triunfo de ARENA ’,Proceso, vol. 1091 (2004), www.uca.edu.sv/publica/proceso/proc1091.html ; Nathaly Guzma´n, ‘ La prensa escrita ’, pp. 138–44. 122 Grigsby, ‘ Did the FMLN Lose ? ’ ; PDDH,Informe especial 2004, p. 326. 123 CIDAI, ‘ (Pre)campan ˜a ’ ; CIDAI, ‘ Derecha inconsecuente ’ ; CIDAI, ‘ Cuestiones abiertas por la muerte de Handal ’,Proceso, vol. 1179 (2006), www.uca.edu.sv/publica/proceso/ proc1179.html. 12 4 Mario Montesino, ‘ Economı´ a remesera y proceso electoral ’,ECA, vol. 665–666 (2004), p. 295 ; PDDH,Informe especial 2004, p. 327. 12 5 OEA,Informe de la Misio´ n de Observacio´ n Electoral de la OEA en El Salvador, Elecciones pre- sidenciales 2004, OEA/Ser.G CP–doc. 3936/04 (2004), p. 10 ; PDDH,Informe Especial 2004, p. 326. 12 6 CIDAI, ‘ Resultados modestos ’,Proceso, vol. 1085 (2004), www.uca.edu.sv/publica/ proceso/proc1185.html. 127 PDDH,Informe especial 2004, p. 315. 128 Ibid., p. 327. Subverting Democracy451 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 Overall, ARENA and the mass media skilfully exploited Salvadorean political culture : the desire for law and order, conservatism and aspirations for success were all tapped into by offering Mano Dura, raising the spectre of the civil war, and threatening job and remittance losses. 129 Although the electoral process was evidently manipulated, some uncertainty remains as to the impact of the manipulation on the ballot box outcome. However, pre- electoral surveys constitute a useful means to ascertain whether Salvadoreans might have voted differently in the absence of conditions that discriminated against opposition parties. Public opinion polls conducted prior to the 2004 elections indicated that both the anti-gang policy and warnings of remittance losses were critical in swaying voter preferences in favour of ARENA. Mano Dura enjoyed high levels of approval in rural areas and the interior of the country, where ARENA’s support had dwindled in 2003, even though gang violence was primarily a problem in the metropolitan area. 13 0 Furthermore, both remittance recipients and Salvadoreans whose relatives live in the United States were more strongly inclined towards the incumbent party than those who did not share these characteristics. 131 On election day the voter rolls exhibited a series of anomalies that had also occurred in previous years and again introduced a degree of bias into the elections. Names had been changed or added to the lists, and some in- dividuals could not vote either because their names had been deleted from the electoral register or because someone had already voted for them. 132 Such practices distort popular preferences by violating some people’s right to vote and amplifying the voice of others. 133 These cases were few and were found not to have affected the final outcome.134 Nonetheless, international stan- dards were broken since the electoral register was not accurate and permitted fraudulent voting. 135 Similarly, no measures were taken to facilitate voting for on-duty members of the PNC and the armed forces, provisionally detained prisoners or patients and nursing staff in national hospitals. Moreover, in San Miguel some individuals were arrested and barred from the polling station on grounds of their tattoos. 136 Although there exists no legal voting ban for either of these groups, these practical suffrage restrictions amount to an informal disenfranchisement of some citizens.

ARENA won the presidency by a substantial margin, a result that was accepted by contenders and observers alike. The FMLN’s errors in campaign message and strategy undoubtedly contributed to its defeat. Voting day ir- regularities, though serious, were not decisive for the outcome. However, the 12 9 CIDAI, ‘ Las elecciones presidenciales. Un triunfo del bloque hegemo´ nico de derecha ’, ECA, 665–666 (2004), p. 239. 13 0 Cruz, ‘ Elecciones presidenciales ’, p. 256. 131 Ibid., p. 302. 132 PDDH,Informe especial 2004, p. 340. 133 Schedler, ‘ Menu of Manipulations ’, p. 45. 134 PDDH,Informe especial 2004, p. 340. 135 IDEA,Electoral Standards, p. 45. 13 6 PDDH,Informe especial 2004, p. 339. 452Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 electoral process violated a series of conditions that are recognised as es- sential attributes of democratic elections. To be sure, democratic norms are not fully adhered to anywhere in the world. Indeed, the bending of rules and the introduction of tactics aimed at gaining a lead over other parties may be considered the very ‘ stuff of politics ’. Mano Dura, which offered a wholly inappropriate response to a complex and long-ignored problem, adroitly exploited the public’s desire for security to boost ARENA’s popularity, and constituted a key means of electoral persuasion. Yet the latter was combined with authoritarian manipulations, largely in the pre-electoral period, and these developments had important implications for the overall democratic quality of the 2004 elections.

Observations on the existence of EA in the 2004 elections can be divided into three areas : the formation of voter preferences, the expression of voter preferences, and institutional bias. First, political parties and candidates need to be able to disseminate their campaign messages without discrimination. A party may fail to pursue a sophisticated enough strategy, select candidates or adopt platforms that fail to impress voters, or disappoint the electorate for a host of other reasons. However, if opposition forces lack reasonable access to financial resources and the media, their campaign will develop in unequal and unfair conditions. ARENA was favoured, financially and otherwise, through the use of state resources and overtly sympathetic news coverage by the elite-controlled media. Thus, the contenders stood no equal chance of success, and citizens, unable to acquire full and fair knowledge about can- didates’ electoral offers, voted on the basis of ‘ induced preferences ’. 137 Second, voters must be insulated from undue external pressures if they are to make an unconstrained choice. Intimidation, even in subtle forms, can prevent individuals from freely expressing their electoral preferences. 138 Media-based politics of fear and direct coercion exercised at workplaces, casting Handal as committed to violence and threatening job and remittance losses, deceived people and appealed to their survival instincts. In a context of high poverty and unemployment/underemployment such tactics may or may not succeed, but they have no place in democratic politics. Third, state agencies, notably the TSE and the FGR, exhibited a lack of independence during the electoral process. While the voting itself proceeded without major problems, campaign abuses and reports of electoral violence and fraudulent practices were not investigated, let alone sanctioned. By not acting impar- tially and in compliance with the law the institutions in question failed to help create the conditions for a democratic election campaign.

Overall, the above analysis makes it difficult to describe the 2004 electoral process as democratic. These remarks are in no way meant to excuse the very 137 Schedler, ‘ Menu of Manipulations ’, p. 40. 138 Ibid., p. 44.

Subverting Democracy453 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 real weaknesses of the opposition. However, there exists a set of minimum conditions that must be fulfilled in a democracy worthy of the name. If any one of these requirements is grossly violated, elections can be considered undemocratic. 139 Structural bias in media access and treatment and campaign financing and expenditure, combined with institutional submissiveness, precluded free and fair competition between the candidates and helped pressure citizens to vote for the incumbent party. Such conditions, marked by flagrant disrespect for human dignity and the rule of law, made demo- cratic elections an unfulfilled aspiration. 14 0 The 2006 Elections : A President in Campaign Mode In many ways the 2006 electoral process differed little from the previous one.

The FMLN remained focused on socio-economic issues but failed once more to present a coherent programme in this area. Instead, it proposed a series of disconnected measures bordering on the simplistic, such as the reintroduction of thecolo´ nand the derogation of CAFTA. 141 ARENA, by contrast, again made public security a key campaign theme and sought to exploit the FMLN’s revolutionary origins. Schafik Handal’s sudden death in January 2006 initially deprived the campaign of its traditional reference point and swung voter preferences in favour of the FMLN. Apparently alarmed by this shift, ARENA began casting the FMLN as an anti-system party, linking it to street gang activities and anti-CAFTA protests. Destabilising actions of this kind, it was argued, demonstrated that the FMLN had not abandoned violence and would only deter the country’s development. 14 2 Strategies of electoral control, however, commenced before campaigning had even started. For example, in November 2005 the PCN-controlledCorte de Cuentasimposed a fine of some US$ 4 million on a number of opposition mayoral candidates for alleged irregularities in payments to a refuse collec- tion service. The ruling targeted mayors of the FMLN and theFrente Democra´ tico Revolucionario(FDR) who were seen to have strong (re-)election chances and, in case of non-compliance with the sentence, could have not have stood for public office. The mass media soon depicted the matter as proof of corruption among local FMLN governments, but the judgment was repealed following an appeal by that party. To all appearances the case had been politically motivated and sought, in subtle ways, to exclude certain 139 Ibid., p. 38. 14 0 PDDH,Informe especial 2004, p. 318. 141 CIDAI, ‘ La campan ˜a electoral a principios de 2006 ’,Proceso, vol. 1176 (2006), www.

uca.edu.sv/publica/proceso/proc1176.html ; CIDAI, ‘ Consideraciones sobre la plata- forma legislativa del FMLN ’,Proceso, vol. 1181 (2006), www.uca.edu.sv/publica/proceso/ proc1181.html. 14 2 Luis Alvarenga, ‘ El desafı´ o del cambio en el FMLN ’,ECA, vol. 665–666 (2004), pp. 203–5. 454Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 opposition candidates from the electoral arena. 14 3 In a separate attempt to manipulate the voting results, ARENA members gave talks to banking and private-sector employees, warning them of the adverse consequences of an FMLN victory for their job security. Some workers publicised these forms of voter intimidation, but the authorities never investigated these reports. 14 4 Similarly, the issue of vote buying arose when a representative of the National Agrarian Institute promised some communities to resolve their land title problems if they voted for ARENA. 14 5 As in previous years the ruling party invested considerable sums of money in its publicity campaign. In 2004 ARENA had already controlled more than 64 per cent of the political advertising market in TV, radio and newspapers, while the FMLN retained a mere 24.61 per cent and the remaining oppo- sition parties were effectively excluded. In 2006 advertising spending re- mained highly asymmetrical and again saw ARENA’s share rise to over 64 per cent. 14 6 The inequitable access to funding was exacerbated by the use of state resources for overt or veiled electoral propaganda. Neither the TSE nor the Corte de Cuentas, however, moved to investigate these incidents. 147 Such glaring differences in campaign expenditure, allowing one party to saturate the media with electoral propaganda and seeing smaller groups struggle to present their platforms to the population, created unfair competition among the contenders. 14 8 As a result, voters were denied the balanced and pluralistic information they required to make an informed choice.

The dominant media, for their part, provided ample and sympathetic coverage of ARENA’s colourful rallies and CAFTA’s entry into force on 1 March 2006, a date that had been shrewdly chosen to coincide with the final campaigning week. Reporting depicted the trade agreement as a major op- portunity for investment, economic growth and employment creation, im- plicitly associating President Saca with economic development. By contrast, news outlets supplied scant and discrediting coverage of the opposition parties, particularly of the FMLN, which was branded an instigator of viol- ence and disorder.

14 9 Thus, a combination of biased news treatment and the 14 3 CIDAI, ‘ Balance polı´ tico ’,Proceso, vol. 1175 (2005), www.uca.edu.sv/publica/proceso/ proc1175.html. 14 4 Luis Gonza´lez, ex-Director, CIDAI, email, 28 Nov. 2008. 14 5 PDDH,Informe especial sobre las elecciones legislativas y municipales de 2006(San Salvador, 2006), p. 49. 14 6 Fundacio´ n Nacional para el Desarrollo (FUNDE),Observatorio de gastos de campan˜ a pub- licitaria electoral : Primer informe, abril–noviembre 2008(San Salvador, 2008), p. 6. 14 7 PDDH,Informe especial 2006, p. 44. 14 8 Organizacio´ n de Estados Americanos (OEA),Informe de la Misio´ n de Observacio´ n Electoral sobre las elecciones legislativas, municipales y de Parlamento Centroamericano de El Salvador, 2006, OEA/Ser.G CP/doc. 4359/08 (2006), p. 11 ; PDDH,Informe especial 2006, pp. 79–80. 14 9 Xiomara Peraza et al., ‘ La prensa escrita y la cobertura de la campan ˜a electoral de 2006 ’, ECA, vol. 688–689 (2006), p. 280. Subverting Democracy455 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 non-regulation of media access meant that coverage of the political parties and their electoral offers was lopsided and lacked the necessary objectivity. 15 0 Again, the information that circulated in the public domain served to con- strain voter preferences rather than to facilitate analysis.

Despite the nature of the 2006 elections ARENA’s campaign centred not on municipal and legislative proposals but on the image of President Saca.

Notwithstanding a constitutional provision barring public officials from participating in electoral activities, a number of ministers, parliamentarians, mayors and mid-level officials ignored this rule. 151 Police appearances in an ARENA publicity spot were similarly illegal. 152 However, the most startling of these interventions were Antonio Saca’s campaigning efforts in favour of ARENA’s candidates. Given his position as head of state, his involvement in party political activities amounted to a particularly serious violation of elec- toral and constitutional norms and testified to a less than minimal respect for the laws governing his conduct. 15 3 The TSE, however, did not pronounce on the issue, nor did the FGR seek to sanction violations of the Electoral Code. 15 4 At the same time, President Saca’s campaign appearances afforded his party additional press coverage and contributed to the already extremely unequal media access by the political parties. 15 5 Reports about an unusual increase in pre-electoral address changes, raising concerns about the occurrence of transfer votes, prompted the TSE to call for an OAS audit of the electoral roll. Despite the OAS mission’s confir- mation that the necessary criteria of reliability had been met, on election day the register continued to exhibit a series of irregularities. The PDDH was informed of almost 500 anomalies, including fake ID cards, the fraudulent registration or removal of voters, deceased voters on the lists, and some individuals’ inability to vote because someone had already done so in their place. 15 6 Although these manipulations were neither pervasive nor decisive for the outcome, they helped distort popular preferences, subtly disen- franchised some citizens, and jeopardised the transparency and credibility of the elections. 157 The final results were mixed. While ARENA reversed existing voting trends, increasing its number of parliamentary seats and municipal govern- ments, it experienced a significant overall drop in votes and failed to achieve its two main objectives : win the capital and secure an absolute majority in the 15 0 OEA,Informe elecciones 2006, p. 11. 151 CIDAI, ‘ Valoracio´ n crı´ tica de la campan ˜a electoral ’,Proceso, vol. 1184 (2006), www.

uca.edu.sv/publica/proceso/proc1184.html. 152 PDDH,Informe especial 2006, p. 49. 15 3 OEA,Informe elecciones 2006, p. 35 ; PDDH,Informe especial 2006, pp. 45–6, 78. 15 4 PDDH,Informe especial 2006, pp. 46, 62. 155 Ibid., p. 47. 15 6 Ibid., p. 27. 157 Ibid., p. 78. 456Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 Legislative Assembly. The FMLN, on the other hand, gained one seat in the Assembly and made a poor municipal showing, gaining control over only 22.5 per cent of local governments. The party’s loss of four departmental capitals and its defeat in several traditional strongholds were surprising and may have been invited by the decision to abandon some of its coalitions and the earlier expulsion of some mayors who were subsequently re-elected on another party ticket. 15 8 While the results were eventually accepted by all contenders, a major dis- pute occurred over the capital. San Salvador, the country’s economic and political power centre and the city with the highest population density, re- mains for all parties the most coveted mayoralty. Following the TSE’s release of incomplete tallies both the FMLN and ARENA declared a premature victory. The tribunal’s failure to disqualify these statements, and its decision to conduct a count of the impugned votes after the results showed the leftist candidate to have won by a slight margin, raised the spectre of fraud among some FMLN supporters. In an attempt to voice their discontent, a group of protesters marched towards the hotel where the TSE magistrates had gath- ered and clashed violently with police. The incident raised fears about yet greater violence, but calm returned to the capital when the tribunal finally proclaimed the FMLN the winner by a mere 44 votes.

The needlessly delayed announcement, the TSE’s non-transparent hand- ling of the electoral process, and the mistrust this provoked in some sectors of society, were debatable. More important for this discussion, however, is the possibility of fraud that loomed on the horizon and its implications for EA in El Salvador. If the right had contemplated the idea of manipulating the final vote count, that option may have been unrealistic and unnecessary for a number of reasons. First, the prospect of stolen elections encouraged opposition supporters to defend their victory through violence, and the in- cipient crisis may have escalated and become uncontainable. Second, and more crucially, San Salvador may be the country’s political bastion, but local elections are necessarily limited in nature. 159 In other words, although ARENA failed to regain control over a prestigious municipal office, it maintained ultimate political power and had to fear neither the eventual discontinuity of its policies nor the loss of traditional privileges. If the party had indeed considered depriving the left of its victory, both the loss of legitimacy this would have invited and the limited incentives for stealing the elections made acknowledging the FMLN’s win a sensible choice. 15 8 CIDAI, ‘ Las elecciones legislativas y municipales de 2006 : Polarizacio´ n sociopolı´ tica y erosio´ n institucional ’,ECA, vol. 688–689 (2006), pp. 208–9. 15 9 Thompson and Kuntz, ‘ After Defeat ’, p. 114. Subverting Democracy457 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 The FMLN’s difficulties in developing a more compelling campaign may have contributed to its losses at the polls, especially its disappointing municipal performance outside San Salvador. Election day irregularities, occasioning the informal disenfranchisement of some voters, were not per- vasive and did not affect the final results. However, a number of criteria that are part and parcel of democratic elections were not met. Issues of structural bias remained critical sources of manipulation prior to and during the cam- paign and made the elections ‘ not less democratic but undemocratic ’. 16 0 Specifically, the 2006 electoral process exhibited signs of EA in four areas :

the development of opposition choices, the formation of voter preferences, the expression of voter preferences and institutional bias.

First, democratic elections presuppose the free formation of alternatives from which voters can subsequently choose. The Corte de Cuentas ruling against FMLN and FDR mayoral candidates, though subsequently revoked, appeared to constitute an (unsuccessful) attempt to exclude contenders from the electoral arena and narrow the range of opposition choices. 161 Second, citizens can only cast an informed vote if they have previously been able to consult pluralistic and independent sources of information. For this to occur, all political parties and candidates need to enjoy free and fair access to the media. If some participants in the contest are given a much smaller oppor- tunity to publicise their platforms than others, they do not stand an equal chance of success, and the electorate is required to vote on the basis of in- duced preferences. The non-regulation of campaign financing/expenditure, together with the abuse of state resources, allowed ARENA to flood news- papers and broadcasting time with its political advertising, putting the com- paratively underfunded opposition parties at a considerable disadvantage.

Similarly, the Electoral Code’s silence on media access and treatment per- mitted the leading press to release biased and unreliable information. Since the oligarchic media system backs ARENA’s agenda, state interference was unnecessary for the production of coverage that favoured the ruling party and marginalised opposition forces. As a result, voters were offered only highly homogenous and uncritical media content to help them decide be- tween the available alternatives. If anything, this situation was exacerbated by President Saca’s campaign activities, which were among the most blatant violations but went unsanctioned and unfairly afforded ARENA an un- rivalled amount of publicity.

Third, in order to express their preferences freely voters need to be shielded from bribery, actual or threatened coercion, or other forms of undue outside pressure. Vote buying, albeit in isolated cases, and voter intimidation at the workplace and through media reports depicting the FMLN as 16 0 Schedler, ‘ Menu of Manipulations ’, p. 41. 161 Ibid., p. 39. 458Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 unwilling to shed its affinity for violence, undermined a basic element of democratic elections. Fourth, public institutions that were expected to per- form key functions in the electoral process did not act independently of political interests and neglected to fulfil some of their chief duties. While the Corte de Cuentas omitted to launch an enquiry into the use of state re- sources, the FGR failed to initiate criminal investigations into the multiple violations that had been committed. 162 Similarly, the TSE’s refusal to pro- nounce on campaign abuses, electoral violence and officials’ participation in party propaganda did nothing to ensure the overall credibility of the electoral process, and the inefficiency it displayed during the San Salvador vote count triggered tensions that might have ended in the breakdown of law and order.

Without the level of politicisation that permeates the state agencies, these could have displayed greater impartiality and professionalism, thus helping to make the elections fairer for all candidates.

Over the years El Salvador’s electoral processes have experienced ap- preciable advances. However, developments in 2006 differed little from previous ones : certain suffrage restrictions persisted, the authorities’ per- formance continued to be inefficient, political actors broke the law with impunity, the campaign was marred by violence, and political competition remained highly unequal. 163 Given both its complexity and relatively recent emergence the electoral system is unlikely to undergo the necessary trans- formations in the near future. However, observation findings also suggest that there is no political will to promote changes that would reduce, if not eliminate, existing problems and make electoral processes fairer. 164 While opposition weaknesses contributed in some way to its losses, the discussion showed that biased structural conditions and the ruling party’s manipulations helped tilt the outcome in its favour. The 2006 electoral process thus re- vealed the persistence of EA.

2009 : El Salvador at the Crossroads In January and March 2009 El Salvador held municipal/legislative and presidential elections, respectively. The first contest permitted the FMLN to increase its mayoralties to 96 and to obtain 35 parliamentary seats, making it the single largest party in the legislature. By contrast, ARENA took control of only 32 parliamentary seats and 122 local governments, but received an important boost through its win in San Salvador. 165 The most serious elec- tion day irregularities concerned voting by foreigners, vote buying and vote 162 PDDH,Informe especial 2006, p. 62. 163 Ibid., p. 28. 164 Ibid., p. 80. 165 OEA,Informe verbal del Jefe de Misio´ n de Observacio´ n Electoral sobre las elecciones municipales y legislativas de El Salvador, 2009(San Salvador, 2009), p. 4. Subverting Democracy459 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 transfers in favour of the ruling party. 16 6 These practices gave rise to ‘ lega- lised fraud ’ in the capital and other important municipalities and produced distrust and a high degree of conflict in some areas of the country. 167 These anomalies, combined with the political tension surrounding the narrow and highly symbolic presidential race, prompted widespread fears of yet greater fraud and violence during the March elections. 16 8 Given the left’s very real chance of seizing national power, this second event undoubtedly attracted much greater interest both domestically and abroad. While ARENA nominated former police chief Rodrigo A´ vila as its candidate, perhaps the more surprising developments had occurred in the leftist party. Contrary to previous presidential contests, the FMLN put for- ward not an historical leader but a party outsider, the widely-respected news anchor Mauricio Funes. Some saw the journalist’s nomination as a sign of the FMLN’s increasing moderation ; for others the choice of former guerrilla commander Salvador Sa´nchez Cere´n as Funes’ running mate cast doubt on the party’s apparent transformation. Ultimately, this formula suggested a degree of political pragmatism, allowing the FMLN both to attract its core supporters and to broaden its electoral appeal.

Pre-electoral polls had consistently predicted a leftist victory, although such projections meant little : in 2004 the FMLN’s earlier lead was over- turned by a substantial margin, and in 2009 Mauricio Funes’ initially sub- stantial lead dwindled considerably during the final weeks of the campaign.

Meanwhile, the campaign developed patterns reminiscent of previous elec- toral processes. By January 2009 ARENA and its front organisationFuerza Solidariawere dominating 73 per cent of the political advertising market. 169 Fuerza Solidaria also sponsored TV and radio spots and gave talks to public- and private-sector workers, ‘ educating ’ voters about the FMLN’s links with Hugo Cha´vez and the adverse consequences of a leftist win for El Salvador. 17 0 Similarly, government officials and the leading media sought to associate the FMLN with Colombian rebels, street gangs and irregular armed groups supposedly active in the country, insinuating that an FMLN win would invite more violence. 17 1 Despite persistent and serious irregularities on 16 6 Concertacio´ n por la Paz,Institucionalidad electoral colapsada(San Salvador, 2009), pp. 16–19 ; OEA,Informe elecciones 2009, pp. 6–7 167 Concertacio´n,Institucionalidad, p. 19. 16 8 OEA,Informe elecciones 2009,p.5. 169 FUNDE,Observatorio de Gastos de Campan˜ a Publicitaria Electoral : Segundo informe(San Salvador, 2009), p. 6. 17 0 El Faro, ‘ Venezolano en gira de campan ˜a contra el FMLN ’, 16 Jan. 2009, www.

elecciones2009.elfaro.net. 17 1 Contrapunto, ‘ Gobierno y prensa : El complot de los fusiles pla´sticos, ’ 22 Dec. 2008, www.contrapunto.com.sv ;El Diario de Hoy, ‘ Se sospecha de alianza de FARC y FMLN en secuestro ’, 15 Oct. 2008 and ‘ Piden al FMLN no utilizar a mareros ’, 8 Jan. 2009, www.elsalvador.com. 460Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 election day, voting occurred in tranquillity and ended in an historic triumph for the FMLN.

Although a comprehensive assessment of the 2009 electoral process ex- ceeds the scope of this article, the left’s victory warrants some brief reflec- tions on the implications for EA in El Salvador. As indicated earlier, this concept does not imply that the opposition can never win, but that the incumbent party systematically manipulates electoral contests to control their substantive outcomes. Regime continuity or change hinges on a number of variables, notably the opposition’s strength, the skilfulness of the manipula- tions, and voter preferences. Widespread wariness with persistent economic problems and violent crime had put ARENA under considerable strain.

Given its reluctance to implement appropriate public policies the party necessarily needed to resort to authoritarian manipulation to reduce the un- certainty of the 2009 election results. Nonetheless, disenchantment with the country’s direction and the popularity of the FMLN’s candidate occasioned a massive voter turnout, which was crucial in neutralising the manipulative strategies.

Since El Salvador had never seen an alternation in power it was unclear whether ARENA would relinquish office peacefully. Rodrigo A´ vila’s prompt acknowledgement of his defeat can be seen in light of both the general risks associated with stolen elections (particularly the most heavily observed contest in the country’s history) and a series of other factors specific to the current Salvadorean context. First, ARENA’s reaction was shaped by its relative strength in the Legislative Assembly, where the party remains in a position to defend elite interests while blocking unwelcome FMLN measures.

A future leftist government could thus be faced with a defiant opposition and find it difficult to implement its policy agenda. Second, El Salvador’s media sector is unlikely to experience significant transformations in the near future. The mass media can be expected to maintain their antagonism to- wards the FMLN and, together with the country’s top entrepreneurs, may develop into an informal opposition force. Third, the FMLN’s past propo- sals were insufficiently credible, which is partly why voters tended to con- sider ARENA the lesser evil. On this occasion the FMLN’s manifesto was more coherent, but past weaknesses and the party’s inexperience in running a country raise questions about its capacity to govern El Salvador effectively.

The FMLN’s failure to deal ably with the nation’s socio-economic problems would play into the hands of ARENA and pave the way for a return to elite rule.

Mauricio Funes’ presidential victory constituted a significant moment for El Salvador’s democracy, and the left is now in a position to promote the changes that Salvadoreans have long been waiting for. Ultimately, this will depend as much on the right’s maturity as on the FMLN’s own competence Subverting Democracy461 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 and ability to work with all sectors of society. The Funes government can begin to address a host of issues that previous administrations were reluctant to tackle, notably institutional reform, respect for human rights, public se- curity, the rule of law and social justice. Time will tell how effective the FMLN will be in this role. At any rate, the country has broken with years of EA and can restart its much-delayed democratic consolidation.

Conclusion When El Salvador emerged from years of authoritarian rule and armed conflict the ruling ARENA party was required to implement the Peace Accords and start building a democratic society. In the early post-war years the country experienced important changes, including the staging of multi- party elections, institutional reforms, transformations in the media sector and greater space for civil society organising. Yet public institutions re- mained largely unresponsive and acute social exclusion persisted, intensifying violent crime and street gang activities. Rather than the perhaps inevitable by-products of a post-conflict situation, these developments were the consequences of a regression from an incipient electoral democracy to EA.

The latter concept offers a novel way of understanding political regimes that have established the institutional fac¸ade of democracy to conceal and reproduce authoritarian governance. Specifically, it invites us to rethink both the nature of the difficulties El Salvador’s democracy has encountered and the reasons for them. The analytical focus rests on incumbents’ manipulative practices, aimed at controlling electoral outcomes and subordinating state agencies to partisan political ends, and enables us to probe alternative explanations for the opposition’s electoral defeats and the ineffectiveness of public institutions. ARENA, and the elite it represents, had more to gain from the preservation of the status quo than from democratic changes. The party had only accepted the introduction of a politically inclusive system to end war-induced economic collapse and restore the oli- garchy’s dominant position through electoral politics. The right lacked in- centives to foster democratic consolidation and found in EA the means to subvert the country’s rules and institutions and defend elite interests and privileges.

Over the years EA has encroached on various areas. In the institutional sphere El Salvador has seen the formation of an authoritarian state where officials largely act not to fulfil institutional mandates but to protect incum- bent power. Successive administrations lacked the political will to equip state agencies with appropriate resources and instead politicised them. At the same time ARENA developed extensive regime influence over other actors. The elite retains control over the state and economic policymaking ; 462Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 lawmaking in the rightist-dominated Legislative Assembly has tended to fa- vour ARENA’s agenda ; conservative leadership in the FGR has served to suppress investigations of powerful individuals and police abuse and in the PNC permitted the formation of a corrupt and abusive police force ; and the TSE, its independence compromised by its composition, has neglected key duties during electoral processes. Finally, shared political and economic in- terests between the media oligarchy and ARENA lead to coverage that is biased in favour of ARENA and its governments, notably, but not only, during electoral cycles.

In the electoral arena irregularities intensified with the 2004 elections, when the right began to fear the FMLN’s ascent to power. EA regimes differ in how they seek to control electoral outcomes, and ARENA relied on a combination of electoral persuasion and manipulation, notably in 2004 when Mano Dura and the threat of remittance losses helped sway voter pre- ferences against the left. However, criteria for democratic elections were repeatedly breached. Voter list anomalies remained sporadic, and technical problems, though oddly recurrent despite El Salvador’s two decades of electoral experience, did not indicate an intention of fraud. Voter intimi- dation, both at the workplace and through the media, affected the free ex- pression of voter preferences and reflected the poverty of ideas that prevails among the political parties and leads to second-rate campaigns. While the FMLN has often failed to convince, so have the other parties. ARENA in particular has shunned debates and instead relied on the media-fuelled anti- communist rhetoric that has featured in all elections despite its long-lost relevance.

Furthermore, electoral processes have been affected by a serious institu- tional bias, notably the TSE’s tendency not to investigate and sanction campaign abuses, electoral violence and fraudulent practices (no matter how intermittent). Most importantly, electoral campaigns regularly unfolded in unequal and unfair conditions, denying contenders an equal chance of suc- cess and undermining the free formation of voter preferences. Again, al- though opposition weaknesses facilitated incumbent victories to a degree, structural bias, notably the non-regulation of media access/treatment and campaign finance/expenditure, constituted a subtle but significant source of electoral manipulation and afforded ARENA substantial advantages over the remaining parties. Ultimately, elections cannot be democratic unless they permit free and fair political competition.

El Salvador is a markedly different country today, and no one would argue that it continues to see the kind of electoral fraud that characterised the past.

However, manipulations typical of an EA regime reappeared, and traditional structures of economic and political dominance remain in place, subverting public institutions to the extent that these guard minority interests and fail to Subverting Democracy463 http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 design the policies that the nation requires. The notion of EA helps reveal how democracy became an instrument for the perpetuation of elite rule and the inequalities underpinning it. The key actors in this process have shown deep disrespect for democratic norms, the rule of law and the country’s institutions, and Salvadorean society has paid a high price for these viol- ations. In the past the FMLN, like other opposition parties, did little to become politically more competitive and reverse EA any sooner. The FMLN’s 2009 presidential victory has restarted democratic consolidation, but this endeavour will require everyone, both the elite and the public, to contribute to it. Regardless of the political regime that develops, or the party controlling the presidency, Salvadoreans will need to begin to press the government of the day to work for society as a whole, not only for some of its members.

Spanish and Portuguese abstracts Spanish abstract.Tras de´cadas de autoritarismo y de una guerra civil de doce an ˜os que finalizo´ con un acuerdo de paz en 1992, El Salvador es un paı´ s marcadamente diferente. A pesar de cambios importantes, sin embargo, las instituciones pu ´blicas se han mantenido en buena medida paralizadas, persiste una aguda exclusio´ n social, y el crimen con violencia se ha disparado. En vez de ser probables productos colaterales de una situacio´ n de post-conflicto, estos y otros desarrollos son la consecuencia de una regresio´ n desde una democracia electoral incipiente hacia un autoritarismo electoral. El partidoAlianza Republicana Nacionalista(ARENA) controlado por las e´lites tiene ma´s que ganar con la preservacio´ n del status quo que con los cambios democra´ticos, y so´ lo acepto´ un sistema polı´ ticamente inclusivo para restaurar la posicio´ n dominante de la oligarquı´ a a trave´s de la polı´ tica electoral. Sin estar com- prometidas con la consolidacio´ n democra´tica, las sucesivas administraciones de ARENAmantienen una fachada institucional de democracia para reproducir la gobernabilidad autoritaria y defender los intereses de las e´lites.

Spanish keywords :calidad de la democracia, autoritarismo electoral, gobierno de e´lite, partidos polı´ ticos, El Salvador Portuguese abstract.Apo´s de´cadas de autoritarismo e uma guerra civil de doze anos encerrada por um acordo de paz negociado em 1992, El Salvador e´ um paı´ s nota- velmente diferente. A despeito de importantes mudanc¸as, no entanto, em grande medida instituic¸o ˜es pu ´blicas continuam na ˜o correspondendo a`s expectativas ; per- siste a exclusa ˜o social aguda enquanto a quantidade de crimes violentos aumentou drasticamente. Ao contra´rio de serem subprodutos talvez inevita´veis de uma situa- c¸a ˜opo´ s-conflito, estes e outros desenvolvimentos sa ˜o consequeˆ ncias do regresso iniciado quando a incipiente democracia eleitoral transformou-se em autoritarismo eleitoral. O partidoAlianza Republicana Nacionalista(ARENA), controlado pelas elites, tinha mais a ganhar pela preservac¸a ˜odostatus quodo que atrave´s de mudanc¸as 464Sonja Wolf http://journals.cambridge.org Downloaded: 10 Feb 2012 IP address: 131.94.103.244 democra´ticas, somente aceitando, inicialmente, um sistema politicamente inclusivo para depois restaurar a posic¸a ˜o dominante da oligarquia atrave´s de polı´ ticas elei- torais. Sem compromisso com a consolidac¸a ˜o democra´tica, sucessivas adminis- trac¸o ˜es da ARENA mantiveram uma fachada institucional de democracia para reproduzir governanc¸a autorita´ria e defender os interesses das elites.

Portuguese keywords :qualidade da democracia, autoritarismo eleitoral, governo das elites, partidos polı´ ticos, El Salvador.Subverting Democracy465