The roots of coercion and insurgency: exploiting the counterfactual case of Honduras

Abstract

Anecdotal evidence points to a significant relationship between repression and rebellion and yet the quantitative civil war literature ignores state strategies, deeming them endogenous or perfectly correlated with polity type. This article seeks to bring the state back in again and examine the causes of states' strategies and the effects of these strategies on non-violent mobilisation. It finds that, under certain circumstances, a state's response to a peaceful opposition movement depends not on its institutions or capacity; rather, it is a function of the state's control of the national security apparatus, autonomy from its constituents, and resources fungible for reform. Additionally, the article concludes that state policy can play a more significant role in explaining the onset of civil conflict than do structural variables such as per capita income, terrain and population size. Historical analysis of coercion and insurgency in the counterfactual case of Honduras illustrates the plausibility of this argument.

Acknowledgements

The author is particularly grateful to James Fearon, Stephen Stedman and Benjamin Valentino for their valuable guidance and feedback on this project. For additional helpful comments, the author wishes to thank Fitini Christia, Stephen Haber, Terry Karl, Janet Lewis, Alexander Montgomery-Amo, Abbey Steele, Stanford CISAC seminar participants, and several anonymous reviewers. All errors remain my own.

Notes

  1. See Fearon, ‘Counterfactuals and Hypothesis Testing’ on counterfactuals, the comparative method, and hypothesis testing.

  2. Kalyvas, Logic of Violence.

  3. This study builds on the excellent scholarship of Tilly, From Mobilization to Revolution; McAdam et al., Dynamics of Contention. This literature looks at the relationship between repression (threat) and mobilisation, but not at why states repress.

  4. To solve this problem, scholars often conflate repression and regime type with the latter proxying for the former (Collier and Hoeffler, ‘Greed and Grievance’; Fearon and Laitin, ‘Ethnicity, Insurgency’). This generates measurement bias; dictatorship is an imperfect measure of indiscriminate state repression. Additionally, despite being lagged, democracy and political instability variables suffer endogeneity as a country is coded as having a civil war only after its level of violence has crossed a certain threshold (and therefore likely already affected the regime type). The quantitative studies that do examine the effect of repression on rebellion are also plagued by endogeneity of a more extreme type where it is impossible to tease out causality (Poe et al., ‘Repression of the Human Right’; Gurr and Moore, ‘Ethnopolitical Rebellion’). And while some scholars engage in sub-national research designs with detailed data on repression, they examine the role of indiscriminate violence only after the wars have begun and thus offer little insight into the role of repression on insurgency onset (Kalyvas, Logic of Violence; Lyall, ‘Indiscriminate Violence’).

  5. See Sambanis, ‘A Review of Recent Advances’; Hegre and Sambanis, ‘Sensitivity Analysis’; and Blattman and Miguel, ‘Civil War’, for reviews of these studies.

  6. Qualitative studies indicate the importance of state policy, but often do so anecdotally. The works of Theda Skocpol and Timothy Wickham-Crowley are exceptions. Unfortunately, since Skocpol brought the ‘state back in’ in the 1980s, state policies have been forgotten as causes of violence; geography and economics have come to dominate the literature (Buhaug and Gates, ‘Geography of Civil War’; Le Billon, Political Economy of War; Ross, ‘Closer Look at Oil’). This article builds on Skocpol's work to bring the state back in again.

  7. Anocracies are regimes that mix democratic with autocratic features. They include regimes that score between − 5 and 5 on the difference between Policy IV's democracy and autocracy measures (the difference ranges from − 10 to10). See Fearon and Laitin, ‘Ethnicity, Insurgency’. Huntington, Political Order, calls autocracies ‘praetorian regimes’.

  8. Honduras and El Salvador share a common colonial history, similar levels of economic development, human development, democratisation, inequality, population size and rough terrain, but exhibited strong and consistent variation on the outcome of state strategy towards popular mobilisation. These cases follow the guidelines for case selection of Lijphart, ‘The Comparable-Cases Strategy’. Other scholars have engaged in comparisons of the two cases: Booth, ‘Socioeconomic and Political Roots’; Brockett, Land, Power, and Poverty, 2nd ed.; Goodwin, No Other Way Out. Goodwin demonstrates the link between repression and civil war. He does not, however, delineate the causal process by which coercion versus reform influences opposition organisations. He also does not provide evidence that repression, as it was applied, did not occur prior to civil war onset. Moreover, in the case that repression did antedate the insurgency, Goodwin fails to provide insight into why, in the absence of armed opposition, a regime would choose a violent (and ostensibly doomed) policy. Booth emphasises variation in income inequality grievances and repression to explain variation in outcomes. However, as Goodwin, he shows correlations, but does not specify causal mechanisms. He also does not account for the marked variation in state strategies of the Central American countries. Brockett documents and explains variation in agrarian reform across Central America. His approach falls into the traditional grievance approach and cannot account for the fact that grievances are nearly universal, but insurgency is not.

  9. See Hegre and Sambanis, ‘Sensitivity Analysis’, for a discussion of the 88 variables found to correlate with civil war onset.

 10. See Trimberger, Revolution from Above; Stepan, State and Society, for a discussion of autonomy.

 11. Fearon and Laitin, ‘Ethnicity, Insurgency’.

 12. Kalyvas, Logic of Violence.

 13. The data for this section comes from Brockett, Land, Power, and Poverty; Williams, Export Agriculture; Browning, Landscape and Society; Dunkerley, The Long War. The population similarly increased 91 per cent in El Salvador during this period.

 14. The percentage of landless rural families swelled from 31.4 per cent in 1970 to 36 per cent in 1974 (Ruhl, ‘Agrarian Structure’, 48; Brockett Land, Power, and Poverty, 74).

 15. Muller and Seligson, ‘Inequality and Insurgency’, 445–446. The gini coefficient reached 0.82 in El Salvador. See Brockett, ‘Measuring Political Violence’, for alternative measures of land inequality including minifundización (the reduction in average plot size of land) and landless scores.

 16. The Polity Project. http://www.cidcm.umd.edu/polity/. From 1949 to 1954, the case could be made that Honduras was a semi-democracy, but not thereafter (See Bowman et al., ‘Measuring Political Democracy’). The Polity score averaged − 0.5 in El Salvador during this period. The polity score for Honduras rose to 6 in 1982 and dropped to − 6 for El Salvador in 1977 after the states had chosen divergent policies towards their popular mobilisation.

 17. El Salvador experienced four coups during these years.

 18. Ruhl, ‘Agrarian Structure’; Rudolph, Honduras: Country Study, 44.

 19. Lapper and Painter, Honduras: State for Sale.

 20. Rudolph, Honduras: Country Study, 47.

 21. El Tiempo, 2 May 1972, quoted in Morris, Honduras: Caudillo Politics.

 22. In addition to members of the conservative National Party to which President López was allied.

 23. Euraque, Reinterpreting the Banana Republic.

 24. The United States influence over Honduran policy grew after 1980. During the 1980s, the US stationed many troops on Honduran soil and supplemented Honduras' military expenditures with 53–99 million US dollars per year (See Acker, Honduras, 117; Ruhl, ‘Agrarian Structure’, 39–44; Morris, ‘Honduras: The Burden’, 211). Moreover, the US strategy with respect to imminent guerrilla movements changed in 1979 and it began to promote ‘development and democratisation’ over ‘coercion and militarisation’. Accordingly, its military and economic aid became conditional on the presence of democracy and reduction in human rights violations. Thus, the Honduran two-pronged strategy of reform and selective coercion was continued, but came to be dictated by a foreign actor. Many studies have focused on the external-dependence of Honduras as an explanatory variable of its relative stability (Coatsworth, United States; Weeks, ‘Interpretation’). However, Nicaragua and Honduras demonstrated the greatest external dependence and experienced divergent outcomes (Goodwin, No Other Way Out, 153).

 25. Stanley, Protection Racket State,1–2. As has been well documented in the literature, in the 1970s, the Salvadoran state engaged in indiscriminate repression of the popular mobilisation and refused reform on nearly all fronts. ‘Guilt’ was determined not on an individual basis, but on a collective one: guilt by association (Kalyvas, Logic of Violence). The UN Truth Commission of 1993 concluded that violence in the countryside was ‘indiscriminate in the extreme’ in the first years of the 1970s (UN, De la locura, 65). Security forces raided villages chosen almost arbitrarily because of the large presence of FECCAS unions. The scale of these massacres increased from those of 1974–78–La Cayetana (19 dead), Aguilares (50 dead)–to those of 1980 — Sumpul River (600 dead), El Mozote (1000 dead) (See Danner, Massacre at El Mozote; Alas, El Salvador). Moreover, the peaceful vehicles for change were mostly halted; elections in 1974 and 1977 proved fraudulent (Webre, José Napoleón Duarte).

 26. Honduras: The Facts Speak.

 27. Schulz and Schulz, The United States, 159.

 28. Benjamin, Don't Be Afraid, xvi.

 29. These families controlled the agrarian sector and 51 per cent of capital in commerce, 49.9 per cent in construction, 43.7 per cent in services and a majority of the top industrial corporations and private-sector associations with powerful lobbying capabilities. See Dunkerley, The Long War; Colindres, Fundamentos económicos. Also, in contrast to Honduras, most of the 50 foreign companies, which invested in El Salvador in the 1960s, entered into joint ventures with the dominant Salvadoran capitalists (the agrarian elite).

 30. Honduras has no one traditional domestic crop like El Salvador (coffee).

 31. Ruhl, ‘Honduras: Militarism and Democratization’, 38.

 32. Euraque, Reinterpreting the Banana Republic, 38.

 33. The domestic agricultural elite effectively retained control over the economy, and the bourgeois groups consequently ‘lacked significant autonomy from the “oligarchic” families’ (Stanley, Protection Racket, 97) The Salvadoran society thus generally lacked economic leaders willing to promote reform contrary to oligarch interests (Paige, Agrarian Revolution).

 34. For parsimony, this article refers to the ‘Honduran military regime’ or state as a singular entity between 1963 and 1982, despite the existence of several regimes during this period. In comparison to El Salvador, however, the regimes' values on the key variables of interest are similar enough to warrant the regimes being pooled together.

 35. See Article 272 of the 1982 Honduran Constitution. This external orientation was so strong that during the 1954 banana strike crisis, the military ‘appeared more disposed to respond to a potential Guatemalan threat to the country's national security than to quell the north coast strikes’. Soldiers were even dispatched to patrol Guatemalan border posts (MacCameron, Bananas, Labor, and Politics, 26). The military's organisational structure reflected this external focus. The air force, a force charged with defence against a conventional external enemy, accounted for 32 per cent of Honduras' armed forces. In El Salvador, the air force comprised only two per cent of the military. (See Euraque, Reinterpreting the Banana Republic; Rudolph, Honduras: Country Study). It may be argued that the Honduran military's 1963 overthrow of Villeda was executed on behalf of the elite. I propose instead that the military used coups to prevent too far a swing to the right (1971) or left (1963) and thereby maintain national order, not ‘order’ as defined by any one group (See Acker, Honduras).

 36. The Salvadoran military's principal raison d'être, namely the defence of national sovereignty in the face of an invading army, constituted ‘the least of its worries’. Instead the military, as an institution, was oriented and designed not to fight a conventional war against a foreign enemy, but ‘to defeat, instead, internal enemies of the state’ (Williams and Walter, Militarization and Demilitarization, 51). The structure and ideology of its forces reflects this internal security orientation. The National Guard, which was founded to facilitate agricultural commercialisation in the late nineteenth century. through large-scale evictions of peasants and repression of the landless, enjoyed the greatest prestige and authority within the military structure. The military, especially the Guard, continued through the twentieth century to preserve order on private estates, arrest people for vagrancy, and serve the interests of the landed elite. See Americas Watch Committee and The American Civil Liberties Union, Report on Human Rights, xx–xxi. See also Stanley, Protection Racket.

 37. Ruhl, ‘Honduras: Militarism and Democratization’.

 38. Euraque, Reinterpreting the Banana Republic.

 39. Quoted in Euraque, Reinterpreting the Banana Republic.

 40. Towards the end of the civil war, even the Salvadoran military acquired autonomy from the oligarchy.

 41. The regime founded ORDEN in the early 1960s as a rural police force charged with informing on and taking action against any subversive activities. By 1970, ORDEN had a vast network in every village, numbering 10,000 combatants and 100,000 collaborators (Jung, ‘Class Struggle’, 74). ORDEN's command structure enabled social elites to gain control of its units (Stanley, Protection Racket).

 42. ANSENAL served as ‘an employment agency for landlords and industrialists looking for so-called supernumerarios: security personnel who would perform security tasks for companies and farms’ (McClintock, American Connection, 220). Additionally, the elites themselves (with aid from military hard–liners) created armed groups: death squads. These answered directly to civilian elite. They included ARENA (National Republican Alliance), the UGB (White Warriors' Union), and FALANGE (Anti-Communist Armed Forces of Liberation by Wars of Elimination) to name a few.

 43. Mao Zedong describes the guerrilla–civilian relationship: ‘Because guerrilla warfare basically derives from the masses and is supported by them, it can neither exist nor flourish if it separates itself from their sympathizers and cooperation […]. The former [the people] may be likened to water and the latter [the guerrillas] to the fish who inhabit it’ (Mao, On Guerrilla Warfare, 44, 92–93). Thus if the civilian populace constitutes the ‘sea’ in which the combatant ‘fish’ swim, counter-guerrilla warfare is a strategy that seeks to catch the fish by draining the sea (Valentino et al., ‘Draining the Sea’, 384).

 44. Guardian, 23 October 1978, quoted in Dunkerley, Long War, 117.

 45. Ropp, ‘The Honduran Army’, 505–506.

 46. Sagan and Waltz, Spread of Nuclear Weapons, 53

 47. The strongly institutionalised Salvadoran military, in contrast, had a myopic vision and organisational structure, which rendered inflexibility and state repression more likely for three reasons: one, the soldiers had extensive training in the use of repression; repression was institutionalised. Two, past experience had demonstrated repression's efficacy. One colonel expressed to Stanford Professor Terry Karl: ‘In 1932 we killed 30,000 peasants, and they were quiet for 50 years. All we are asking for is another 50 years’ (Karl, ‘Expert Testimony’, 71). Three, the military's organisational structure (brief presidential term and the Defence Minister's exclusive control of officer assignments) enabled hardliners to marginalise reformists (Stanley, Protection Racket). See also the ‘Woerner Report’, a secret Pentagon document produced in 1981 by Brig. Gen. Fred F. Woerner.

 48. Fearon and Laitin, ‘Ethnicity, Insurgency’, 6. See also Olson, Logic of Collective Action; Gurr, Why Men Rebel; and Poe and Tate, ‘Repression of Human Rights’ who argue that economic growth proves destabilising, prompting states to use force to maintain control.

 49. Kalyvas, Logic of Violence.

 50. Penn World Tables.

 51. It gained a 52.4 per cent majority in the free and fair elections of 1982.

 52. The ‘relative strength’ model of Fearon and Laitin, ‘Ethnicity, Insurgency’ is based on the contest models of Gates, ‘Recruitment and Allegiance’, which builds on the conflict success functions of Hirshleifer, ‘Macrotechnology of Conflict’.

 53. Of this arable land, in the 1960s, less than 25 per cent was incorporated into farms in Honduras, while in El Salvador, 75 per cent was.

 54. Durham, Scarcity and Survival, 102; Seligson, ‘Thirty Years of Transformation’.

 55. Lapper and Painter, Honduras.

 56. The banana companies suffered exorbitant losses in the September 1954 hurricane and consequently donated 62, 291 acres of land to the government for colonisation projects. The same occurred following 1974 Hurricane Fiji (MacCameron, Bananas, Labor, and Politics). See also Anderson, Politics in Central America.

 57. United Brands paid a $1.25 million bribe to the Honduran economic minister. It transferred land as reparation (Volk, ‘Honduras’).

 58. Prior to the outbreak of war, Salvadoran migrants occupied 293,000 of the best manzanas in Honduran territory and constituted 20 per cent of the Honduran agriculturally active population (Durham, Scarcity and Survival, 125). See also Torres-Rivas, Interpretación del desarrollo social; Carías and Slutzky, La guerra inútil.

 59. Durham, Scarcity and Survival.

 60. Large landowners assumed this logic, perceiving Salvadorans to be ‘a convenient scapegoat [whose expulsion] offered a means of reducing the threat of land occupations and agrarian reform.’ (Ibid, 125).

 61. No less than 78.7 per cent of the Salvadoran refugees had laboured in an agricultural capacity, 81.8 per cent of which had worked land for themselves. See Capa and Stycos, Margin of Life; Durham, Scarcity and Survival.

 62. Most of the land redistributed in Honduras was publicly owned and thus did not threaten the landed elite. Grants from the two major banana companies accounted for 28 per cent of the lands distributed by INA through 1980. The other lands divided as follows: expropriation or purchase (15 per cent, predominantly from foreign fruit companies), colonisation of unoccupied areas (44 per cent), and recovery of illegally occupied public lands (13 per cent).

 63. Land reform finally initiated in 1980 reflects this; the reform involved expropriations of large properties, much of which was devoted to export agriculture. Phase One of the reform expropriated the 472 properties exceeding 1235 acres. This land accounted for 22 per cent of the nation's farmland: 31 per cent of the land devoted to cotton, 24 per cent of that for sugar, and 14 per cent of coffee land (See US Congress, Status of Land Reform; Simon and Stephens, El Salvador Land Reform; Brockett, Land, Power, and Poverty).

 64. Pearce, Promised Land, 193.

 65. Individuals, it is posited, will not join a rebellion (a public, collective good) if they can instead free ride. Thus to achieve collective action, the armed group must either provide selective incentives (Popkin, Rational Peasant; Olson, Logic of Collective Action) or co-opt networks (Granovetter, ‘Strength of Weak Ties’; Coleman, ‘Social Capital’; Tarrow, Power in Movement; Weinstein, Inside Rebellion).

 66. Petersen, Resistance and Rebellion; Scott, Domination.

 67. Wood, Insurgent Collective Action.

 68. Wood, Insurgent Collective Action, employs James Coleman's term ‘zealots’ to describe the initial insurgent participants who she claims, ‘appear to have been unusually inclined to defiance’.

 69. Weinstein, Inside Rebellion, asserts that nascent insurgencies, hoping to mobilise on a large-scale basis, require resources: either natural resources or networks (ethnic, religious or ideological). This literature, however, cannot provide an explanation for how resource-poor insurgencies (characterised by a lack of natural resource wealth, criminal opportunities and external support) embed themselves in pre-existing networks and why these networks' members opt for armed rather than peaceful mobilisation.

 70. I calculated these figures from Honduras: The Facts Speak, which reports the victim's name and profession, the date of the human rights abuse, and the perpetrator.

 71. Most of the non-nationals were temporarily in Honduras and suspected of having links with the Salvadoran, Nicaraguan or Guatemalan rebels. Some of these were killed by Nicaraguan security forces (FDN and the Contras): see italics in Table I.

 72. Kalyvas, Logic of Violence.

 73. Tactics such as ‘disappearances,’ however, are extremely cruel because families do not know the truth as to what happened to their family member, do not have the remains of their loved ones, and thus, are unable to move forward with their lives and reconcile with the past. Author's interviews with victims of paramilitary and guerrilla violence, Colombia, September 2007–August 2008.

 74. For example, subject to execution, exile and arrest, the Salvadoran National Opposition Union began ‘a series of meeting with labour, peasant and other mass groups in hopes of coming up with a plan to pressure the government from several directions’ (Quoted in Menéndez, Voices from El Salvador, 120) The Rural Workers Federation, meanwhile, came into existence when the Union of Rural Workers (UTC) ‘realized that FECCAS was suffering the same forms of repression as [it]. At the next opportunity [it] formed an alliance to create a united front’ (Pearce, Promised Land, 161). Becoming targets of state violence demonstrated to academics, mostly from middle-class backgrounds, that they would receive the same brutal treatment as the peasantry. Accordingly, the student union (National Association of Salvadoran Educators) united with the peasant (FECCAS and the UTC) and shantytown organisations (UPT) to forge the Bloque Popular Revolucionario 30 de Julio. This trend in organisation intensified over the course of the 1970s. See Almeida, Waves of Protest, for a comprehensive review of this process in El Salvador.

 75. The Unión Nacional Campesina (UNC), Asociación Nacional de Campesinos Hondureños (ANACH), and Federación de Cooperativas de la Reforma Agraria de Honduras (FECORAH) briefly worked together in 1975 for land reform. However, the ‘unity front’ proved superficial with each group working independently towards its own objectives (Morris, Honduras: Caudillo Politics; See also Barry and Preusch, AIFLD in Central America).

 76. Quoted in Benjamin, Don't Be Afraid, 111.

 77. This is the American Federation of Labor's organisation in Honduras, which had links to the US government, CIA and Honduran state. It sought to make ‘Honduran organized labor as politically impotent as possible’ (McClintock, The American Connection, 123).

 78. The US Agency for International Development also operated programmes in El Salvador. For example, it created the Salvadoran Communal Union programme (UCS) which benefited rural small-holders (McClintock, The American Connection, 156). Thus, the US' involvement in organising labour to prevent communism taking root in the unions occurred equally in Central America's other countries as in Honduras.

 79. Williams and Walter, Militarization and Demilitarization, 89.

 80. Acker, Honduras.

 81. Prior to civil war onset, security forces in El Salvador had murdered 51 priests, forced 60 into exile and assaulted 300 clergymen. There were also 19 bombings, 43 shootings, and 30 robberies of churches. In 1976, fliers circulated urging Salvadorans to ‘Be a Patriot! Kill a Priest!’ See US Congress, Religious Persecution; Montgomery ‘The Church’; Menéndez, Voices from El Salvador; Goodwin, No Other Way Out.

 82. Morris, Honduras: Caudillo Politics, 82

 83. Ruhl, ‘Agrarian Structure’, 55

 84. Those who protested human rights abuses in Honduras did not blame the Honduran government; rather they held the United States and the Contras responsible. The US military bases and Contra refugees occupied large amounts of land, displacing peasants and introducing a non-domestic war to Honduran soil. Rather than polarising the country as occurred in El Salvador, these protests had a unifying effect in Honduras, stirring nationalistic sentiments and creating common ground for the Left and Right (Acker, Honduras). This was clear in the protests between 1984 and 1986 in which more than 100,000 Hondurans took to the streets to protest against the Contra war and the American and Nicaraguan military presence (Binns, United States in Honduras; Benjamin, Don't Be Afraid).

 85. Quoted in Benjamin, Don't Be Afraid, 137.

 86. See Tilly, Regimes and Repertoires, on repertoires of mobilization.

 87. Villalobos, ‘Why is the FMLN Fighting?’, 27.

 88. Weinstein, Inside Rebellion. These networks provide bonds of trust, collective action, and within-group policing. They provide ‘activist or investor’ combatants, recruited by appealing to nonmaterial interests and norms, and willing to assume a higher level of risk for less-assured returns.

 89. A prolonged war, in which the armed faction patiently waits for each peaceful group to autonomously realise the imperative of combining the political with a military strategy.

 90. The revolutionaries were actively involved in population mobilisation and utilised it in the latter part of the 1970s.

 91. Becerra, Evolución Histórica de Honduras, estimates the number of armed combatants to have been a few dozen. Schulz and Schulz, United States, estimate 200. Despite this discrepancy, the guerrilla groups ‘had never shown any signs of having mass support, much less the ability to coordinate their activities’ (Schulz and Schulz, United States, 217).

 92. Armstrong and Shenk, El Salvador, 142. See also, UN, De la locura.

 93. Benjamin, Don't Be Afraid, 157–8.

 94. Quoted in Menéndez, Voices from El Salvador, 62

 95. Ibid, 50.

 96. The Honduran military focused its operations on the border areas in order to decrease the flood of Salvadoran and Nicaraguan guerrillas seeking refuge in Honduran territory and to police the areas in dispute after the 1969 Salvadoran/Honduran War (Rudolph, Honduras: Country Study; Binns, United States in Honduras). Additionally, in 1980, the Honduran and Salvadoran armed forces began joint counter-insurgency operations and intelligence. Finally, the US military and Contras also militarised the Honduran/Nicaraguan border, blocking the transport of Sandinista arms.

 97. In contrast, in El Salvador, many civilians ‘kept a gun and some ammunition in their hut or farm’ to aid insurgents. In cities, neighbourhood committees supplemented the insurgents' resources with arms, ammunition, food, water, medicine and logistical support (Menéndez, Voices from El Salvador).

 98. These guerrilla groups were either eliminated or went into exile. Some of those in exile returned in the early 1990s to launch political parties. For example, the FMLH established the Morazanist Liberation Party.

 99. Schulz and Schulz, The United States, 215.

100. It is beyond the scope of this article to extend the analysis to the present. However, for an excellent review of the contemporary landscape in Honduras, see Ruhl, ‘Honduras Unravels’.

101. These scholars focus on relative deprivation, inequality, poverty, and land distribution as driving violence. Representative of this approach are Gurr, Why Men Rebel; Huntington, Political Order; Paige, Agrarian Revolution; Muller and Seligson, ‘Inequality and Insurgency’; and Russett, ‘Inequality and Instability’.

102. These include grievances arising from the denial of political participation and cultural self-determination. See Linz and Stepan, Problems of Democratic Transition who emphasise ethno-nationalist policies as engendering armed movements; and Burton, Conflict: Human Needs Theory, who highlights the denial of basic human needs (identity, recognition, role/participation, and psychological security) as the propeller of violence.

103. Fearon and Laitin, ‘Ethnicity, Insurgency’ and Collier and Hoeffler, ‘Greed and Grievance’, find political dictatorship, income inequality and discriminatory ethnic polices (‘grievances’) uncorrelated with the probability of civil war. They conclude that ‘what is critical is not whether people actually have reason to commit violence, but what enables them to carry it out in particular circumstances […] feasibility is a rare phenomenon’ (Collier quoted in Sherman, ‘Economics of War’, 28). Combined, these circumstances weaken the state's policing capacity and confer rebels access to recruits, a sanctuary, and materiel and financial resources: the requisites of insurgent viability. The circumstances include rough terrain, cross-border sanctuaries, lootable natural resources, poverty and large populations.

104. For the mechanism-view of politics, see Tilly, ‘Mechanisms in Political Science’.

105. Tilly, From Mobilization to Revolution; Francisco, ‘Coercion and Protest’.

106. Wickham-Crowley Guerillas and Revolution in Latin America, 228.

107. Calculated from Honduras: The Facts Speak for Themselves.

108. Poe and Tate, ‘Repression of Human Rights’, 854 coined this term. It refers to the grievances, which result from the denial of freedoms from arbitrary arrest, torture and death. Also emphasising personal integrity grievances are Goodwin, No Other Way Out; Lichbach, ‘Deterrence or Escalation’; Gurr and Moore, ‘Ethnopolitical Rebellion’; and Mason, Caught in the Crossfire.

109. This does not imply correlation between income inequality and conflict. Fearon and Laitin, ‘Ethnicity, Insurgency’ Collier and Hoeffler, ‘Greed and Grievance’, and the cases of El Salvador and Honduras undermine this correlation empirically. However, there may exist an indirect effect of income concentration on war in that the smaller the economic elite, and the more homogenous their interests, the greater the challenge for the state in generating durable multi-class support for reform.