The Political Economy of Death Squads: Toward a Theory of the Impact of StateSanctioned
Terror
Author(s): T. David Mason and Dale A. Krane
Source: International Studies Quarterly , Jun., 1989, Vol. 33, No. 2 (Jun., 1989), pp. 175-
198
Published by: Wiley on behalf of The International Studies Association
Stable URL: https://www.jstor.org/stable/2600536
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms
The International Studies Association and Wiley are collaborating with JSTOR to digitize,
preserve and extend access to International Studies Quarterly
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
International Studies Quarterly (1989) 33, 175-198
The Political Economy of Death Squads:
Toward a Theory of the Impact of
State-Sanctioned Terror
T. DAVID MASON
Mississippi State University
AND
DALE A. KRANE
University of Nebraska at Omaha
A central theoretical question in the literature on state-sanctioned terror is
whether, and under what conditions, repressive violence deters or stimulates
a shift in popular support away from the regime and toward the
opposition. By combining a rational choice model of the nonelite response
to escalating levels of death squad violence with a structural analysis of the
global and domestic conditions under which the escalation of statesanctioned
terror can be expected, we demonstrate theoretically that
carefully targeted repressive violence may in fact reduce the level of active
popular support for the opposition, at least temporarily. However, as the
level of repressive violence escalates and its application becomes more
indiscriminate, it may in fact produce increases in active support for the
opposition because nonelites can no longer assure themselves of immunity
from repression by simply remaining politically inert. Thus, they turn to the
rebels in search of protection from indiscriminate violence by the state.
Why, then, would a regime, itself composed of supposedly rational individuals,
pursue a policy of escalating repression if such measures are ultimately
counterproductive? We argue that the conditions of structural dependence
characterizing these regimes leave them without the institutional machinery,
economic resources, or political will to address opposition challenges
through more accommodative programs of reform. Thus, escalating
repression is perpetrated not because it has a high probability of success but
because the weakness of the state precludes its resort to less violent
alternatives. The utility of this approach is illustrated with a case study of
reform, repression, and revolution in El Salvador.
Authors' note: An earlier version of this paper was presented at the 1985 Annual Meeting of the American Political
Science Association. The authors are indebted to Mark Lichbach, Bob Maranto, Thomas Rasmussen, Michael
Urban, Tyffany Mason, the editors of this journal, and several anonymous reviewers for their insightful
suggestions an-d criticisms of earlier drafts. Any flaws in the analysis, of course, remain the responsibility of
authors.
? 1989 International Studies Association
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
176 The Political Economy of Death Squads
Although the term "death squad" has entered the lexicon of public discourse on
political violence in the Third World, this phenomenon has only recently become the
subject of systematic analysis in the social sciences. No doubt this omission has been
due in part to the politically charged nature of the subject, exemplified by the heated
debates over proper U.S. policy toward regimes that use death squads and other
forms of repression as instruments of control. From an analytic perspective,
inattention to state-sanctioned terror is symptomatic of the tendency among students
of revolution to treat repressive violence as largely reactive in character, with the
forms and levels of such violence being determined by the scope and intensity of the
insurgent violence, and not vice versa.' Proactive or preemptive violence by the state
seldom has been accorded independent analytical attention.
This tendency arises in part from what Denardo (1984:ch. 1) has termed the
"spark and tinder" logic of most theories of revolution. According to this view,
explaining revolution is a matter of accounting for those "underlying conditions"
that render a society so volatile that one spark can ignite mass discontent into a
revolutionary conflagration. Which precipitating event ignites the tinderbox is of less
concern than how the society became so volatile in the first place, for when the
underlying conditions are favorable it is highly probable that a revolution will occur.
This logic underlies Davies' (1962) "J-curve," Skocpol's (1979) and Johnson's (1982)
structural analyses of revolution, Paige's (1975) theory of agrarian revolt, as well as
most empirical analyses of civil violence, including Eckstein (1965), Bwy (1968), Gurr
(1969), Feierabend and Feierabend (1972), and Hibbs (1973).
Yet, as both Scott (1985:29) and Denardo (1984:16-17) point out, mass revolts
occur far less often than do those syndromes of "underlying conditions" which one
or another theory of revolution has posited as being conducive to revolution. Scott
(1985:xvi) adds that "it is rare for peasants to risk outright confrontation with
authorities over taxes, cropping patterns, development policies, or onerous new
laws." Denardo (1984) argues that mass discontent will crystallize into active support
for revolutionary violence only through the initiative of a dedicated revolutionary
elite guided by a carefully formulated revolutionary strategy. It is our contention
that, along with the organizational catalyst, what is required to convert normally
risk-aversive peasants into revolutionary soldiers is a high level of indiscriminately
targeted repressive violence.
Using the logic of rational choice theory, we present a model of the decision
calculus by which nonelites choose between supporting the regime, supporting the
rebels, or remaining uninvolved. Following Scott's (1985) image of peasant political
behavior, we assume that nonelites caught in the crossfire between regime and rebels
would prefer to remain uninvolved, devoting their efforts instead to the everyday
tasks of securing subsistence. However, as the level of repressive violence escalates
and becomes more indiscriminate, the option of remaining uninvolved is eventually
precluded because nonelites can no longer assure themselves of immunity from
repression by simply remaining politically inert. Under such conditions, they can be
induced to support rebel organizations by the promise of protection from indiscriminate
violence by the state.
This leads us to the question of why a regime, itself composed of supposedly
rational individuals, would pursue a policy of escalating repression if such measures
l Nardin- (1971:64) has criticized theories of civil violence for- treating the violence of opposition groups as a
political problem and that of the state as conflict managemen-t. Gurr- (1985a:2-3) points to a similar tendency
among most empirical theories of civil violence. He notes that, for instance, Tilly's (1978:ch. 4) theory of state
coercion treats it as one among several responses (the others bein-g toleration and facilitation) that a state can adopt
when faced with challenlges from grouLps seeking to improve their position; the implication here (among others) is
that state coercion is reactive, not proactive.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
T. DAVID MASON AND DALE A. KRANE 177
are ultimately counterproductive. We argue that the conditions of structural
dependence characterizing these regimes leaves them without the institutional
machinery, economic resources, or political will to address opposition challenges
through more accommodative programs of reform. Thus, escalating repression is
perpetrated not because it has a high probability of success but because the weakness
of the state precludes its resort to less violent alternatives. By combining a rational
choice model of individual response to death squad violence with a structural
dependence model of regime response to opposition challenges, we hope to explain
the political economy of repression with more rigor and subtlety than either of these
two research traditions-rational choice theory and dependency theory-can by
itself. Certainly these two are seldom treated as readily complementary; therefore
our approach is at least novel. We shall illustrate its utility with a case study of
reform, repression, and revolution in El Salvador.
State-Sanctioned Terror in Comparative Perspective
Over the last decade, a coherent body of scholarship on the sources, characteristics,
and effects of state-sanctioned terror has begun to coalesce. A core of conceptual
works have defined the range of actions included under the rubric of statesanctioned
terror (Sloan, 1984) by distinguishing it from "agitational terrorism" of
opposition groups (Sloan, 1983), specifying characteristics of the state, the opposition,
and the environment (domestic and international) that are conducive to
"governance by terror" (Duvall and Stohl, 1983; Lopez, 1984; Gurr, 1985a, 1985b,
1986), and developing models of the decision calculus by which the state and the
opposition choose between violent and nonviolent tactics in their struggle for
political power (Lichbach, 1984, 1985a, 1985b, 1987). In the empirical realm, several
scholars have explored the methodological issues inherent in the empirical analysis
of state-sanctioned terror (Mitchell et al., 1986; Nicholson, 1986) and have documented
empirical dimensions of state terror in the Third World (e.g., Wolpin, 1986),
including its most extreme manifestations, genocide and politicide (Harff, 1986).
Others have applied elements of these theories in analyses of the patterns, techniques,
and impact of state terrorism in Argentina (Pion-Berlin, 1984), Brazil
(Krane, 1983), Guatemala (Aguilera Peralta, 1980; Premo, 1981), Latin America
generally (Sloan, 1984; Lopez, 1986), the Philippines (Zwick, 1984), and South
Africa (Denemark and Lehman, 1984).
Within this body of research, there remains one rather conspicuous theoretical
void. Most theoretical works have focused on the strategic behavior of rival elites in
the regime and the opposition as they compete for popular support by supplying a
variable mix of benefits and sanctions to nonelites (e.g., Leites and Wolf, 1970,
Jackson et al., 1978; Denardo, 1984; and Lichbach, 1984, 1985a, 1985b, 1987).
Missing in these analyses is any explicit attention to the ways in which differing mixes
of benefits and sanctions affect the political preferences and behavioral choices of
rural and urban nonelites. This would seem to be a crucial question since most of the
victims of political violence are found among these nonelites, and their support and
loyalty are what ultimately determine the outcome of the struggle between the
regime and its opposition. By focusing on the individual caught in the crossfire
between regime and rebels, we can integrate elements of a diverse array of theories
(e.g., resource mobilization, structural dependency, demographic transition) in an
effort to discern just what syndromes of social, economic, and political dynamics
would induce an otherwise politically inert peasant (or urban nonelite) to risk violent
death by taking up arms and joining a movement intent upon the overthrow of the
incumbent regime. For these reasons, our analysis of death squad violence will use
the decision calculus by which nonelites choose between supporting the regime, the
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
178 The Political Economy of Death Squads
opposition, or neither as the common yardstick with which to assess the impact of
varying levels of repressive violence and different syndromes of structural conditions
on the extent of opposition support and level of opposition violence.2
The term "death squad" denotes those military, paramilitary, and irregular units
that engage in violent acts against a population in order to deter them from lending
support to opposition groups. "Death squad violence" is repressive violence intended
to induce compliance through fear. It may be employed reactively or proactively. Its
most critical distinguishing feature is that it is violence sanctioned by the regime,
either explicitly through policy pronouncements or implicitly through lack of effort
to curtail such acts. The extent to which the state sanctions such violence is reflected
in the composition of the death squads and the patterns of their violence. They may
consist of secret police, special counterinsurgency units, or regular units of the
armed forces engaged in various forms of institutionalized violence. They may also
consist of armed units structured along military lines but organizationally autonomous
from state security forces, at least in a formal and legal sense. Among these are
regional and local irregulars, mercenaries, private armies, and vigilante units
engaged in noninstitutionalized violence in the service of local elites who share the
government's desire to preserve the economic and political status quo (Aguilera
Peralta, 1980:110; Lopez, 1984:63-64).
Death squads are most prevalent in societies where an authoritarian alliance
between the military and a powerful economic elite is faced with a serious challenge
to its legitimacy and authority. What the ruling elite fear most is an opposition
movement that can mobilize the rural and urban poor behind efforts to redistribute
political power and economic resources. Death squads and other agents of repression
are instituted as means of suppressing such movements by eliminating their leaders
and terrorizing their sympathizers into withholding support from opposition organizations
(Jackson et al., 1978; Aguilera Peralta, 1980:110; Sloan, 1983; and Gurr,
1985a, 1985b, 1986).
The dominance of the military in the authoritarian state makes repression (as
opposed to accommodation) the more likely response to mass opposition. Repression
is the response for which the requisite institutional machinery (i.e., the military) is
most highly developed. As Seligson (1987:4-5) notes, the military comes to see itself
as the only institution capable of enforcing the political tranquility that is presumed
(by the military and by local and foreign economic elites) to be a prerequisite for
economic development. For many of these regimes, regular infusions of U.S.
military equipment, training, and funding have served as constant reinforcements
for this belief, making violent repression the almost reflexive response to mass
political opposition.
Given this conceptualization of death squad violence, the central theoretical puzzle
becomes that of sorting out the conditions under which such violence deters or,
alternatively, stimulates nonelite support for the opposition. The evidence on this
issue is by no means conclusive, as both theoretical and empirical arguments have
been advanced for the stimulative (Eckstein, 1965; Gurr, 1969:579), deterrent
(Snyder and Tilly, 1972:527; Hibbs, 1973:82-93; Tilly, 1978), and curvilinear (Bwy,
1968; Gurr, 1969, 1970; Feierabend and Feierabend, 1972; Muller, 1985; Seligson
2 The usual assumptions and caveats concern-ing rational choice theory hold here. We simply assuLme that actors
will choose from among available alternatives that course of action (or inaction) which, to the best of their knowledge,
most enhances their own well-being, however conceived. In choosing, they discouLnt each alternative in accordance
with some subjective estimate of the likelihood of that ouLtcome and then choose the one with the greatest expected
utility. In accordance with the revealed preference approach, there is no need to presume that nonelites actually
make the calculations posited in the theory, only that they behave as if they did. Nor is there any presuLmption that
individuals are concerned solely with money or other material commodities. See Popkin (1979:31) on these assuLmptions
as they apply to the analysis of peasant behavior in Third World societies. We are well aware of the many
critiques of rational choice theory. For a recent exchange, see Grafstein (1987) and Hadari (1987).
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
T. DAVID MASON AND DALE A. KRANE 179
and Muller, 1987) effects of state-sanctioned terror on opposition support. In an
effort to sort out these findings, we turn now to a rational choice analysis of the
impact of escalating repressive violence on the political support of individual
nonelites.
Targets of Death Squad Violence
From a rational choice perspective, nonelite response to death squad violence can be
represented as the rational individual's choice between supporting the government,
its opposition, or neither. Rival elites compete for the support and loyalty of citizens
by providing them with benefits (both private and collective in nature) in return for
acts of support, and by threatening them with sanctions (including violence) for
supporting the other elite group (Leites and Wolf, 1970). The political choice facing
a nonelite can be represented as the difference in the individual's expected utility
from supporting, respectively, the government, the opposition, or neither (Mason,
1986b). Under conditions of violent civil conflict, the critical element in this decision
calculus should be the threat of sanctions by the regime and/or the rebels. First,
part of the benefits offered by each of the rival elites consists of nonexcludable public
goods that, as such, are subject to free rider effects (Olson, 1965; Tullock, 1971).
Second, whatever excludable private benefits one elite group may provide will rarely
if ever be sufficient to offset the risk of violent reprisals at the hands of the rival elite.
Thus, neither collective benefits nor private rewards can offset the risk of death
squad violence, and we are left with the threat of violent sanctions as the major
determinant of nonelites' willingness to support the rebels, the regime, or neither.
The impact of death squad violence on the distribution of popular support
between regime and opposition will vary depending upon the targeting strategy
adopted by the regime. The regime may target its violence against (1) leaders of
opposition organizations, (2) the rank and file members and supporters of those
organizations, or (3) randomly selected groups of the mass public who have little or
no demonstrable affiliation with the opposition (see figure 1.)
Opposition Rank and File Politically
Leaders Supporters Inactive
= Withdraw Withdraw (Free Ride?) Remain Inactive
or or
" Go Underground" Provide Covert Support (Free Ride)
o n and Shift to Violent or
a O Tactics (Terrorism?) "Go Underground" and
C> Participate in Violent
Opposition
-: 4a Shift Emphasis Flee the Country Provide Covert
from Terrorism or Support to Opposition
I va to Insurgency Join Insurgent or
Opposition Remain Inert
c: E Insurgency and Continue as Flee the Country
n Conventional Tactics Insurgents or
CK S_ Join Insurgent
Opposition
FIG. 1. Participation in opposition activities.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
180 The Political Economy of Death Squads
Opposition Leaders
The most restrictive targeting strategy is one confined to the leadership of opposition
organizations. Usually, this is the earliest and most common form of death squad
violence to occur, and it appears before the political struggle has escalated to an
intractable civil war.3 Death squad violence directed against opposition leaders will
reduce nonelites' willingness to engage in active support of the opposition. However,
it does so not by inducing the fear that they may become victims themselves but by
reducing their estimate of the opposition's ability to translate individual support into
private or collective benefits. Although the free rider effect (Olson, 1965) would
imply that rational individuals will not contribute to the provision of collective goods
unless coerced or promised private benefits, Frohlich and Oppenheimer (1970),
Frohlich, Oppenheimer, and Young (1971), Chamberlin (1974), and others have
shown that when the public benefits in question are inclusive public goods (i.e., one
person's consumption of them does not reduce the amount available for others),
individuals can be induced to contribute their provision even in the absence of selective
incentives. However, when opposition leaders are eliminated by death squads,
nonelites begin to doubt whether opposition organizations can deliver collective
benefits. Hence, the level of active support for the opposition should diminish, more
because it has become futile rather than because it is dangerous. The level of latent
support for the opposition should rise, however, as the loss of benefits from the
opposition is not likely to be offset by a comparable increase in benefits from the
regime. The level of latent support may be conceived of as those individuals who,
given a choice, would prefer the opposition and its programs to those of the current
regime but who, for fear of the consequences, are unwilling to express those
preferences with overt support behavior.
Rank and File Supporters of the Opposition
The repression of opposition leadership seldom spells an end to political opposition
because the grievances that gave rise to their demands in the first place are not
resolved by repression. Instead, the repression of nonviolent forms of opposition
(e.g., demonstrations, strikes, and electoral politics) will simply induce the new
opposition leaders to adopt violent tactics of their own (Lichbach, 1987). The choice
of tactics at this point will be influenced by a number of concerns. First, the new
leaders will avoid actions which leave them exposed to the current repressive
practices of the state. Second, they must choose tactics of their own that have some
promise of overcoming free rider effects induced by the initial targeting of
opposition leaders. Programs that mobilize a different segment of the population
(e.g., peasants as opposed to factory workers), or that have lower expected costs of
participation (e.g., mass demonstrations as opposed to strikes confined to known
union members) may be adopted.
If the new opposition leaders succeed in restoring a significant level of antigovernment
activity, the regime may respond by expanding the range of death
squad targets to include rank-and-file members of opposition organizations as well as
participants in specific opposition activities. This escalation should have two different
effects on different segments of the targeted population. First, known members of
opposition organizations will not necessarily be intimidated into refraining from
3 The regime also has at its disposal a numhbe of repressive techniques shor-t of death squad violen-ce. It can issue
decrees or other quasi-legal measures to impede the operation of these organizations. It can impr-ison, detain, or
exile the leadership. It can restrict their right to en-gage in certain- political activities. Krane (1983:56-57) provides
evidence of this phen-omenon in Brazil. Aguilera Peralta (1980) and Premo (1981) discuss the use and effects of
these and other repressive actions in Guatemala.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
T. DAVID MASON AND DALE A. KRANE 181
further anti-government activity. They must assume that, as known members of the
opposition, they remain probable targets of repression. Therefore, they have little to
lose by continuing to challenge the regime, though for obvious reasons they can be
expected to resort to more covert and perhaps violent tactics, such as guerrilla
warfare (Lopez, 1984; Sloan, 1984; Lichbach, 1987). By contrast, violence against
occasional participants in opposition activities should intimidate the heretofore
uninvolved into remaining so. To join or support the opposition overtly now is to risk
victimization at the hands of the death squads. Thus, if the regime can confine its
targeting to known members and occasional supporters of opposition organizations,
it may temporarily preempt any expansion in the base of overt support for the
opposition. Latent support for the opposition should continue to increase, however,
in that compliance through fear cannot be equated with positive support. Should the
opposition later demonstrate the ability to secure people against the risk of violent
reprisals, it may be able to mobilize this pool of latent supporters.
The Politically Uninvolved
If the targeting of nonelite participants in opposition activities results in a shift to
insurgent violence on the part of the opposition, the regime may further escalate the
application of repressive violence. Selecting the targets with little if any previous
involvement in anti-government activities is intended to induce the fear among the
uninvolved that, if they do offer aid to the insurgents, they risk reprisals by the death
squads. For such violence to achieve this effect, however, it must be targeted in such a
way that one need not fear coercive sanctions so long as one withholds support from
the insurgents. The difficulty in targeting death squad violence so precisely is the
problem of distinguishing the guerrilla irregular from the politically inert peasant.
As death squads become more arbitrary in their application of violence, the regime
risks the possibility of converting previously indifferent peasants into active
supporters of the insurgents.4 Simply refraining from supporting the insurgents will
no longer exempt one from death squad violence.5
Under these circumstances, the insurgents need only offer an individual the
means to avoid state-sanctioned terror to win his or her support. Insurgents may
provide armed protection or the sanctuary of secure base areas. They may also target
those peasants who serve as informants for the death squads. With their informants
silenced, the death squads may become even more indiscriminate in their targeting,
imposing a greater risk on all nonelites and rendering those who would otherwise
remain uninvolved even more susceptible to rebel appeals that promise sanctuary
from repressive violence. In this manner, indiscriminate repressive violence ceases to
deter and, indeed, stimulates a shift in nonelite support to the insurgents.
Rural Social Change and the Escalation of Death Squad Violence
Our analysis of the effects of different targeting strategies leads us to the question of
why a regime would choose to escalate the level of repressive violence if at some point
this strategy actually stimulates a shift in popular support to the insurgents. We
4 Any shift in peasan-t loyalty from the r-egime to the insurgents is a gradual process, constrained by the risk
aversiveness of individuals living on the margin-s of subsistence. Migdal (1974:226-56) discusses the levels of
peasant support for insurgent organizations an-d the conditions that contribute to incr-easing levels of such support.
5 Frohlich and Oppenheimer (1973, 1974) present a formal treatment of this logic as it applies to the decision to
pay or evade taxes. Mason (1986b, 1987) has applied these theoretical constructs in a formal analysis of, first,
nonelite response to state-sanctioned terror and, second, competing strategies of insurgency and coun-terin-
surgency.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
182 The Political Economy of Death Squads
would argue that the economic and demographic transformations that accompany
dependent modes of development erode the already marginal economic security of
nonelite households, rendering them more vulnerable to subsistence crises and
therefore more susceptible to mobilization appeals of opposition organizations.
These same conditions of structural dependency also constrain the ability of the state
to pursue accommodative alternatives to the escalation of repression.6
Many Third World nations were first integrated into the global trade regime as
producers of agricultural commodities for export to the metropolitan North. The
shift to export agriculture transformed land use and land tenure patterns in these
nations, disrupting traditional patron-client networks and leaving peasants exposed
to fluctuations in local markets for land, labor, and subsistence goods.7 In response to
expanding export markets, landlords act to consolidate land holdings into larger
commercial estates devoted to the production of profitable export crops. As a
consequence, both the numbers and proportions of the rural population existing as
landless wage laborers increase dramatically, disrupting the diffuse ties to local
patrons that traditionally afforded peasants some margin of economic security
(Migdal, 1974:229-31; Mason, 1986a:492-95).
The eroding economic plight of displaced peasants is further compounded by
rapid population growth in the countryside.8 Population growth combines with the
consolidation of land holdings to raise the effective price of tenant plots and to
depress agricultural wages (Harris and Samaraweera, 1984:141-42). Concomitantly,
the market price of subsistence goods rises as the growth of the landless population
raises demands for subsistence goods, while the displacement of food crops by
export crops reduces local supplies (Durham, 1979:30-36).
Faced with these conditions, the displaced peasant has three choices: (1) accept the
status quo and obtain subsistence through whatever means are immediately available,
(2) attempt through collective action to force an increase in wages or access to land, or
(3) migrate to other regions where unused lands, better wages, or occupational
alternatives to agriculture are available (Scott, 1972:93-94). However, with the value
of land and the size of the landless population both increasing, there is not likely to be
much unused land available. The concentration of export production in one or two
major crops ensures that the terms of trade between landlords and peasants in
different regions of the nation will become uniform. Those who profit from export
agriculture do not often invest their profits locally in new job-producing enterprises.
Instead, they prefer increases in consumption or safer, more lucrative investments in
the metropolitan North. Even if landowners were to invest locally, it is doubtful that
the rate ofjob creation would be sufficient to absorb the expansion in the labor force
that is typical of dependent Third World societies. Gendell (1986:64-65) estimates
that an average annual growth in GDP of 8 percent would be required to absorb the
growth in Latin America's labor force that is projected to occur between 1975 and
2000. Thus, the very conditions that render the peasant's immediate economic plight
untenable also serve to foreclose the major "escape valves." Large numbers of peasants
are compelled to migrate to urban barrios where they sustain a marginal existence in
the informal and service sectors of the economy (Walton, 1984).
6 Among the major works on dependency theory are Frank (1972), Cardoso (1973, 1977), O'Donnell (1973),
Cardoso and Faletto (1979).
7A number of works have analyzed these changes as the erosion of traditional patron-client networks. Among
them are Migdal (1974), Paige (1975), Scott (1975), Kerkvliet (1977), Scott and Kerkvliet (1977), Popkin (1979),
Mason (1986a).
8 Migdal (1974:92-103), Scott and Kerkvliet (1977), Durham (1979), Harris and Samaraweera (1984), and
Mason (1986c) have discussed the impact of popLlation growth on the economic plight of peasants.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
T. DAVID MASON AND DALE A. KRANE 183
In the absence of state action to remedy their economic distress, nonelites turn to
alternative institutions in search of some form of relief. Among these are grass roots
organizations of various sorts, including traditional communal organizations, peasant
cooperatives, and (in the case of Latin America) the Catholic Church.9 Through
these mechanisms, the poor begin organizing for the purpose of providing themselves
with relief from the conditions of poverty and impotence. Participation is
concentrated initially in less explicitly political activities such as setting up health and
hygiene programs, child care cooperatives, literacy programs, clinics, and food
distribution networks. Such programs address the immediate economic needs of the
displaced and are less risky than more explicitly political activities (Chaffee, 1979).1
As nonelites experience some measure of success in cooperating for economic
self-help, grass-roots organizations begin expanding their activities to include
mobilization for policy change and reform. Popular organizations serve as links
between previously disenfranchised nonelites and more traditional opposition
organizations such as labor unions and political parties. Thus the uninvolved are
mobilized for political action not directly by opposition parties themselves but
indirectly through the parties' ties to emerging grass-roots organizations.
Faced with an expansion in the opposition's base of mass support, the regime has
two broad strategies to choose from in its effort to preempt a successful challenge to
its authority. It can increase its reliance on coercion in an effort to intimidate
nonelites into withholding active support from the opposition; or it can attempt
through reforms to accommodate at least minimally the demands of economically
marginalized nonelites. The particular mix of coercion and accommodation that the
regime adopts will depend upon the resources at its disposal and the difference in
the expected payoffs (in terms of political stability) from investing marginal
resources in coercion versus accommodation (Lichbach, 1984).
However, the character of the authoritarian alliance between the military, the
economic oligarchy, and their external supporters makes such regimes generally
resistant to reforms that would redistribute wealth or political power at the expense
of this alliance. " The highly skewed distribution of land, wealth, and income and the
high rates of unemployment and underemployment typical of dependent societies
limit the tax base from which the state can extract the resources needed to finance
reforms. If the state attempts to shift the costs of reform onto the economic elite,
they in turn can threaten to liquidate their assets and emigrate, or they can activate
their ties with the military (through a coup if necessary) to preempt redistributive
reforms. External financial assistance, whether from private banks, foreign governments,
or international agencies such as the IMF, also reinforces the tendency
'On the role of the Catholic Church and the doctrine of "Liberation Theology," see Lernotux (1980), and
Montgomery (1982, 1983).
10 In Central America, a further consequence of the dislocations described above has been the brea
peasant families and an increase in the number of female-headed households, especially in ur-ban barrios. Out of
economic necessity, women have come to serve as the backbone of many of the grass-roots communal
organizations. When these organizations become targets of governmenlt repression, women are not spared.
Consequently, they too have been mobilized-independen-tly of their spouses-for participation in Central
American guerrilla movements as combatants and not just as support personnel. In Nicaragua, for instance,
women constituted an estimated 30 percent of the Sandinista army and held positions as commanders of
everything from small combat units to full battalions (Flynn, 1983). Women made up half of the membership of the
grass-roots popular organizations that eventually came to serve as a support base for the FMLN in El Salvador
(Montgomery, 1982). On the role of women in Central American revolutions, seeJaquette (1973), Ramirez-Horton
(1982), Brewer (1983), Chin-chilla (1983), and Mason (1986d).
" The characteristics of the authoritarian state are the subject of major works by Collier (1979) and O'Donnell
(1973). Sloan (1984) and Gurr (1985a, 1985b, 1986) discuss its propensity for repressive rather than accommodative
strategies.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
184 The Political Economy of Death Squads
toward repression. Such assistance is usually conditional upon the regime's adoption
of conservative monetary and fiscal policies. Such measures intensify the economic
hardships facing nonelites, which in turn fuel increased opposition activity. The
regime reacts with further repression because to accede to opposition demands
would be to violate IMF conditionality and thus jeopardize international sources of
assistance (Pion-Berlin 1984:107-8). Foreign benefactors are usually willing to
support repressive policies through generous grants of security assistance.
In short, the authoritarian regime lacks the institutional machinery, redistributable
resources, and political inclination to pursue more accomodative strategies.
Instead, regimes that have been successful in the past in suppressing opposition
through violent repression are likely to resort to such tactics in the future, regardless
of the character of the opposition organizations or their demands (Gurr, 1986:54-
57). Coercion is the response for which the requisite institutional machinery (i.e., the
security forces) is the most extensive and well-established in the authoritarian state.
The ideology of the military is such that political tranquility is presumed to be a
prerequisite for economic development. On this basis the military often comes to
serve as the ultimate arbiter of political authority in the system (Pion-Berlin, 1984;
Lopez, 1986; Seligson, 1987:4-5). Hence, the almost reflexive response to opposition
challenges is to increase the level of officially sanctioned violence directed
against the opposition leaders and their actual, suspected, or potential supporters.
However, the very forces that fuel growing opposition in the first place will in time
erode the effectiveness of a strictly coercive strategy of control. Unrelieved poverty
and continued population growth limit the coercive capacity of the regime in several
ways. First of all, widening economic distress contributes to expanded nonelite
involvement in grass-roots organizations. If the regime begins targeting the organized
opposition, a greater proportion of the nonelite population will see themselves
as potential victims of this repression because far more of them are involved in
grass-roots organizations. Hence, more of them will choose to join an insurgent
opposition-rather than do nothing-in search of some protection against statesanctioned
terror.
Second, high rates of population growth shift the age distribution in such a way
that the number and proportion of young adults in the population increases
dramatically (Gendell, 1986:52-59). In general, young adults are more willing to
participate in political violence than are their older counterparts, and the probability
of civil disorders in a locale is positively related to the number and proportion of
disaffected young adults in the local population.'2 From a rational choice perspective,
young adults are more willing to participate in civil violence because they have
more to gain and less to lose than do their older counterparts. They have fewer assets
to risk and fewer responsibilities to tie them down and prevent them from joining an
insurgent organization. Furthermore, whatever costs young adults do risk will be
discounted to present value over a longer remaining life span, and any collective
benefits that might result from successful revolt will be compounded over a longer
remaining life span. In short, the expected lifetime benefits are greater and the
expected costs are less for young adults than for their elders. The shift in the age
distribution of the population limits the coercive capacity of the regime by presenting
12 For instance, Miller, Bolce, and Halligan (1976) foun-d that age was more str-ongly related to riot participation
among American blacks than was any other demographic variable (i.e., income, occupation, education). Mason anld
Murtagh (1985) showed that age and gen-der distinguished riot participants from non-participan-ts (with young
males bein-g most likely to riot). The disproportionate par-ticipation of youth in the events of 1968 in Franlce, in
anti-war protests in America, and in the disturbances of the Cultural Revolution in Chin-a in-dicate that the
relationship between youth and political instability is not culture-bound or time-bound: the larger the young adult
age cohort, the greater the probability of political disturbances (Kelly and Galle, 1984:107; see also Moller-, 1968).
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
T. DAVID MASON AND DALE A. KRANE 185
it with an expanding pool of potential insurgents that it must control through
coercion. Identifying actual insurgents from among the large and growing population
of potential insurgents becomes increasingly problematic, and death squad
violence becomes less precisely targeted. Consequently, the risk of death from
indiscriminate repression begins to approach or exceed the risk of death from
participation in an insurgent organization.
Finally, the need for a constant increase in coercive capacity, combined with the
limited financial resources at the disposal of the regime, implies that at some point the
quality of the added security forces will begin to decline. The regime will be compelled
to rely on ill-trained recruits and draftees, as well as paramilitary units, irregulars, and
local death squads that are less subject to direct supervision and control by either the
central military command or civilian political authorities. As raw trainees and paramilitary
death squads are unleashed on the countryside, their blunt, arbitrary application
of force will exacerbate the process by which the otherwise uninvolved turn to
the insurgents for security from state-sanctioned terror. In this manner, the exclusive
reliance on a coercive strategy will eventually contribute to, rather than retard,
the growth in support for the insurgent opposition (Mason, 1987).
Eventually, the regime's financial resources will be strained by the cost of
constantly expanding its coercive machinery. Fewer resources will remain for
investment in what accommodative programs the state may be willing to undertake.
The growing financial burden of coercive control will come to be felt by the very
elites (domestic and foreign) in whose interests the state is attempting, through
repression, to maintain the political and economic status quo. These pressures may at
some point create strains within the authoritarian alliance itself, leading to a coup (as
in Guatemala in 1973 and El Salvador in 1979), loss of the support of a significant
segment of the economic or political elite (as in Brazil from the mid-1970s), or the
exodus of that elite (as in Cuba in the late 1950s). Whatever the outcome the
consequences for the regime's stability are potentially devastating, as evidenced by
the fall of the Somoza dynasty in Nicaragua in 1979. In the following section, we shall
apply the theoretical framework developed thus far to an analysis of the impact of
death squad violence on regime stability in El Salvador.
El Salvador: The Political Economy of Repression, Reform, and Revolution
The current crisis in El Salvador can be traced to the pattern of dependent
development described earlier: severe land pressures catalyzed by a shift to export
agriculture and the rapid growth of the rural population, the failure of an
authoritarian regime to relieve these pressures through redistributive reforms or
developmental initiatives, and the subsequent violent struggle between opposition
and regime. These dynamics have shaped both the state's use of repressive violence
and the nonelites' response to state-sanctioned terror.'3
When independence was declared in 1839 the peasant population of El Salvador
was distributed largely among four land tenure arrangements: (1) as colonos or
permanent workers on large haaciendas, (2) as smallholders or tenants on the
surrounding minifundia, (3) as landless laborers and squatters, or (4) as residents of
communal lands (ejidos). As long as total population remained relatively stable, the
power of the landed elite over the peasantry was moderated by the fact that a
plurality of the land was owned communally in the form of the ejidos. Those lands
13 Browning (1971) and Burns (1984) present detailed treatments of the impact of export agriculture on
Salvadoran development. Analyses of recent events in the crisis include Baloyra (1982), Montgomery (1982), and
Shulz (1984). Durham (1979) presents an especially rigorous an-alysis of the impact of economic change and
population growth on political stability.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
186 The Political Economy of Death Squads
served as an escape valve of sorts: to the extent that disaffected peasants could be
absorbed into the ejidos, intermittent land pressures and peasant discontent could b
relieved by internal migration.
The opening of Pacific coast shipping in the 1870s made it possible for Salvadoran
landowners to enter the booming coffee trade. The shift to coffee production
provoked profound changes in the land tenure system that eventually eroded the
traditional clientelist stability in the countryside of El Salvador. Because the lands best
suited for coffee cultivation were in the ejidos, the Salvadoran state undertook the
statutory abolition of communal ownership in the late 1800s in order to enable the
plantation oligarchy ("the fourteen families") to acquire these lands and convert
them to coffee production. This land redistribution displaced thousands of peasants
and foreclosed the major mechanism for relief of periodic land pressures (Browning,
1971 :ch. 5). Structural dislocations in the countryside were exacerbated by the
expansion of total population from 554,000 in 1878 to 1,493,000 by 1931 (Browning,
1971:171). Thus, the traditional mechanisms by which land pressures had been
relieved were either dismantled by the shifts in land use and land tenure or were
overwhelmed by the accompanying demographic transition (Mason, 1986a:496-99;
Burns, 1984:298-302). Eventually, these pressures erupted into the peasant revolt
of 1932 which the state brutally suppressed by killing between ten and thirty
thousand peasants in what has come to be known as "la matanza."'14
La matanza marks the point at which an authoritarian alliance between the
Salvadoran military and the landed elite turned to an almost exclusive reliance on
coercion as a way of preempting further unrest in the countryside (Hoeffel, 1984: 101,
Shulz, 1984:194-98). However, coercion alone could not contain the momentum of
further economic and demographic transformations that continued to erode the
foundations of social stability. The introduction of additional export crops (e.g.,
cotton, cattle, sugar) in the post World-War II era raised the value of lands not suited
for coffee production and thus contributed to further concentration of land
ownership and displacement of peasant households. As the landless labor force
grew, demand for agricultural labor declined because the production characteristics
of the new crops did not require a large permanent labor force. Cotton production
has been increasingly subject to mechanization, and cattle production has a high
land-to-labor ratio. Furthermore, their production displaced food crops, reducing
the local supply of subsistence goods (Durham, 1979; DeWalt, 1985). Food grain
imports rose throughout most of the 1960s and 1970s and prices rose with them, as
evidenced by the 182 percent increase in the consumer price index between 1970
and 1977 (Montgomery, 1982:29-30).
By 1975 almost 60 percent of the land was owned by only two percent of the
population. By contrast, over 40 percent of the rural population had no access to
land, and of those that did, 49 percent had less than one hectare and 94 percent had
less than ten hectares (Shulz, 1984:191; see also Simon and Stephens, 1982:160-61,
Montgomery, 1982:27-28; Graham, 1984:174-76). For the 40 percent of the
population that was landless, wages provided only 52 percent of family income. For
smallholders and tenants with less than two hectares of land, crops provided only
one-third of household income (Brewer, 1983:400).
Rapid population growth accelerated the erosion of the nonelites' economic
plight. Total population more than doubled between 1960 and 1982, and annual
growth rates have remained between 2.5 and 3.5 percent. The dependency ratio has
14 Estimates of the nuLmber of people killed in la ?nataslza vary widely. See An-derson (1971:135) for a discussio
he validity of the differ-ent estimates.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
T. DAVID MASON AND DALE A. KRANE 187
remained above 90 percent (the figure for the U.S. in 1980 was 51 percent).'5 Thus,
local capital formation has been overwhelmed by growing consumption demands.
By 1970, agricultural density (the ratio of population to arable land) had increased to
266.2 persons per square kilometer, the highest ratio in Latin America (Taylor and
Jodice, 1983). Some 70 percent of the total land surface was already under
cultivation (Shulz, 1984:191). Migration to Honduras, a traditional escape valve for
land-poor Salvadorans, was foreclosed by the Soccer War of 1969 (Durham, 1979).
Those who have access to land have overworked it to the point that soil depletion and
erosion have become major problems. Those who lacked access to land have resorted
to planting food crops in the most inhospitable of locales, including the slopes of
active volcanoes. In the aggregate, these actions have contributed to serious
deforestation and soil erosion problems (Shulz, 1984:191).
The Salvadoran state has lacked both the institutional capacity and the political
volition to create occupational alternatives to agriculture. Industrial growth has
never been sufficient to absorb those displaced by shifts in land use, and population
growth has simply exacerbated this problem. Although manufacturing grew by 24
percent between 1961 and 1971, industrial employment grew by only six percent
(Shulz, 1984:193). Between 1960 and 1982, the size of the working age population
grew by 1.4 million, while industrial employment grew only from 230,000 to
600,000.16 The one million new workers not absorbed by what is often characterized
as the industrial "boom" of the 1960s were left to compete for a limited supply of low
paying jobs in the agricultural and service sectors of the economy. Indeed, the
proportion of the economically active population employed in industry actually
declined from 13 percent in 1961 to only 10 percent in 1971 (Shulz, 1984:193).
Thus, the demographic and economic transformations occurring in El Salvador
since 1945 have left the majority of the pesantry devoid of any security against
fluctuations in the deteriorating markets for land, labor, and subsistence goods.
In the absence of reforms by the state, nonelites increasingly have turned to a new
generation of Catholic priests and lay workers who, inspired by the ideals of
liberation theology, have begun mobilizing the poor in a network of mass-based
organizations intended to provide relief from poverty and impotence. Initially, these
grass-roots organizations addressed the immediate economic needs of the poor by
providing services such as health and hygiene programs, child care cooperatives,
food distribution networks, and literacy programs. By the 1970s, popular organizations
had evolved into a mass support base for political opposition groups challenging
the policies and eventually the legitimacy of the military-dominated regime
(Lernoux, 1980; Montgomery, 1982:123-30).
As the level of opposition activity grew, the mass organizations (and especially their
leaders) became the targets of an escalating campaign of violent repression perpetrated
by government security forces and paramilitary organizations such as ORDEN
(the Nationalist Democratic Organization), the UGB (Union Guerrillera Blanca or
White Warrior's Union), the Maximiliano Hernandez Martinez Squad, Mano Blanco
(the White Hand), and FALANGE (Fuerzas Armadas de Liberacion Anticommunista or
Anti-Communist Armed Forces of Liberation). State-sanctioned terror was coupled
with electoral fraud to overturn elections in 1972 and 1977 that would have brought to
15 The dependency ratio is defined as the combined population under 15 and over 64 years old as a percentage of
the population between those ages.
16 The figures on industrial employment are calculated using figures for 1960 and 1982 on total population s
percent of the population between 15 and 64, and percent of the wor-kforce in industry. The data wvere taken from
Taylor and Jodice (1983) and the World Bank's World Development Report 1984.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
188 The Political Economy of Death Squads
power reform-minded Christian Democratic governments. Instead, opposition
leaders (including former President Jose Napoleon Duarte) were jailed or exiled.
Between 1972 and 1980 repression of the rank-and-file supporters of the
opposition became more widespread and lethal. The National Guard killed fifty
peasants and wounded two hundred others at La Cayetana in 1974 (Shulz,
1984:202). In 1975, participants in demonstrations protesting exorbitant government
expenditures on the Miss Universe pageant held in San Salvador were gunned down
by government troops. In February of 1978 hundreds were wounded or killed and
thousands arrested or detained by the National Guard following a major demonstration
protesting electoral fraud (Americas Watch, 1982:39). In 1979, twenty-one
peasants were killed during a peaceful demonstration, and two months later another
twenty were massacred in Coatepeque (Wheaton, 1980:8). In 1980, the largest
demonstration in El Salvador's history-estimates run from 80,000 to 200,000
participants-was fired upon, leaving dozens dead and hundreds injured (Leiken,
1984:117).
With the escalating repression of legitimate groups, opposition organizations
began adopting more militant and violent tactics of their own. The emergence of
groups such as the Popular Revolutionary Bloc (BPR), the Unified Popular Action
Front (FAPU), and the Popular Leagues of the 28th of February (LP-28) marked the
opposition's shift to a strategy to topple the regime by revolutionary means.'7 In
1980, the Faribundo Marti National Liberation Front (FMLN) and the Democratic
Revolutionary Front (FDR) were formed as umbrella organizations uniting the
fragmented insurgent groups into a coordinated revolutionary opposition pledged
to remedying the economic distress of nonelites through the institution of democratic
reforms, the redistribution of land, the creation of rural credit and marketing
services, wage reforms, and industrial expansion.
Despite these promises, the willingness of nonelites to shift their loyalties to the
FMLN-FDR has been constrained by the accompanying risk of lethal reprisals at the
hands of the death squads. However, as the level of coercion has increased, its
application has become grossly indiscriminate with respect to both the selection of
targets and the amount of force used. Statistics compiled by the Archbishop's Legal
Aid Office document that five to six persons a day were assassinated by death squads
in the first half of 1979. The figure rose to nine to ten per day by the year's end
(Wheaton, 1980:8).
We have argued that as repression becomes more pervasive and more arbitrary, its
deterrent effect on opposition support should diminish because the likelihood of
becoming a victim is no longer related to one's support or non-support of the
insurgents. The growth of the FMLN's active forces and its base of covert support
during the late 1970s suggests such a shift in peasant loyalties and its correlation with
the escalation of indiscriminate repressive violence. By 1979, the FMLN had an
estimated 6,000-8,000 active guerrillas, as many as 100,000 part-time militia, and a
reported one million sympathizers who provided covert support in the form of food,
intelligence, and sanctuary (Leiken, 1984:118).
Increasing civil violence heightened the long-standing conflict between the landed
oligarchy and the military, on the one hand, and the nonrevolutionary opposition led
by the Christian Democratic Party (PDC), on the other, over the efficacy of the state's
exclusive reliance on repression as a strategy of control. By October 1979, the more
17 Other such organizations included the Popular Forces of Liberation (FPL) and its military arm, the Armed
Forces of Popular Liberatioin (FAPL), the Armed Forces of Liberation (FAL) which is affiliated with the Salvadoran
Communist Party (PCS), the Armed Forces of National Resistance (FARN), the People's RevoluLtionary Army
(ERP), and others. The origin and composition of these and other insurgent oiganizations are discussed in
Armstrong aind Shenk (1982), Leiken (1984), ShuLlz (1984).
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
T. DAVID MASON AND DALE A. KRANE 189
progressive elements of the Salvadoran military had come to the conclusion that
accommodative reforms of some sort would be required to stop the further erosion
of popular support. Hence, they staged a coup that installed a civilian-military junta
pledged to institute reforms. Despite the defection of key civilians dissatisfied with
the pace and scope of the reforms and the lack of civilian control over the security
forces, a reconstituted junta announced a land reform program and a return to
civilian rule under a new constitution to be written by a popularly elected Constituent
Assembly.
The Salvadoran reform program of 1980 represents a shift in the regime's strategy
for winning popular support from one based almost exclusively on coercion to one
based more heavily on remedying the deteriorating economic plight of the rural and
urban poor. The success of this strategy would depend, first of all, on the extent to
which the announced reforms actually were implemented and, second, on the extent
to which their provisions actually relieved the economic distress of displaced peasants.
Most importantly, success would depend on the extent to which the civilian government
could exert sufficient control over the security forces and their paramilitary
counterparts to reduce the level and arbitrariness of repressive violence.'8
The 1980 land reform was designed to proceed in three phases. Phase I would
convert estates larger than 500 hectares into peasant cooperatives. Phase II would
affect farms between 100 and 500 hectares. Phase III would transfer title of rented
land to the current tenants, up to a limit of seven hectares per household. In each
phase, former landowners would be compensated with a combination of cash and
redeemable bonds (Browning, 1983; Simon and Stephens, 1982). Phase I was largely
completed by the time of the 1982 Constituent Assembly elections, and Phase III's
implementation had already begun. Phase II, which would have included most of the
coffee acreage, was postponed indefinitely, reflecting the still-powerful influence of
the coffee oligarchy.
The effectiveness of Phases I and III in restoring popular support was hampered by
flaws in their design and implementation. Most significantly, only peasants who were
already employed as permanent workers on the estates were allowed to enter Phase I
cooperatives, and only peasants already renting land could apply for title under Phase
III. Therefore, only about 100,000 people out of a population of five million received
land under Phases I and III, and the landless population remained virtually untouched
by the benefits of reform. Implementation of reform was hampered by the
sheer volume of administrative work involved in the transfer of title on the estimated
100,000 parcels of land covered by Phase III. As a result, there were serious delays
which allowed landlords time to evict peasants before they could obtain title to their
land (Strasma et al., 1983:135-42). Many who qualified for land under Phase III
were reluctant to apply for title because they were renting from relatives or from
other peasants who owned small amounts of land scattered among several plots. The
specific provisions of Phase IL were not delineated until the December 1983
promulgation of the new constitution. The most important provision was that within
three years, affected landowners must sell to peasants or peasant organizations all
land in excess of 245 hectares or face expropriation of that land by the government
without prior compensation. This ceiling would free 20,000-70,000 acres of land for
18 Of couLrse, the suLccess of the program also depended on a substanitial infuLsioni of U.S. economllic aind technical
assistance ($3.2 billion between 1980 and 1987). Likewise, the wvillingness aind ability of the state to coiltrol the
excesses of the secuLrity forces and death squLads also depended in pai-t on the willingness of the U.S. to manipulate
its suLbstantial infuLsioin to militai-y aid (some $1 billion ovei- the same time peiriod) to that eind. Whethei U.S. aid has
in fact had any instruLmental impact on the human rights situation in El Salvador is certainly debatable. For a
discussion of the ansouints and forms of this aid aind some of its effects, see Siegel and Hackel (1988:115-28).
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
190 The Political Economy of Death Squads
redistribution, according to the U.S. State Department (1984:22). To date, the actual
redistribution of land under Phase II has not proceeded very far, if at all.
The opponents of reform used the time between the announcement of the
reforms and the election of the Constituent Assembly to escalate the level and
arbitrariness of death squad violence. Phase I was implemented by a "state of siege"
imposed on the estates by the military, ostensibly to enforce the implementation of
the reforms but effectively to intimidate peasants into abandoning the cooperatives.
During the reform's first year, death squads and security forces assassinated over 500
peasant leaders, dozens of officials with ISTA (the Salvadoran Institute of Agrarian
Transformation, the agency charged with implementing Phase I), Catholic priests
and lay catechists (including Archibishop Oscar Romero, who was gunned down
while saying mass), land reform workers (including Rodolpho Viera, the head of
ISTA, and two U.S. land reform advisors), as well as hundreds of members of
peasant unions and cooperatives (Simon and Stephens, 1982; Wheaton, 1980).
Similar violence reduced the effectiveness of Phase III's implementation. Because
Phase III beneficiaries were isolated renters, they were more vulnerable to intimidation
by local death squads. The administrative delays that characterized Phase III
simply provided the oligarchy with ample time to short-circuit the program through
the use of death squad violence. According to figures compiled by Socorro Juridico,
12,501 "extralegal killings of civilians not engaged in armed combat" occurred
during 1981, the year in which Phases I and III were being implemented. Over 5,000
of these victims were identified as campesinos (Simon and Stephens, 1982:61).
As the figures in Table 1 suggest, death squad violence became increasingly
widespread and random as the struggle both within the regime and between the
regime and the FMLN heightened in anticipation of the 1982 elections. In one
incident, hundreds of unarmed peasants including women, children, and elderly
people were massacred as they attempted to flee across the Sumpul River into
Honduras (Blundy, 1983). In December, some nine hundred civilians, including
women and children, were taken from their homes and executed by government
troops during a sweep through the Morazan province (Shulz, 1984:230). As a
priest in the Catholic Church's Legal Aid Office put it, "to be a victim . . . it is
enough to be a relative of a militant or to be suspected of having collaborated
with the insurgents. The concept of collective responsibility is being progressively
extended to the individual, the family, town, and even province" (quoted in Shulz,
1984:214).
The ultimate outcome of this struggle within the regime and between it and the
FMLN is, of course, yet to be determined. There is some evidence, however, that since
the reforms were effectively ratified in the elections of 1982 and 1984, the new civilian
leadership has been able to make limited progress in reversing the shift of popular
TABLE 1. Deaths from repression and counterinsurgency in El Salvador, 1983-1986.
1983 1984 1985 1986
Targeted Victims of
Death Squads/Military 1,259 224 240 123
Civilian Noncombatant Victims
of Counterinsurgency Operations 3,625 2,657 1,416 830
Disappeared 535 195 81 39
TOTAL 5,419 3,076 1,737 992
Sources: Americas Watch (1984, 1985a, 1985b, 1987).
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
T. DAVID MASON AND DALE A. KRANE 191
support to the FMLN. In the 1982 Constituent Assembly elections, the PDC gained a
plurality of seats but was denied the presidency by a coalition of right-wing parties.
However, the 1984 elections brought PDC leader Jose Napoleon Duarte to power.
Crucial to the PDC's success in restoring a base of popular support to the troubled
regime has been the government's ability to reduce (though by no means eliminate)
death squad violence. Indeed, the elections themselves, despite their flaws, have been
characterized not so much as referenda on land reform as votes for an end to violence.
The level of repressive violence, though still brutally high, showed a steady decline
between the 1982 and 1984 elections. In 1983, there were 1,259 documented killings
by the death squads, 3,625 civilian noncombatants killed by the military, and another
535 counted among the "disappeared." Although these figures represent a substantial
decline from the 12,000 killed in 1981, they also indicate the inability (or unwillingness)
of the Constituent Assembly to exert effective control over the agents of
repression. In 1984, death squad murders declined to 185 in the first six months and to
39 in the six months following Duarte's election. "Disappearances" declined as well in
each half of 1984 to 139 and 56 respectively. The killing of civilian noncombatants by
the security forces, though declining, remained abominably high at 2,657 for 1984
(Americas Watch, 1984, 1985a, 1985b). These trends continued downward under
Duarte's administration. In 1985, death squad murders declined to 240, disappearances
to 81, and civilian combat deaths to 1,416 (Americas Watch, 1986:2); in 1986,
the figures were 123, 39, and 830, respectively (Americas Watch, 1987:7).
The total of 950+ deaths at the hands of the regime in 1986 represent an
improvement only in comparison to the bloodbath of 1979-1983. There is no
evidence of a dramatic decline in the number of full-time guerrillas, nor should we
expect there to be since, in the absence of a credible amnesty program, active
guerrilla fighters can expect death after surrender."9 By the same token, there is n
evidence of a significant growth in the rebel forces. The decline in the level and
arbitrariness of repression, coupled with the limited but positive effects of the
reform program, may well have reduced the level of covert support for the FMLN as
well. Though the Reagan administration liked to attribute these developments to
more effective counterinsurgency tactics, it is our contention, based on the theory
presented here, that the explanation lies elsewhere.
First, the land reform program, despite its limitations, has co-opted a large
number of direct beneficiaries from participation in insurgency. More than likely,
the promise of further reforms has made other nonelites less willing to lend support
to the rebels. For them, the political choice is now between some chance of reform
benefits with reduced risks of violence so long as the PDC remains in power versus a
more remote chance of reform benefits from an FMLN victory-and those coming
only after a protracted and bloody civil war. Second, the FMLN's own tactics of
disrupting the economy have apparently alienated a large number of peasants
(Leiken, 1984) and made them more susceptible to the promise of peaceful reform
by the PDC. Most important, the decline in nonelite support for the FMLN can be
attributed to the dramatic decline in the level and arbitrariness of death squad
violence. As repression has become less pervasive and arbitrary, the option of
remaining uninvolved in the conflict has been restored.
The largest block of victims since 1984 has been noncombatant victims of
counterinsurgency operations by the Salvadoran military, not of death squad
19 The Arias Central American Peace Plan contained amnesty provisions. However, the Salvadoran Assemb
enacted a law that gran-ted amnesty nlot just for- r-ebels buLt for- anyone who committed a crime "in connilection with
the conflict before October 22, 1987. ThuLs, the members of the military and the death squtads who are responsible
for some 44,000 non-combat-related civilian deaths over the couLrse of that last decade were granted amnesty along
with any rebels willing to lay down- their arms. While several exiled politicians have returned to participate in the
1988 elections, few if anly rebels have participated in the amnesty program (America's Watch, 1988:2-5).
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
192 The Political Economy of Death Squads
violence. This development is attributable to the government's shift to tactics that
rely on air strikes and the designation of free-fire zones in regions of heavy rebel
activity (Americas Watch, 1984, 1985a; Lawyers Committee for International Human
Rights, 1984). In these circumstances, a rational strategy for a war-weary
nonelite would be to withhold any covert or overt support for the guerrillas because
any actions that marginally increase the level of guerrilla activity in an area also
increase the likelihood of local counterinsurgency operations by the military. Given
current tactics of the Salvadoran military, life in regions of heavy counterinsurgency
activity is extremely dangerous, to say the least. A nonelite would have a strong
incentive either to migrate or to join the rebels. The growth in the population of El
Salvador's internal refugee camps (Americas Watch, 1985a; Lawyer's Committee for
International Human Rights, 1984) and the continued high incidence of civilian
battle deaths are testimony to these effects in the Salvadoran countryside.20
While the substitution of official counterinsurgency operations for indiscriminate
death squad violence may have stemmed the erosion of the government's stability, it
has by no means increased the level of active support for the regime. Because
counterinsurgency violence is more predictable than death squad violence, there are
measures that nonelites can take to avoid becoming victims. In this respect,
substituting counterinsurgency operations for death squad violence probably has
reduced the level of covert support for the FMLN. However, it does so by converting
the civilian population of the affected region into either refugees who are forcibly
relocated or active rebel supporters who stay behind. In a free-fire zone, no other
choices exist. Those who are torn from their home for relocation in a refugee camp
can hardly be counted upon as active supporters of the regime, any more than can
landlords whose property was expropriated-with compensation-in the land
reform program. Refugees who left friends or family behind would also have some
incentive to provide whatever covert support they could to the rebels. In the absence
of any political efforts (such as an amnesty program) that would induce rebels to lay
down their arms, the government is faced with the prospect of a protracted and
costly military stalemate with an insurgent army that has nothing to lose by
continuing the battle.
CONCLUSION
A central theoretical question in the literature on state-sanctioned terror is whether,
and under what conditions, repressive violence deters or induces a shift in popular
support away from the regime and toward the opposition. What is perhaps unique
about this analysis is that we have approached this question by combining a rational
choice model of the nonelite response to escalating levels of death squad violence
with a structural analysis of the global and domestic conditions under which the
escalation of state-sanctioned terror can be expected. In particular, we have
described the ways in which variations in the targeting of such violence affect the tide
of popular support in societies torn by civil violence and repression. Of significance is
the argument that carefully targeted repressive violence may reduce the level of
active popular support for the opposition, while widespread indiscriminate violence
may in fact produce increases in active support for the opposition.
20 The FMLN's reaction to the escalation- of the war by the military has been to break its forces into small nLnits of
eight to twelve persons that ar-e less vnLlnerable to detection an-d attack by the military. The increased uLse of air
strikes in FMLN zones of control has led the FMLN to expand the war geographically by dispersing these units
throughout the cotuntry and shifting to a greater reliance on hit-and-run attacks as opposed to battlefield
confrontations. However, highly centralized leadership still allows the FMLN to reconcentrate forces for large-scale
attacks, such as the raids on military facilities in La Union (October, 1985), San MiguLel (June, 1986), and El
Paraision (April, 1987). See Siegel and Hackel (1988).
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
T. DAVID MASON AND DALE A. KRANE 193
However, even carefully targeted repression does not redress the grievances that
gave rise to opposition in the first place; it simply terrorizes nonelites into not acting
overtly to seek redress of those grievances. This raises a question that needs to be
addressed in future efforts to elaborate more precisely the formal elements of this
theory. Specifically, our analysis implies that carefully targeted repression may deter
nonelite behaviors that support the opposition, but it does not shift nonelite preferences
in favor of the regime. Quite the contrary, nonelite preferences should shift in favor
of the opposition, even though their willingness to express them with overt acts of
support may be deterred by repression. This distinction between preferences and
supportive behaviors, expressed eloquently by Christopher Dickey (1984) in the
phrase "obedezco pero no cumplo" ("I obey but I do not comply"), was raised by Leites
and Wolf (1970), but its theoretical implications were never sufficiently explored
because of their near-exclusive concern with competing elite strategies of insurgency
and counterinsurgency. By assessing the behavioral calculus of individual nonelites,
we have begun to lay the theoretical groundwork for a more refined model of the
ways in which repression affects the preferences and behaviors of individual
nonelites and, therefore, the aggregate level of popular support for the regime (cf.
Mason, 1986b; 1987).
This distinction contains certain implications for policy makers as well. In
particular, U.S. policy toward insurgencies has consistently treated them as military
problems whose resolution is simply a matter of implementing the appropriate
counterinsurgency tactics. The fact that current policy debates over "low intensity
conflict" seem to focus more on questions of force structure and use doctrine is
indicative of this tendency (see Klare and Kornbluh, 1988). While conventional
counterinsurgency tactics may stem the growth in active support for the rebels, they
do nothing to reduce the incentives for those already marked as rebels or rebel
supporters to continue the battle. Protracted low-intensity warfare is a burden that
the war-weary populations and fragile economies of Third World nations can ill
afford to bear. Our analysis suggests that, ultimately, non-military initiatives such as
negotiated cease-fires and amnesty programs designed to integrate an insurgent
opposition back into the mainstream of conventional politics may be the only way to
bring an insurgency to a peaceful conclusion. In the absence of any credible
assurances against violent reprisals, rebels have nothing to lose by continuing their
own program of violent opposition.
Our analysis also suggests that structural reform in the local economy is essential to
the long-term stability of any regime in these nations. Where non-military options
such as land reform have been tried, the assumptions underlying their design and
implementation often have been equally simplistic. The idea behind such U.S.backed
policies in Vietnam and El Salvador was that "land to the tiller" programs
would "win the hearts and minds" of rural populations by giving them a stake in the
status quo (Prosterman, 1981). However, such programs ignore the marginal
economic existence of the newly created peasant smallholder and, by themselves, do
not reduce his or her vulnerability to either the economic risks of agriculture or the
political risks of death squad violence. Any land reform program that does nothing
to reduce the economic marginality of beneficiaries or to create occupational
alternatives to agriculture for non-beneficiaries has little hope of remedying the
underlying structural sources of instability. Moreover, economic reforms that do
nothing to reduce the threat of repressive violence are unlikely to succeed, because
unchecked repression by foes of reform within the regime can undermine the
success of land reform by intimidating potential beneficiaries into abandoning their
right to claim title to land.
In principle, the propositions presented in this article can be tested empirically
(e.g., by testing the pattern of relationships between land reform, death squad
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
194 The Political Economy of Death Squads
activity, and the level of support for the regime in El Salvador; see Mason, 1986a).
The analysis presented here has contributed to that broader effort by elaborating
more fully the theoretical link between repression and nonelite support for the
regime. Much work remains to be done before our understanding of the politics of
repression even approaches the state of knowledge on the dynamics of revolution.
This study has focused analytical attention on nonelite response to repressive
violence, because the effects of repression can never be fully understood until we
explore the ways in which the calculus of fear determines nonelite response to
different levels of repressive violence under different sets of structural conditions.
References
AGUILERA PERALTA, G. (1980) Terror and Violence as Weapons of Counterinsurgency in Guatemala.
Latin American Perspectives 7:91-113.
AMERICAS WATCH. (1982) Report on Human Righlts in El Salvador. New York: Vintage.
AMERICAS WATCH. (1984) Free Fire: A Report on Humnan Rights in El Salvador. New York: Americas
Watch Committee.
AMERICAS WATCH. (1985a) Draining the Sea: Sixth Supplement to the Report on Humian Rights in E
Salvador. New York: Americas Watch Committee.
AMERICAS WATCH. (1985b) The Continuing Terror: Seventh Supplement to the Report on Human Rights in
El Salvador. New York: Americas Watch Committee.
AMERICAS WATCH. (1986) Settling into Routine: Human Rights Abuses in Duar-te's Second Year. New Y
Americas Watch Committee.
AMERICAS WATCH. (1987) The Civilian Toll, 1986-1987. New York: Americas Watch Committee.
AMERICAS WATCH. (1988) Human Rights in El Salvador on the Eve of Elections. New York: Americas
Watch Committee.
ANDERSON, T. (1971) Matanza: El Salvador's Commuunist Revolt of 1932. Lincoln, NE: University of
Nebraska Press.
ARMSTRONG, R. AND J. SHENK. (1982) El Salvador: The Face of Revolution. Boston: South End Press.
BALOYRA, E. (1982) El Salvador in Transition. Chapel Hill, NC: University of North Carolina Press.
BLUNDY, D. (1983) Victims of the Massacre that the World Ignored. In Revolution in Central America,
edited by Stanford Central American Action Network, pp. 337-42. Boulder, CO: Westview Press.
BREWER, T. H. (1983) Women in El Salvador. In Revolution in Central America, edited by Stanford
Central America Action Network, pp. 400-08. Boulder, CO: Westview Press.
BROWNING, D. (1971) El Salvador: Landscape and Society. Oxford: Clarendon Press.
BROWNING, D. (1983) Agrarian Reform in El Salvador. Journal of Latin American Studies 15:399-426.
BURNS, E. B. (1984) The Modernization of Underdevelopment: El Salvador, 1858-1931. Journal of
Developing Areas 18:293-316.
BWY, D. P. (1968) Political Instability in Latin America: The Cross-Cultural Test of a Causal Model.
Latin American Research Review 3:17-66.
CARDOSO, F. H. (1973) Associated Dependent Development: Theoretical and Practical Implications.
In Authoritarian Brazil: Origins, Policies, and Future, edited by A. Stepan, pp. 142-76. New Haven,
CT: Yale University Press.
CARDOSO, F. H. (1977) The Consumption of Dependency Theory in the United States. Latin American
Research Review 12:7-24.
CARDOSO, F. H. AND E. FALETTO. (1979) Dependency and Development in Latin America. Berkeley, CA:
University of California Press.
CHAFFEE, W. A., JR. (1979) Let Jorge Do It: A Rational Choice Model of Political Participation. In
Politics and the Poor: Political Participation in Latin America, Vol. II, edited by M. A. Seligson and
J. A. Booth, pp. 18-34. New York: Holmes and Meier.
CHAMBERLIN, J. (1974) The Provision of Public Goods as a Function of Group Size. Amierican Pol
Science Review 68:707-15.
CHINCHILLA, N. S. (1983) Women in Revolutionary Movements: The Case of Nicaragua. In
Revolution in Central America, edited by Stanford Central America Action Network. Boulder, CO:
Westview Press.
COLLIER, D. (1979) The New Authoritarianism in Latin America. Princeton, NJ: Princeton University
Press.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
T. DAVID MASON AND DALE A. KRANE 195
DAVIES, J. C. (1962) Toward a Theory of Revolution. American Sociological Review 27:5-18.
DENARDO, J. (1984) Power in Numbers: The Political Strategy of Protest and Rebellion. Princeton, NJ:
Princeton University Press.
DENEMARK, R. A. AND H. P. LEHMAN. (1984) South African State Terror: the Costs of Continuling
Repression. In The State as Terrorist: the Dynamics of Government Violence and Repression, edited by
M. Stohl and G. A. Lopez, pp. 143-65. Westport, CT: Greenwood Press.
DEWALT, B. R. (1985) The Agrarian Bases of Conflict in Central America. In The Central American
Crisis: Sources of Conflict and the Failure of U.S. Policy, edited by K. W. Coleman and G. C. Herring,
pp. 43-54. Wilmington, DL: Scholarly Resources.
DICKEY, C. (1984) 'Obedezco Pero No Complo' (I obey but I do not comply). In Central America:
Anatomry of Conflict, edited by R. L. Leiken, pp. 33-48. Elmsford, NY: Pergamon Press.
DURHAM, W. H. (1979) Scarcity and Survival in Central America: Ecological Origins of the Soccer War.
Stanford, CA: Stanford University Press.
DUVALL, R. D. AND M. STOHL. (1983) Governance by Terror. In The Politics of Terrorism, 2nd ed.,
edited by M. Stohl, pp. 179-219. New York: Marcel Dekker, Inc.
ECKSTEIN, H. (1965) On the Etiology of Internal Wars. History and Theory 4:133-63.
FEIERABEND, I. K. AND R. L. FEIERABEND. (1972) Systematic Conditions of Political Aggression: An
Application of Frustration Aggression Theory. In Anger, Violence, and Politics: Theories and
Research, edited by I. K. Feierabend, R. L. Feierabend, and T. R. Gurr, pp. 136-83. Englewood
Cliffs, NJ: Prentice-Hall.
FLYNN, P. (1983) Women Challenge the Myth. In Revolution in Central America, edited by Stanford
Central America Action Network. Boulder, CO: Westview Press.
FRANK, A. G. (1972) The Development of Underdevelopment. In Dependence and Underdevelopment:
Latin America's Political Economry, edited by J. D. Cockroft et al. New York: Anchor Books.
FROHLICH, N. AND J. A. OPPENHEIMER. (1970) I Get by with a Little Help from My Friends. World
Politics 23:104-20.
FROHLICH, N. AND J. A. OPPENHEIMER (1974) The Carrot and the Stick: Optimal Program Mixes fo
Entrepreneurial Political Leaders. Public Choice 19:43-61.
FROHLICH, N. ANDJ. A. OPPENHEIMER. (1973) Governmental Violence and Tax Revenue. In Violence
as Politics, edited by H. Hirsch and D. Perry. New York: Harper.
FROHLICH, N., J. A. OPPENHEIMER, AND 0. YOUNG. (1971) Political Leadership and Collective Goods.
Princeton, NJ: Princeton University Press.
GENDELL, M. (1986) Population Growth and Labor Force Absorption in Latin America, 1970-2000.
In Population Growth in Latin America and U.S. National Security, edited by J. Saunders, pp. 49-78.
New York: Allen and Unwin.
GRAFSTEIN, R. (1987) Rational Choice and Social Theory. Social Science Quarterly 68:358-63.
GRAHAM, D. H. (1984) The Economic Dimensions of Instability and Decline in Central America and
the Caribbean. In Revolution and Counterrevolution in Central America and the Carribbean, edited by
D. E. Shuilz and D. H. Graham, pp. 157-88. Boulder, CO: Westview Press.
GURR, T. R. (1969). A Comparative Study of Civil Strife. In Violence in America: Historical and
Comparative Perspectives, edited by H. D. Graham and T. R. Gurr. New York: Signet.
GURR, T. R. (1970) Why Men Rebel. Princeton, NJ: Princeton University Press.
GURR, T. R. (1985a) The Role of the State in Political Violence: a Theoretical Analysis. Paper
presented at the Session on Domestic Violence, International Political Science Association
Annual Meeting, Paris.
GURR, T. R. (1985b) Structures of Repression and Conflict: Foundations for a General Theory of
Political Coercion. Paper presented at the International Studies Association Annual Meeting,
Washington.
GURR, T. R. (1986) The Political Origins of State Violence and Terror: A Theoretical Analysis. In
Government Violence and Repression: An Agenda for Research, edited by M. Stohl and G. A. Lopez,
pp. 45-72. Westport, CT: Greenwood Press.
HADARI, S. A. (1987) What Are Preference Explanations? The Interpretive Core of Economic
Modeling. Social Science Quarterly 68:340-57.
HARFF, B. (1986) Genocide as State Terrorism. In Government Violence and Repression: An Agenda for
Research, edited by M. Stohl and G. A. Lopez, pp. 165-88. Westport, CT: Greenwood Press.
HARRIS, J. R. AND V. SAMARAWEERA. (1984) Economic Dimensions of Conflict. In Multidisciplinary
Perspectives on Population and Conflict, edited by N. Choucri, pp. 123-56. Syracuse, NY: Syracuse
University Press.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
196 The Political Economy of Death Squads
HIBBS, D. A., JR. (1973) Mass Political Violence: a Cross National Causal Analysis. New York: Wiley
Interscience.
HOEFFEL, P. H. (1984) Autumn of the Oligarchs. In Revolution and Counterrevolution in Central America
and the Carribbean, edited by D. E. Shulz and D. H. Graham, pp. 91-116. Boulder, CO: Westview
Press.
JACKSON, S., B. RUSSETT, D. SNIDAL, AND D. SYLVAN. (1978) Con-flict and Coercion in Dependenlt
States. Journal of Conflict Resolution 22:627-57.
JAQUETTE, J. S. (1973) Women in Revolutionary Movements in Latin America. Journal of Marriage and
Family 35: 344-54.
JOHNSON, C. (1982) Revolutionary Change, 2nd ed. Stanford, CA: Stanford University Press.
KELLY, W. R. AND 0. R. GALLE. (1984) Sociological Perspectives anld Evidence on Links between
Population and Conflict. In Multidisciplinary Perspectives on Population ancd Conflict, edited by N.
Choucri, pp. 91-122. Syracuse, NY: Syracuse University Press.
KERKVLIET, B. J. (1977) The Huk Rebellion: A Study of Peasant Revolt in the Philippines. Berkeley, CA
University of California Press.
KLARE, M. T. AND P. KORNBLUH, EDS. (1988) Low Intensity Conflict: Counterinsurgency, Proinsurgency,
and Antiterrorism in the Eighties. New York: Pantheon.
KRANE, D. A. (1983) Opposition Strategy and Survival in Praetorian Brazil, 1965-1979. Journal of
Politics 45:28-63.
LAWYERS COMMITTEE FOR INTERNATIONAL HUMAN RIGHTS. (1984) El Salvador's Other Victims: T
on the Displaced. New York: Americas Watch Committee.
LEIKEN, R. (1984) The Salvadoran Left. In Central America: Anatomqy of Conflict, edited by R. L. Leike
pp. 111-30. Elmsford, NY: Pergamon Press.
LEITES, N. AND C. WOLF. (1970) Rebellion and Authority. Santa Barbara, CA: Rand Corporation.
LERNOUX, P. (1980) Cry of the People. New York: Penguin Books.
LICHBACH, M. (1984) An Economic Theory of Governability: Choosing Policy and Optimizing
Performance. Public Choice 44:307-37.
LICHBACH, M. (1985a) Deterrence or Escalation? The Paradox of Government Coercion and Concession.
Paper presented at the International Studies Association Annual Meeting, Wash-
ington.
LICHBACH, M. (1985b) Deterrence or Escalation? A Microeconomic Model of Opposition Responses
to Government Coercion. Paper presented at the Annual Meeting of the American Political
Science Association, New Orleans.
LICHBACH, M. (1987) Deterrence or Escalation in Repression and Dissent.Journal of Conflict Resol
31:266-97.
LOPEZ, G. A. (1984) A Scheme for the Analysis of Government as Terrorist. In The State as Terrorist:
the Dynamics of Governmental Violence and Repression, edited by M. Stohl and G. A. Lopez, pp.
59-81. Westport, CT: Greenwood Press.
LOPEZ, G. A. (1986) National Security Ideology as an Impetus to State Violence and Terror. In
Government Violence and Repression: An Agenda for Research, edited by M. Stohl and G. A. Lopez,
pp. 74-96. Westport, CT: Greenwood Press.
MASON, T. D. (1986a) Land Reform and the Breakdown of Clientelist Politics in El Salvador.
Comparative Political Studies 18:487-516.
MASON, T. D. (1986b) Nonelite Response to State Sanctioned Terror. Paper presented at the Annual
Meeting of the American Political Science Association, Washington, D.C.
MASON, T. D. (1986c) Population Growth and the 'Struggle for the Hearts and Minds' of the Rural
Population in Central America. In Population Growth in Latin America and U.S. National Security,
edited by J. Saunders, pp. 183-222. New York: Allen and Unwin.
MASON, T. D. (1986d) Women's Participation in Central American RevoluLtions: A Theoretical
Perspective. Paper presented at the Annual Meeting of the International Studies AssociationSouth,
Atlanta.
MASON, T. D. (1987) Insurgency, Counterinsurgency, and the Rational Peasant. Paper presented at
the Annuial Meeting of the American Political Science Association, Chicago.
MASON, T. D. ANDJ. A. MURTAGH. (1985) Who Riots? An Empirical Examination of the New Urban
Black versus the Social Marginality Hypotheses. Political Behavior 7:352-73.
MIGDAL, J. S. (1974) Peasants, Politics and Revolution: Pressutres Toward Political and Social Change in the
Third World. Princeton, NJ: Princeton University Press.
MILLER, A. H., L. BOLCE, AND M. HALLIGAN. (1976) The New Urban Black Revisited. Ethnicity
3:338-67.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
T. DAVID MASON AND DALE A. KRANE 197
MITCHELL, C., M. STOHL, D. CARLETON, AND G. A. LOPEZ. (1986) State Terrorism: Issues of Concept
and Measurement. In Government Violence and Repression: An Agenda for Research, edited by M.
Stohl and G. A. Lopez, pp. 1-26. Westport, CT: Greenwood Press.
MOLLER, H. (1968) Youth as a Force in the Modern World. Comparative Studies in Society and History
10:237-60.
MONTGOMERY, T. S. (1982) Revolution in El Salvador: Origins and Evolution. Boulder, CO: Westview
Press.
MONTGOMERY, T. S. (1983) Liberation and Revolution: Christianity as a Subversive Activity in
Central America. In Trouble in our Backyard: Central America and the United States in the Eighties,
edited by M. Dishkin, pp. 75-100. New York: Pantheon Books.
MULLER, E. N. (1985) Income Inequality, Regime Repressiveness, and Political Violence. American
Sociological Review 50:47-61.
NARDIN, T. (1971) Violence and the State: a Critique of Empirical Political Theory. Beverly Hills, CA: Sa
Publications.
NICHOLSON, M. (1986) Conceptual Problems of Studying State Terrorism. In Government Violence and
Repression: An Agenda for Research, edited by M. Stohl and G. A. Lopez, pp. 27-44. Westport, CT:
Greenwood Press.
O'DONNELL, G. (1973) Modernization and Bureaucratic Authoritarianism: Studies in South American
Politics. Berkeley, CA: Institute of International Studies.
OLSON, M. (1965) The Logic of Collective Action. Cambridge, MA: Harvard University Press.
PAIGE, J. (1975) Agrarian Revolution: Social Movements and Export Agriculture in the Underdeveloped
World. New York: Free Press.
PION-BERLIN, D. (1984) The Political Economy of State Repression in Argentina. In The State as
Terrorist: the Dynamics of Governmental Violence and Repression, edited by M. Stohl and G. A. Lopez,
pp. 99-122. Westport, CT: Greenwood Press.
POPKIN, S. L. (1979) The Rational Peasant: The Political Economy of Rural Society in Vietnam. Ber
CA: University of California Press.
PREMO, D. (1981) Political Assassination in Guatemala. Journal of Inter-American Studies and World
Affairs 23:429-56.
PROSTERMAN, R. L. (1981) El Salvador's Land Reform: The Real Facts and the True Alternatives. In
El Salvador: Central America in the New Cold War, edited by M. Gettleman et al., pp. 170-78. New
York: Grove Press.
RAMIREz-HORTON, S. (1982) The Role of Women in the Nicaraguan Revolution. In Nicaragua in
Revolution, edited by T. W. Walker, pp. 161-80. New York: Praeger.
SCOTT, J. C. (1972) Patron-Client Politics and Political Change in Southeast Asia. American Political
Science Review 66:91-113.
SCOTT, J. C. (1976) The Moral Economy of the Peasant: Rebellion and Subsistence in Southeast Asia. N
Haven, CT: Yale University Press.
SCOTT, J. C. (1985) Weapons of the Weak: Everyday Forms of Peasant Resistance. New Haven, CT: Yale
University Press.
SCOTT, J. C. AND B. J. KERKVLIET. (1977) How Traditional Rural Patrons Lose their Legitimacy-A
Theory with Special Reference in Southeast Asia. In Friends, Followers, and Factions: A Reader
in Political Clientelism, edited by S. W. Schmidt, et al. Berkeley, CA: University of California
Press.
SELIGSON, M. A. (1987) Democratization in Latin America: The Current Cycle. In Authoritarians and
Democrats: Regime Transition in Latin America, edited by J. M. Malloy and M. A. Seligson, pp. 3-12.
Pittsburgh, PA: University of Pittsburgh Press.
SELIGSON, M. A. AND E. N. MULLER. (1987) Democratic Stability and Economic Crisis: Costa Rica,
1978-1983. International Studies Quarterly 31:301-26.
SHULZ, D. E. (1984) El Salvador: Revolution and Counterrevolution in the Living Museum. In
Revolution and Counterrevolution in Central America and the Caribbean, edited by D. E. Shulz and D.
H. Graham, pp. 189-268. Boulder, CO: Westview Press.
SIEGEL, D. AND J. HACKEL. (1988) El Salvador: Counterinsurgency Revisited. In Low Intensity Conflict:
Counterinsurgency, Proinsurgency, and Antiterrorism in the Eighties, edited by M. T. Klare and P.
Kornbluh, pp. 112-35. New York: Pantheon.
SIMON, L. R. AND J. C. STEPHENS. (1982) El Salvador Land Reform 1980-1981: Impact Audit, 2nd ed.
Boston: Oxfam America.
SKOCPOL, T. (1979) States and Social Revolution: A Comparative Analysis of France, Russia, and China. N
York: Cambridge University Press.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms
198 The Political Economy of Death Squads
SLOAN, J. W. (1983) Political Terrorism in Latin America. In The Politics of Terrorism, 2nd ed., edited
by M. Stohl, pp. 377-96. New York: Marcel Dekker.
SLOAN, J. W. (1984) State Repression and Enforcement Terrorisrn in Latin America. In The State as
Terrorist: The Dynamics of Governmental Violence and Repression, edited by M. Stohl and G. A. Lopez,
pp. 83-98. Westport, CT: Greenwood Press.
SNYDER, D. AND C. TILLY. (1972) Hardship and Collective Violen-ce in France, 1830 to 1960. American
Sociological Review 37:520-32.
STRASMA, J., P. GORE, J. NASH, AND R. I. RoCHIN. (1983) Agrarian Reformi in El Salvador. Washington,
DC: Agency for International Development.
TAYLOR, C. L. AND D. A. JODICE. (1983) World Handbook of Political and Social Indicators, 3rd ed. New
Haven, CT: Yale University Press.
TILLY, C. (1978) From Mobilization to Revolution. Reading, MA: Addison Wesley.
TULLOCK, G. (1971) The Paradox of Revolution. Public Choice 11:89-99.
U.S. DEPARTMENT OF STATE (1984) Report on the Situation in El Salvador (nimeo).
WALTON, J. (1984) Reluctant Rebels: Comparative Studies of Revolution and Underdevelopment. New York:
Columbia University Press.
WHEATON, P. (1980) Agrarian Reform in El Salvador: A Program of Rural Pacif cation. Washington, DC:
EPICA Task Force.
WOLPIN, M. (1986) State Terrorism and Repression in the Third World: Parameters and Prospects.
In Government Violence and Repression: An Agenda for Research, edited by M. Stohl and G. A. Lopez,
pp. 97-164. Westport, CT: Greenwood Press.
WORLD BANK (1984) World Bank Development Report, 1984. New York: Oxford University Press.
ZWICK, J. (1984) Militarism and Repression in the Phillipines. In The State as Terrorist: the Dynamics of
Governmental Violence and Repression, edited by M. Stohl and G. A. Lopez. Westport, CT:
Greenwood Press.
This content downloaded from
178.40.241.180 on Thu, 04 Jan 2024 15:27:40 +00:00
All use subject to https://about.jstor.org/terms