Caught in the middle, Colombia’s war on drugs and its effects on forest and people

Abstract

Coca plantations are the largest illegal agribusiness in the world, and Colombia is the world’s leading coca producer. Since 1994, the Colombian state, with the aid of the US, has waged a war on drugs based on air fumigation of coca plantations. This article evaluates the social and environmental impacts of this policy. We construct and analyse statistically for the first time a spatial database with social, economic, environmental, coca production and fumigation data for all 1125 municipalities of Colombia for the period 2001–2008. We complement statistical analysis with in situ observations and secondary literature review. We find that even if the questionable government claims that overall extent of coca plantations has been reduced were to be true, still coca activity has been diffused in the territory, with devastating environmental and social consequences. Biodiversity hotspot areas are being deforested, and local populations, especially Afro-Colombian communities, are being displaced from their territories. Our statistical analysis provides quantitative evidence to back up previous claims based on victims’ experience, single case-studies and ethnographic observation. We question the effectiveness of the fumigation policy and suggest that what is actually eradicated by the war on drugs is not coca, but humans and the forest.

Full text

Caught in the middle, Colombia’s war on drugs and its effects on forest and people
Alexander Rincón-Ruiz
a,
⇑
, Giorgos Kallis
a,b
a
ICTA, Autonomous University of Barcelona, Spain
b
ICREA, Autonomous University of Barcelona, Spain
article info
Article history:
Received 14 September 2011
Received in revised form 24 November 2012
Available online xxxx
Keywords:
Cocaine
Coca crops
War on drugs
Colombia
Fumigation
Deforestation
abstract
Coca plantations are the largest illegal agribusiness in the world, and Colombia is the world’s leading coca
producer. Since 1994, the Colombian state, with the aid of the US, has waged a war on drugs based on air
fumigation of coca plantations. This article evaluates the social and environmental impacts of this policy.
We construct and analyse statistically for the first time a spatial database with social, economic, environ-
mental, coca production and fumigation data for all 1125 municipalities of Colombia for the period 2001–
2008. We complement statistical analysis with in situ observations and secondary literature review. We
find that even if the questionable government claims that overall extent of coca plantations has been
reduced were to be true, still coca activity has been diffused in the territory, with devastating environ-
mental and social consequences. Biodiversity hotspot areas are being deforested, and local populations,
especially Afro-Colombian communities, are being displaced from their territories. Our statistical analysis
provides quantitative evidence to back up previous claims based on victims’ experience, single case-stud-
ies and ethnographic observation. We question the effectiveness of the fumigation policy and suggest
that what is actually eradicated by the war on drugs is not coca, but humans and the forest.
Ó2012 Elsevier Ltd. All rights reserved.
1. Introduction
Coca production is the largest illegal agribusiness in the world.
The global cocaine retail value is estimated at US$80–$100 billions,
equivalent to 0.15% of global GDP, and at the level of the annual
GDP of countries such as Iraq or Slovakia (UNODC, 2010). Cocaine,
consumed mostly in North America and Europe, is primarily pro-
duced in the Andean region. The plantation of coca crops for co-
caine is concentrated in three countries: Colombia, Peru and
Bolivia. Since 1997, Colombia is the main coca producer, account-
ing for more than 50% of total world production, with some
81,000 ha of coca cultivated and 450 metric tons of cocaine pro-
duced in 2008. Coca production in Colombia accounted for 623
millions of dollars of revenue in 2008, 0.3% of GDP and 3% of agri-
culture’s GDP (UNODC, 2008a). Unlike Peru and Bolivia, whose
anti-drug policy is based on manual eradication, Colombia is the
only country in the world to use aerial fumigation.
Colombia’s fumigation policy began cautiously in the end of the
1970s in order to fight marijuana plantations, but was extended in
1994 to the expanding cultivations of coca. Aerial fumigation
intensified and proliferated with the signing of ‘Plan Colombia’ in
1999 by Colombia and USA and the subsequent creation of the
‘‘Program of Eradication of Illicit Crops with Glyphosate’’ in 2000.
Plan Colombia has been celebrated as a great success in reducing
the total area of the country occupied by coca from 144,800 ha in
2001 to 81,000 ha in 2008 (UNODC, 2010), presumably liberating
local populations from the grip of the illegal business and its dev-
astating consequences. The Colombian government has also her-
alded the environmental benefits of the war on drugs; the coarse
hypothesis behind such statements is that coca has negative envi-
ronmental effects and any policy that reduces must by definition
have positive ones (Álvarez, 2007; Bernal, 2007). Yet, other
researchers argue that fumigation goes hand-in-hand with defor-
estation and environmental degradation (Ávila et al., 2007; Vargas,
2004; Walsh et al., 2008a), negative health effects (Ávila et al.,
2007; IDEA, 2005; Nivia, 2001a), and social impacts, including
forced displacement, disproportionately falling on Afro-Colombian
groups and low-income population (Defensoria del Pueblo, 2007;
OAIPC, 2010).
How does aerial fumigation affect coca production, the liveli-
hood and settlement patterns of human populations and the state
of ecosystems? This is an important question if one wants to know
how and why anti-drug interventions ‘‘from a safe distance’’, such
as aerial fumigation, may produce counterproductive results at the
ground that undermine their proclaimed intentions. We provide
new evidence at a finer spatial scale than ever before, which sub-
stantiates the claim that aerial fumigation has negative social
and environmental effects, and we then explain why this is the
case. We argue that the aerial fumigation policy is ill-suited for
the socio-environmental interdependencies present at the complex
0016-7185/$ - see front matter Ó2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.geoforum.2012.12.009
⇑
Corresponding author. Present address: Transv 80 No. 80-20, Bloque M, Apto
501, Bogotá, Colombia.
E-mail address: Alexander.risvid@gmail.com (A. Rincón-Ruiz).
Geoforum xxx (2013) xxx–xxx
Contents lists available at SciVerse ScienceDirect
Geoforum
journal homepage: www.elsevier.com/locate/geoforum
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

socio-ecosystem of the Colombian forests, where most of coca pro-
duction is concentrated. In this, we position our research as a case
study of the broader thesis about the failure of State improvement
schemes based on schematic visions that do violence to complex
socio-ecosystem interdependencies (Norgaard, 1994; Scott, 1998).
We are not the first ones to study the social and ecological im-
pacts of cocaine production or the war on drugs. There is a long lit-
erature on the failures of the US-driven war on drugs in Latin
America and particularly the negative effects of the forced dis-
placement of cultivations (Guáqueta, 2005), compared to more
structural solutions offering employment alternatives to producers
(Guridi, 2002). The failure of forced policies to make peasants to
leave cocaine production has been documented for the case of Bo-
livia (Guridi, 2002) and Peru (Cabieses, 1999). A grand part of the
literature on aerial fumigation evaluates direct impacts, most nota-
bly on health (e.g. Ávila et al., 2007; Hewitt et al., 2009; IDEA,
2005; Nivia, 2001a; Solomon et al., 2005a, 2005b), and the environ-
ment and agriculture (Ávila et al., 2007; Eslava et al., 2007; Mes-
sina and Delamater, 2006; Nivia, 2001a, 2001b; Varona et al.,
2009). Concerning indirect effects, there are studies, which have
looked at the displacement of peasants and legal crops in Bolivia
and Peru (Bradley and Millington, 2008). For Colombia there is
anecdotal evidence that the fumigations destroy the revenue base
of the peasant economy and displace both coca production and
peasants to new areas (Vargas, 2004). Scrutinizing the official data
at the national level, González (2006) finds inconsistencies that
raise questions about the proclaimed effectiveness of the eradica-
tion policy. Also an inter-temporal econometric analysis at the na-
tional level by Moreno-Sanchez et al. (2003) shows that the
cultivation area of coca in Colombia has increased as eradication
efforts have intensified, because farmers compensate for eradica-
tion by cultivating the crop more extensively. This pattern is con-
firmed by a statistical analysis at the level of the 32 sub-national
departments of Colombia by Dion and Russler (2008), who find
that fumigation displaces, but does not eradicate, coca production.
This displacement effect has been called in the drugs literature the
‘‘balloon effect’’ (Laffiteau, 2010; UNODC, 2008b) and attributed to
an inelastic demand, that will be satisfied in one way or the other
by the producing regions.
Whereas this literature offers many useful reference and entry
points, there are several gaps if one wants to get a more accurate
picture of how aerial fumigations affect production, settlement
and ecological patterns in Colombia. First, the national or regional
scale analyses hide important shifts and effects at lower spatial
scales, where complex interdependencies are at play. We provide
here for the first time data coverage on fumigations and coca cul-
tivation down to the municipal level (1125 municipalities). Second,
much of the interest until now has been on production patterns,
and the effects of fumigation on the acreage and location of coca
cultivations. Despite claims for the dislocation of people or the un-
even impacts of the policy on the basis of race or class, no other
study to our knowledge has examined such effects rigorously.
We cover a greater number of variables per year (also for a more
recent period, 2001–2008, than other studies) identifying new
associations between coca cultivation and its social impacts, espe-
cially dislocation, which has not been evaluated before. Third, con-
cerning environmental impacts, whereas Dávalos et al. (2011)
before us also looked at the complex relations between illicit crops
and deforestation in Colombia at the municipal level, we extend his
analysis by using a different methodology on the basis of a map-
ping of ecosystems which permitted us to evaluate land-use
changes at the ecosystem level (see methods below). Fourth, and
most importantly, this is the first study that attempts an integrated
and multi-dimensional analysis of both direct and indirect effects
of fumigations at the most refined scale possible. Whereas other
studies before focused either on health, environmental or produc-
tion effects, we examine all these together. This gives us the oppor-
tunity to offer a more accurate understanding of the multi-faceted
effects of fumigation on people and the territory, and through it
draw wider claims on how improvement schemes and anti-drug
policies from a distance produce negative effects in complex so-
cio-ecosystems such as those of Colombia.
In summary, our main claim is that the fumigation policy is fail-
ing in Colombia, because it does not eradicate, but diffuses coca
production, shifting it to forests of ecological importance and to
areas inhabited by low-income, especially Afro-Colombian and
indigenous communities, which as a result are increasingly dis-
placed. The broader significance of our claim is the confirmation
of a broader pattern whereby government ‘‘improvement’’ policies
imagined from a distance fail miserably in the face of complex local
socio-ecological interdependencies.
Section 2presents the methods used to generate the evidence
for this claim and the new data mobilized or constructed for this
analysis. We employ a novel spatial approach to respond to the
above questions demonstrating the importance and contribution
of geographical analysis. In particular, we analyze statistically a
newly-compiled geographical and longitudinal dataset of aerial
fumigation, coca production and various socio-economic and
demographic variables at the municipal level, complementing it
with qualitative information from interviews and secondary docu-
ments, as well as in situ assessments of the impacts of aerial
fumigation.
Section 3presents the empirical evidence that supports our
claim. We find that:
1. Fumigation has not eradicated, but displaced coca production to
other regions. Such a ‘‘Balloon effect’’ has been noted by others
for manual eradication and at the macro-regional level (Bradley
and Millington, 2008; Laffiteau, 2010; The Economist, 2001;
UNODC, 2008a). Our intra-national study finds in addition that
aerial fumigation not only displaces, but actually diffuses the
production of coca in the territory, and that the effect of fumi-
gation is temporary, as production often returns after a while.
This creates a negative spiral of fumigation and cultivation that
affects more and more territories and people.
2. Fumigation in Colombia displaces production to areas of pri-
mary forest of great environmental significance.
3. Fumigation causes negative health impacts but these are con-
tested and hard to verify. The level of complaints launched by
local communities suggests that fumigations do impact nega-
tively local livelihoods.
4. Fumigation is associated with increased human displacement.
5. Less developed communities, including indigenous and Afro-
Colombian communities, are disproportionately impacted by
fumigation and coca displacement. There is no evidence how-
ever to suggest discriminatory fumigation by the authorities.
Section 4discusses the main findings of our research and at-
tempts to explain why is the policy failing. We argue that the pol-
icy overlooks complex interdependencies at the local level, and in
particular does not account for the lack of alternative sources of
income, as well as the particular socio-ecological features of the
coca economy, which make it selectively shift to areas of primary
forest and low development. Section 5reinstates our main con-
clusion and draws its policy implications: the Colombian anti-
drugs policy of aerial fumigation has caused a displacement and
diffusion of coca cultivation in the territory, impacting socially
and ecologically vulnerable areas and expanding the war on drugs
to new areas, affecting the livelihoods of more people. We add
our voice to those who argue that the US and Colombian govern-
ments should reconsider thus policy and shift resources instead
to policies that curb demand for drugs at its source or that
2A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

provide meaningful livelihood alternatives to local populations at
the production regions.
2. Data sources and methods
The research findings discussed in this article were gathered
following a multi-evidentiary strategy consisting of four compo-
nents. First, an extensive literature review was conducted of all
peer-reviewed and government publications concerning antidrug
policies and the social, economic, environmental and political as-
pects of coca plantations in Colombia. This preliminary phase of
the research benefited by discussions with experts of the Inte-
grated Illicit Crops Monitoring System (SIMCI in Spanish), the Na-
tional Office of Narcotics of Colombia (DNE in Spanish), the
Ministry of Defense, and the NGOs Transnational Institute (TNI),
the Arcoiris Foundation and Acción Andina, as well as other
researchers from Colombian Universities.
Second, and at the heart of the research findings is a database at
the municipal level, the first of its kind, including data for all 1125
municipalities of Colombia with respect to social, economic, envi-
ronmental and institutional features, as well as information on the
extent of coca production and aerial fumigation. We used existing
data but we are the first ones to construct this comprehensive
database for the purposes of this research. The data for municipal
coca production (2001–2008) are taken from the official data pro-
duced by SIMCI, whereas for fumigation we used municipal level
data from the DNE, which has been reported by UNODC (UNODC,
2008b, 2009). Such official data is highly suspicious and controver-
sial, as it is often invoked to support the success of the govern-
ment’s anti-drug policy. Data from CIA’s Crime and Narcotics
Center (CNC) suggests a lesser impact for the fumigation program
(GAO, 2008). However the CNC data is available only at the na-
tional and not the municipal level. Unfortunately, the SIMCI/UN-
ODC data is the only source available at the municipal level and
with this we had to work. Since we are interested more on longitu-
dinal cross-sectional differences the possible biases in the absolute
levels are less crucial for our case. Furthermore, we will show that
even with this government data that may have been manipulated
to paint a better picture, a disaggregated analysis at the municipal
level shows that the fumigation policy is failing.
Table 1 details the variables used in the database, their defini-
tion and the sources of the data. This list was constructed on the
basis of the literature review and our initial research questions
after consultation with experts and taking into account the avail-
ability of data at the municipal level. On the basis of this database,
we tested (spatial association and correlation analysis) hypotheses
concerning the socio-economic characteristics of the expanding
coca frontier, and the relationships between aerial fumigation,
the spatial distribution of coca production, environmental effects
and population displacement. The statistical significance of the
correlations was carried out using 10,000 permutations (Anselin,
2005; Anselin et al., 2002). Analysis of spatial information and
the statistical tools were done using the software SPSS, Geoda
and ArcGIS.
Third, we analyzed the deforestation and ecosystem impacts of
coca cultivations by overlapping information about land coverage
from the national map of ecosystems in 2000 in shape format
(IDEAM et al., 2007) with geospatial information about coca culti-
vations from SIMCI’s maps for the 2001–2008 period. A total of 154
ecosystems were grouped into three principal and 32 sub-biomes.
Sixteen types of land cover were grouped into eight natural classes
(natural continental waters, shrubs, natural forests, grasslands,
grasses and coastal bushes, continental hydrophytes, coastal lakes
and estuaries, and secondary vegetation) and eight transformed
classes (heterogeneous agricultural areas, largely alternated areas
– agro industrial crops, urban areas, artificial continental water,
forest plantation, annual or transitory cultivations, (semi-) perma-
nent crops and grasses). Our analysis was based in the natural cov-
er and ecosystems estimated in 2000 (IDEAM et al., 2007) and the
expansion of coca crops during 2001–2008 (Coca census 2001–
2008). We estimated the share of natural ecosystem transformed
at the municipality level. Where relevant, we also used maps
(shape format) divided according to the legal status of Colombian
territories into: Collective territories of Afro-Colombian communi-
ties, indigenous territories, forest reserve, natural national parks
and subtracted area from the forest reserve (subtracted area means
the area that ceases to be in forest reserve) (Accion Social, 2009),
allowing in this way to document differentiated changes in cultiva-
tion and fumigation in each of these territories. The results were
complemented with data from the United Nations Office on Drugs
and Crime. Dávalos et al. (2011) before we also looked at the com-
plex relations between illicit crops and deforestation in Colombia.
He used however remote sensing images, and where heavy clouds
did not allow a clear picture, the data was classified as missing.
This was a problem especially for the Pacific region and likely to
lead to an underestimation of deforestation. Our analysis which
overlaps municipal and ecosystem data does not suffer from this
problem, though we might have overestimated deforestation due
to coca in cases where what was classified as natural forest in
2000 was already transformed to agriculture and another type of
land use and only after that turned into coca cultivation.
Fourth, a rapid assessment was conducted in the department of
Nariño (Map 1), one of the areas most affected by the expansion of
illicit coca crops in the 2000s. Between 2001 and 2008 the depart-
ment of Nariño had an increase of 162% of the area cultivated with
coca crops and it has been a growing target of fumigations. During
a period of 1 month of field-work, the first author interviewed a to-
tal of 18 people, 10 of them representing those to different degrees
of affected by or involved in coca production (including indigenous
leaders, Interior departmental advisers, members of the health
department secretary, members of the national police) and eight
people from local communities. Additional information on the im-
pacts of fumigation on the local population was collected through a
review of governmental documents from the Ombudsman, the lo-
cal police, the health secretary and the hospitals of Nariño and
through direct conversations with local people, particularly with
five peasants from the coca-growing areas of the department of
Nariño and leaders of the indigenous community indigenous Awá
that inhabit the southwest of Colombia, one of the communities
most affected by armed conflict in Colombia and the war on drugs
(Saavedra, 2009).
Below we report on the key findings of our analysis structured
around each of the five sub-claims identified in the introduction.
The core findings are based on the municipal statistical research
and the ecosystem impact assessment, grounded where relevant
with material from secondary literature research and the Nariño
visit.
3. Evidence
3.1. The effects of aerial fumigation on coca production
According to the government data that we used, between 2000
(commencement of Plan Colombia) and 2008, coca cultivations at a
national level have been reduced from an area of 163,000 ha to
half, i.e. 81,000 ha. However, unlike what the government claims,
CIA’s Crime and Narcotics Center (CNC) has reported that between
2001 and 2007 the area cultivated with coca has remained stable at
around 170,000 ha (GAO, 2008). Rather than such national aggre-
gates, we are more interested in the territorial distribution and im-
A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx 3
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

Table 1
Description of variables included in the analysis.
Name Variable name Years Source Unit Description
PRD Primary road density 2005 Road map (shape) by Geographic Institute
Agustín Codazzi – estimation by the authors
M/ha Meters of primary roads per hectare
FDP Forced Displacement of
Population
2001–
2008
Presidency of the Republic of Colombia –
Presidential Agency for Social Action and
International Cooperation
Number of displaced
people
Number of forced displaced people by
violence and conflict. This information is
taken from the National System of attention
to Displaced People (‘‘Sistema Nacional de
Atención Integral a la Población Desplazada’’)
RUBN Rural Unsatisfied Basic
Needs Index
2005 National Administrative Department of
Statistics – (DANE Spanish acronym for
‘‘Departamento Administrativo Nacional de
Estadística’’)
Index from 0 to 100
(from ‘no basic need
satisfied’ to completely
satisfied)
The index is determined through five
indicators: adequacy of housing, degree of
household overcrowding, adequacy of basic
household services, degree of economic
independence of the household; household
with children at school-age which are not
attending school
IMD Index of municipal
development
2001–
2008
National Planning Department of Colombia
(DNP – Spanish acronym for ‘‘Departamento
Nacional de Planeación’’) and Direction of
territorial sustainable development (DDTS –
Spanish acronymun for ‘‘Dirección de
Desarrollo Territorial Sostenible’’)
Index from 0 to 100
(where 0 means low
municipal
development)
Synthetically measuring the performance of
municipalities in social and financial indices,
including: % of households with water
supply, % of households with sewage, %
households with energy services, % of people
without unsatisfied basic needs in urban area
VAIA Number of violent acts
by illegal armed groups
2001–
2006
Los Andes University Bogotá and Ministery of
defense Colombia
Number of violent acts Number of violent acts by Illegal Armed
Groups (FARC, AUC, ELN) per Municipality,
Including terroristic acts, assaults, attacks,
roadblocks, ambushes, harassment, attacks
on population
MR Murder Rate by Illegal
armed groups
2001–
2008
Colombian National Police – Estimation by
the Authors
Number of homicides Number of homicides per 100,000
inhabitants committed by illegal armed
groups (FARC, AUC, ELN). Homicides
committed by common crime are not taken
into account
MIAG Number of murders by
Illegal armed groups
2001–
2008
Colombian National Police Number of homicides
by Illegal armed groups
Homicides committed by common crime are
not taken into account
PPF Percentage of primary
forest area
2000 Colombia Ecosystem map (shape) – Institute
of Hydrology, Meteorology and
Environmental Studies of Colombia (IDEAM –
Spanish acronymun for ‘‘Instituto de
Hidrología, Meteorología y Estudios
Ambientales de Colombia) – Estimation by
the authors
Percentage Hectares of primary forest as percentage of
the total area of the municipality
PCOCA Percentage of coca area 2001–
2008
Coca maps (shape) – Integrated Illicit Crops
Monitoring System (SIMCI: Spanish acronym
for ‘‘Sistema Integrado de cultivos ilicitos’’)
Percentage Hectares of cultivated coca as percentage of
the total area of the municipality
NCOMP Number of complaints
to the Ombudsman by
citizens concerning
aerial spraying
2001–
2008
Local and National Ombudsman’s Office Number of complaints Number of complaints to the Ombudsman by
citizens concerning aerial fumigation
AF Aerial fumigation 2001–
2008
National Direction of Narcotics (DNE –
Spanish acronym for ‘‘Dirección Nacional de
Estupefacientes’’
Number of hectares Number of fumigated hectares per
municipality
CA Coca area 2001–
2008
Coca maps (shape) – Integrated Illicit Crops
Monitoring System (SIMCI: Spanish acronym
for ‘‘Sistema Integrado de cultivos ilicitos’’)
Number of hectares Ha of area cultivated with coca in the
municipality
RPOP Rural population 2005 National Administrative Department of
Statistics – DANE, estimation by the authors
Number of persons Number of persons living in the rural zones of
each municipality
AF01-
08
Area fumigated
between 2001 and 2008
Total
2001–
2008
National Direction of Narcotics (DNE –
Spanish acronym for ‘‘Dirección Nacional de
Estupefacientes’’
Number of hectares Total area fumigated between 2001 and 2008
SDC01-
08
Indicator of variation in
coca cultivation
Total
2001–
2008
Coca maps (shape) – Integrated Illicit Crops
Monitoring System (SIMCI: Spanish acronym
for ‘‘Sistema Integrado de cultivos ilicitos’’)
Number of hectares Standard deviation of the area of coca
cultivated from the 2001–200
ABC
and
AIT
Area of the municipality
belonging to Indigenous
territories and black
communities
Maps of black communities and Indigenous
territories (shape) – Geographic Institute
Agustín Codazzi
Number of hectares Number of hectares under legal status of
indigenous territory (AIT) or under title of
black communities (ABC)
AECO Area of natural cover
and natural ecosystems
at municipal level
2000 Colombia Ecosystem map (shape) – Institute
of Hydrology, Meteorology and
Environmental Studies of Colombia (IDEAM –
Spanish acronymun for ‘‘Instituto de
Hidrología, Meteorología y Estudios
Ambientales de Colombia) – Estimation by
the authors
Number of hectares Hectares of natural cover of a certain
ecosystem type within the municipality
4A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

pacts of coca production. We will show that even with the use of
the official government data that should have been favorable to
the effects of fumigation, it appears that fumigation policy has
not eradicated, but diffused production through the territory, espe-
cially in socially and ecologically vulnerable zones, affecting the
lives of more people.
Before the start of fumigations, most coca cultivations were
concentrated in the northern region of Colombia, at the Colombian
Amazon (Map 1). In 2000 the three departments of the Amazon re-
gion alone (Putumayo, Guaviare and Meta), out of the total of 32
departments in which Colombia is divided, accounted for 58% of
the national production of coca. And it was there that 56% of the
fumigations in 2000 concentrated. However, following the fumiga-
tion policy coca production was dispersed to new regions princi-
pally in the Pacific region (Nariño and Chocó departments). We
demonstrate this ‘‘balloon effect’’ (Paredes and Correa, 2007)in
Map 1. Departments and natural (continental) regions of Colombia.
Table 1 (continued)
Name Variable name Years Source Unit Description
CALS Area cultivated with
coca in each of the main
territorial divisions
according to legal status
2000–
2008
Coca maps (shape) – Integrated Illicit Crops
Monitoring System (SIMCI: Spanish acronym
for ‘‘Sistema Integrado de cultivos ilicitos’’)
and maps of black communities territories,
Indigenous territories, National Natural
parks, Subtracted area from forest reserve –
Geographic Institute Agustín Codazzi
Number of hectares
with coca in each of the
legal status of the
territory selected, by
year
Number of hecatares with coca crops in Afro-
Colombian territories, indigenous territories,
forest reserve, subtracted area from the forest
reserve and national natural parks
A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx 5
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

four ways: (1) graphically (Maps 2 and 3), (2) statistically (Table 2),
(3) with a cross-sectional municipal analysis (Fig. 1 and Map 4) and
(4) in Section 5, with a case-study (Fig. 2). Since our analysis fo-
cuses on the fumigation policy, we do not include manual eradica-
tion. This makes it likely that we are overestimating the effects of
fumigation in causing the ‘‘balloon effect’’. Note however that
fumigations have affected a greater part of the territory than man-
ual eradication (172,000 ha vs. 41,000 ha in 2006, and 13,300 vs.
95,000 ha in 2008; (UNODC, 2006b, 2008b, 2009).
Maps 2 and 3, which are derived from our municipal database
(Table 1), compare changes in area fumigated (2002–2003 and
2006–2007) with changes in the area occupied by coca crops in
the subsequent year (2003–2004 and 2007–2008). We find that
in the municipalities where fumigations increased (black color –
Maps 2a and 3a), the extent of land covered by coca area declined
in the subsequent period (gray color – Maps 2b and 3b). However
there was an increase in the extent of the cultivated area in the
municipalities neighboring the areas fumigated (black color –
Maps 2b and 3b). Therefore the shrinking of coca production in
one part came at the expense of expanding in another. In other
words, even if overall coca cultivation were to be decreasing (as
the official data claims), it is nonetheless diffusing in the territory.
In addition to the visual representations of Maps 2 and 3,we
test the proposition of a balloon effect by analyzing statistically
the spatial association between the fumigated area by municipality
for the year nand the area under coca in the bordering municipal-
ities in the year n+ 1. The estimation was done using the multivar-
iate Moran’s I coefficient, an indicator of spatial correlation
(Anselin et al., 2002). We found a positive association between
the area fumigated in a municipality in year n, and the area under
Map 2. (a) Change fumigated area 2002–2003 and (b) change coca area 2003–2004.
6A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

coca plantation in bordering municipalities in the year n+1 (Ta-
ble 2). In other words, there is a direct association between aerial
fumigation of a municipality and subsequent increase of coca pro-
duction in a nearby one. This finding of a ‘‘balloon effect’’ is not un-
ique to aerial fumigation or Colombia, but has also been observed
in the cases of manual eradication programs in Bolivia and Peru
(Bradley and Millington, 2008; Salisbury and Fagan, 2011).
Correlation is not causation. First, while we can ascertain
expansion in neighboring areas after fumigation, we do not have
data to document actual displacement. However it is not far-
fetched to hypothesize that expansion in neighboring areas is the
result of displacement, as corroborated by many of our interviews.
Second, it can be that the departure of coca production from one
area is the result of other factors, such as a coca production satura-
tion effect, increased local conflict or changes in labor conditions.
The hypothesis here would be that the areas experiencing these
changes would be the ones with more intense development of coca
production and hence the ones more likely to be fumigated. Fumi-
gation therefore would correlate with expansion in neighboring
areas (through displacement) but with no causal relation. Again,
such factors were not identified as important either in the litera-
ture review or the interviews, but of course these alone cannot rule
them out. Given however that we compare fumigation at time n
with neighboring production at time n+ 1, we find it less plausible
that such a clear ‘‘cat and mouse’’ dynamic would emerge simply
by fumigation following year after year saturated or conflicting
areas, i.e. zones where production was already at the point of mov-
ing to a neighboring zone. What we cannot rule out however is the
possibility that fumigation acted in concert with some of these or
other factors to cause the migration of production; for example it
Map 2. (continued)
A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx 7
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

can be that the areas fumigated were also the ones where there
were other forms of confrontation or violence causing displace-
ment of cultivation, and so displacement was not the effect of
fumigation alone. Further research is necessary to isolate the cau-
sal contribution of each factor and their interactions.
Contributing to the understanding of the balloon effect, we
find that it is not so much the case that the geography of coca
production is shifted from one area to the next, but that produc-
tion is diffused across municipalities. The number of municipali-
ties with coca plantations within their territories increased from
164 in 2001 to 202 in 2008 (Fig. 1). All the new (to coca) munic-
ipalities have plantations that exceeded 100 ha, suggesting that
there is extensive cultivation going on, and that this is not a min-
or side-effect. Furthermore, while by 2001, only 85 municipalities
had coca cultivations exceeding 100 ha, by 2008 the number of
municipalities with such extensive cultivations had increased to
106.
This expansion of the cultivation to new territories, has pro-
duced a subsequent increase of aerial fumigation in the new terri-
tories. In turn, and in a vicious cycle mode, this appears to have
caused new displacement of coca cultivations. Interestingly in
some cases, this has caused a return to areas previously fumigated.
A consequence is that the number of municipalities fumigated in-
creased accordingly. Fig. 1 shows that in 2001, 50 municipalities
were fumigated, but this number increased to 97 in 2008. Rather
than an intensification of the policy, this can be seen as evidence
of its failure to eradicate coca production in the targeted areas.
The result is the geographical expansion of both coca and the
war on drugs frontier. Map 4 illustrates spatially the persistence
of coca production in the territory by indicating the number of
Map 3. (a) Change fumigated area 2006–2007 and (b) change coca area 2007–2008.
8A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

years that each municipality has experienced coca cultivation
(minimum 1, maximum 8): many municipalities have seen a con-
tinuous production of coca in their territory. In none of the munic-
ipalities fumigated during the study period were crops eliminated
completely. In fact 81 of the 143 fumigated municipalities show an
increasing trend of coca crops.
1
Our data suggests also that coca plantations go away during
fumigation, but come back after the territory stops being fumi-
gated. The evidence is the statistically significant correlation
2
be-
tween the size of the area fumigated between 2001 and 2008 and
an indicator of variation in coca cultivation, measured as the stan-
Map 3. (continued)
Table 2
Moran’s I – spatial correlation of area fumigated per municipality in year nwith area
under cultivation in the same municipality in year n+1/
*
Significant correlation
(0.05).
Area sprayed by
municipality in the year n
Coca area in the neighborhood
municipalities in year n+1
Moran’s I
2001 2002 0.105
*
2002 2003 0.102
*
2003 2004 0.217
*
2004 2005 0.192
*
2005 2006 0.121
*
2006 2007 0.235
*
2007 2008 0.179
*
*
Correlation is significant at the 0.05 level (2-tailed).
1
The trend was estimated by lineal regression: coca area = f(time).
2
Pearson Correlation 0.728/significant at the 0.01 level.
A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx 9
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

dard deviation of the area of coca cultivated from the 2001–2008. In
other words, the more a municipality is sprayed the more ‘‘back and
forth’’ of coca production it experiences.
In conclusion: fumigation is associated with expansion of pro-
duction to other areas, which can be taken as evidence of displace-
ment. Such displacement diffuses the problem in the territory,
with coca production affecting more areas, and presumably more
people. Furthermore, production returns, even if at a lower level,
to the areas from which it was supposedly eradicated by aerial
fumigation. Our local level analysis confirms national and regional
level studies, which have claimed that fumigation is not an effec-
tive approach in eradicating coca production. Furthermore, in addi-
tion to the displacement pattern identified by these studies, we
highlight a broader pattern of diffusion.
Map 4. Persistence of coca cultivations. Number of years that the municipalities had coca cultivations in the 2001–2008 period.
Fig. 1. Growth in municipalities with coca cultivations and in municipalities being
fumigated (2001–2008).
10 A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

3.2. Fumigation and deforestation
The official government discourse is one that links fumigation
and sustainability, in the sense that more fumigation means less
coca production and hence less deforestation. Instead we find that
the territorial diffusion of coca activities leads to continued and
expanding deforestation, and we attempt to characterize the envi-
ronmental characteristics of the new areas of the forest affected by
coca and fumigations.
On the basis of the crossing of the ecosystem map of Colombia
in 2000 and the coca maps (2001–2008) in shape format, we esti-
mated the extent of natural ecosystems (forest) affected by coca
diffusion year by year, based upon data on existing ecosystems
and natural cover for 2000. Table 3 presents the extents of natural
forest affected by coca during the period 2001–2008. At the begin-
ning of the decade, the Amazon and Orinoco region ecosystems
were those most affected, but impact upon them has decreased
over time: 71,920 ha of coca in 2001 were occupying natural forest
in 2000, but in 2008 this area decreased to 22,270 ha. In contrast in
2008 the natural forests of the Caribbean region and especially of
the Pacific region, which is considered one of the biodiversity hot-
spots of the world, showed significant increases in coca conversion.
1982 ha that were natural forests in 2000 in the Pacific became
coca cultivations by 2001, and by 2008 this area had increased to
8166 ha.
Even if fumigations had reduced the total area of coca cultiva-
tions in Colombia, they have increased deforestation since the bal-
loon effect means that new areas of primary forest are deforested
as the war pushes the frontier to new territories. Furthermore,
the primary forest that is lost to coca plantations is irreplaceable.
Even though fumigation may displace coca, the previous state of
the forest is not recoverable. According to the data of SIMCI the dis-
placement of coca crops to new areas has generated a deforestation
of primary forest of 110,026 ha between 2001 and 2008.
Our research at Nariño confirms this pattern at the level of a
department. About 40% of the area under coca in Nariño in 2008
was natural forest in 2000. Nariño has 10 eco-regions, five of them
occupied by coca cultivations; between 2003 and 2008 18%, of the
total of deforestation caused by coca in Colombia took place in
Nariño. According to SIMCI data, 13,000 ha of natural forest has
been converted to coca in Nariño in 2003–2008. 22% of the total
area cultivated with coca in 2008 was tropical rain forest in
2000. 8% was riparian forest.
Whereas there is a clear link between coca production and
deforestation, this should not be read as an argument in favor of
fumigation (and the common replacement by oil palm planta-
tions). Fumigation diffuses and expands deforestation, while an
industrial tree plantation is no substitute for primary forest loss.
Fig. 2 illustrates the persistence of coca crops and the vicious cycle
of fumigation, that characterizes Nariño. An increase in fumigation
is associated with a decrease in deforestation (without deforesta-
tion disappearing altogether). Nevertheless, when fumigation de-
creases, the deforestation due to coca cultivation starts
increasing again. Interviews with peasants from the area confirm
that coca cultivations that had moved to nearby territories return
to where they had left from. Rather than an argument for a contin-
uous or intensified fumigation, the point here is that fumigation
causes merely spatial displacement and additional destruction,
and is ultimately ineffective as a strategy of reduction of coca cul-
tivation. An exception to this pattern of displacement are the last
years depicted in Fig. 2 (2007–2008), where despite the increase
in fumigation there is an increase also in coca-driven deforestation.
A possible explanation is that coca production is no longer dis-
placed to other department but to new municipalities within
Nariño itself.
In conclusion: fumigation does not reduce deforestation. It dis-
places production from areas where the primary forest is already
lost biodiversity hotspots where additional primary forest is
destroyed.
3.3. The effects of fumigation on health and agriculture
The impacts of aerial fumigation on the health of local popula-
tion and the legal crops are intensely debated in Colombia. Some
researchers cannot find statistically significant evidence given that
local people usually have contact with many other toxic substances
that can cause health effects (such as pesticides and herbicides
used on crops) (Varona et al., 2009). Plan Colombia fumigation con-
tinues on this basis, the anti-drug police spraying the territory with
a mixture of the herbicide ‘Roundup Ultra’, the proprietary name of
a Monsanto product that contains glyphosate and the surfactant
polyethoxylated tallowamine (POEA) and ‘Cosmo-flux 411F’. In
Colombia, pesticides containing glyphosate such as ‘Roundup’ are
registered under the toxicological class IV (slightly toxic) (Nivia,
2001a). The US Environment Protection Agency (EPA) registers gly-
phosate in Toxicity Category III (with Category I being the most
toxic and IV the least). In terms of carcinogenicity it has been
placed in Category E, i.e. with evidence of non-carcinogenicity in
humans (Sherret, 2005).
Various government sources argue that fumigations have not
exceeded health or environmental norms and have not had nega-
tive effects, an argument that has been used in favor of the contin-
uation of the policy (United States Department of State, 2010). The
Organization of American States (OAS) also published a study in
2005 noting that the chemicals used to aerially eradicate coca
did not pose significant risks to humans and most wildlife
Table 3
Natural forest in 2000 converted into coca cultivations between 2001 and 2008.
Natural forest in 2000 converted into coca cultivations between 2001 and 2008 2001 2002 2003 2004 2005 2006 2007 2008
Amazon and Orinoquian Region 71,910 61,536 39,027 33,616 39,684 33,120 35,613 22,270
Pacific Region 1982 4640 4526 3821 3792 2788 5839 8166
Caribe Region 3797 3352 3918 3258 4039 2336 6125 5712
Fig. 2. Aerial fumigation (ha sprayed) and deforestation into coca crops in Nariño/
Colombia (2001–2008).
A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx 11
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

(Solomon et al., 2005a, 2005b). Similar is the conclusion of the
International Narcotics Control Strategy Reports financed by OAS,
which also added that the damage from drug crop production
and processing far outweighs the negligible risk from exposure to
glyphosate due to coca or poppy spraying (Bernal et al., 2009;
Bolognesi et al., 2009; Brain and Solomon, 2009; Hewitt et al.,
2009).
On the other hand other researchers and NGOs working in the
region have provided evidence on the negative impacts of the
fumigations on the population’s health and the environment (Ávila
et al., 2007; Eslava et al., 2007; IDEA, 2005; Nivia, 2001a, 2001b;
Walsh et al., 2008a). In several countries ‘‘Roundup’’ was among
the first pesticides that was reported to cause human poisoning,
and effects reported after exposure include nausea, dizziness,
respiratory problems, increased blood pressure and allergic reac-
tions (Nivia, 2001a). And according to the Farm Chemicals Hand-
book (1990), it is not recommended to use glyphosate via aerial
application due to environmental effects. Colombia’s fumigations
with glyphosate have reportedly generated environmental prob-
lems in neighboring Ecuador, as verified by government institu-
tions and scientists (Ávila et al., 2007). Glyphosate has been
developed to be applied directly to plant leaves and not through
the air (Haney et al., 2000). Similarly (Sherret, 2005) argues that
the problem with aerial fumigation in Colombia is not the toxico-
logical profile of glyphosate per se, but the open violation of the
norms for its application whether through ignorance or intent.
Local communities insist on their experience of the negative im-
pacts of the aerial fumigation, and have denounced the crime per-
petrated against them with national-level protests and
communiqués (Defensoria del Pueblo, 2007, 2009; La Nació-
n.com.co, 2010; OAIPC, 2010; Oslender, 2010). Studies also of the
Institute of Environmental Studies of the National University of
Colombia question the results of the OAS studies (Solomon et al.,
2005a, 2005b). Of course, such struggle over scientific uncertainty
and complexity in an environmental health issue is not unique to
fumigation in Colombia, but characteristic of many other environ-
mental controversies. The government continues to deny any link
between fumigations and adverse health effects, whereas NGOs
call for an application of the ‘‘precautionary principle’’, i.e. a pre-
cautionary banning of fumigation given its unknown and disput-
able, yet highly risky health effects, but in vain (Kennedy and
Stefani, 2009).
Health effects are very difficult to verify and this is beyond the
purpose of our study. Still, our data confirms a significant correla-
tion between area fumigated and the number of complaints sub-
mitted to the Ombudsman and local authorities, a rough
indicator of impacts on livelihoods (Table 8). At the beginning of
the decade, aerial fumigation was especially concentrated in the
department of Putumayo, which received 47% of all aerial fumiga-
tion in Colombia (of 224,516 ha sprayed in Colombia between 2001
and 2002, 104,397 ha were in Putumayo). According to our analy-
sis of archives between mid 2001 and mid-2002 the Ombudsman
received 318 complaints concerning health impacts or the loss of
(legal) crops from aerial fumigation in the three municipalities of
Putumayo where 6076 families live. A 2002 study conducted by
the health department of Putumayo on the impact of fumigations
in community territories showed that 4883 (81.5%) of the 5929
people that had filed complaints reported health problems when
interrogated by municipal officials. Furthermore, a 2007 report of
the Ombudsman’s Office in Putumayo, based on direct observation
of the people recovered in the local hospitals, revealed that vomit-
ing and diarrhea, headaches and respiratory problems were com-
mon symptoms of those exposed to fumigation. O’Shaughnessy
and Branford (2005) documents also negative health effects in
the poorer segments of the population, based on field-work in
Putumayo. After the increase of claims between 2001 and 2003,
there was a decreasing tendency of claims starting in 2004. Our
interviews suggest that farmers stopped reporting due to the lack
of response by PECIG, the authority overseeing the program of
eradication. People no longer believed that the state will attend
their claims and complaints. And this may explain why a total of
2559 claims in 2003 were reduced to 781 in 2008. Our field-work
and our collection of interviews and photographic material con-
vinced us of the actual impact of fumigations on legal crops, even
though we could not conduct proper scientific assessments. Media
have also documented with interviews and reports the impacts
fumigation in local communities of the Colombian pacific region
(TeleSur_TV, 2011, 2011a).
In Nariño, the arrival of aerial fumigation was followed by an in-
crease in formal complaints and claims by the local population. Of
the reported claims on the aerial fumigation in the country be-
tween 2001 and 2008, 45% came from Nariño and concerned
health and crops loss, while at the peak of 2003, 76% of the total
claims (1950 claims) were from Nariño (Defensoria del Pueblo,
2007, 2009; Policia Nacional de Colombia – Dirección de Antinar-
coticos, 2010). Our analysis of the archive of the complaints shows
that between 2000 and 2006, a total of 1177 families reported ef-
fects from aerial fumigation including death of domestic animals
(ducks, chickens, pigs and cows), pollution and destruction of legal
crops used for self-consumption (chiro, chilma, cassava, papacum,
chontaduro, banana, coconut, cacao, corn, etc.) and impacts on
their health. Indigenous testimonies reported the deaths of three
children and two abortion cases between 2000 and 2006 due to
the fumigation. CODHES, an organization monitoring human rights
and displacement, denounced the death of 25 indigenous children
from starvation due to the impact of fumigation on food crops (El
Espectador, 2008). During our interviews, indigenous leaders told
us that there was no previous consultation or warning about the
fumigations and claimed that their water sources have been con-
taminated and that they have lost seeds and medicinal plants. At
the time of writing of this article, indigenous groups continue to
denounce publicly the displacement that fumigations cause (Auto-
ridades Indígenas AWÁ – UNIPA, 2011). There is still no evidence of
any intervention or assistance from public or private entities as a
response to the indigenous claims, despite the repeated denuncia-
tions of the terrible impacts (Oslender, 2010; Walsh et al., 2008b).
In conclusion: it is not possible to verify beyond doubt the neg-
ative impacts on the health of the people residing in the fumigated
areas. There are however serious indications that fumigation af-
fects the health of people and their legal crops. Relevant evidence
includes the explosion in the number of formal complaints associ-
ated to the fumigations and anecdotal experiential evidence, such
as that collected in our interviews with local people and the pho-
tographs we took of agricultural crops affected by fumigations.
3.4. Aerial fumigation and human displacement
Nariño has been one of the departments of Colombia that has
suffered the most in terms of forced human displacement. In
2000 when aerial fumigation started in Nariño, 732 cases of dis-
placed persons were reported, representing a 0.3% of the total in
Colombia. But between 2001 and 2008 and as production and
fumigations increased so did displacements reaching a total of
31,314 in 2008, corresponding to a 10% of the total of the popula-
tion displaced in Colombia (301,754).
3
What is happening in Nariño is part of a broader pattern. Based
on our analysis, 70% of the municipalities that experienced
3
Displacement in Colombia and subsequent migration also had to do with the
Pudricion de Cogollo (PC), a disease the oil palm had, and which hit gravely all the
farmers that had adopted monoculture palm systems and dedicated less time to
subsistence crops.
12 A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

increased fumigations between 2001 and 2008 also represent
increasing tendencies in forced displacement of the population,
as measured by the National Registry on forced population. There
is also a significant correlation between fumigated areas and forced
displacement for all the years analyzed. In other words, the more
the area fumigated in a municipality, the higher the number of
people that leaves it.
According to our interviews there are two factors at play: first,
part of the population was economically dependent on coca and
traditional crops and the destruction of cultivation by the fumiga-
tions forced them to move to other municipalities. Second, the aer-
ial fumigation impacted the traditional crops even of those families
that were not involved in coca, affecting food security and forcing
them to migrate (see also Messina and Delamater (2006) for Putu-
mayo, who documented that fumigation does not affect only areas
with coca but also areas with other cultivations). Indigenous and
Afro-Colombian communities have denounced the displacement
of population from their communities as a consequence of the
water contamination, land degradation and loss of food security
caused by fumigations (CODHES, 2011; Martinez-Alier et al.,
2010; OAIPC, 2010; Oslender, 2010; Solomon et al., 2005a; Walsh
et al., 2008a).
The government emphasizes the first driver, i.e. the displace-
ment of coca farmers. From its perspective this is not necessarily
a problem, since these are workers who are involved in illegal
activities. Furthermore, the government puts blame on other fac-
tors and argues that it is not the fumigations that have impacts
on the population (Solomon et al., 2005a). and force them to leave
The proposition is that displacement is mostly a result of the gen-
eral armed conflict and the violence of the armed groups and not
the fumigations per se. Fumigations target the areas where coca
and illegal groups presence is strong; according to this narrative,
it is not the fumigations that make the people leave but the vio-
lence of the coca business and the illegal groups. However other
researchers and NGOs (CODHES, 2008, 2009; Ibáñez and Moya,
2007; Ibáñez and Vélez, 2008) suggest that aerial fumigation is di-
rectly related and implicated in the armed conflict, and is in many
and different ways a cause of displacement in Colombia.
To test the claim of a relation between fumigation and displace-
ment we investigate whether there is a statistically significant cor-
relation between the extent of area fumigated and the number of
people displaced. To isolate this from the direct displacement ef-
fects of violence, we control for murder rate and violent acts by
illegal groups (state-used indicators of violence). In both cases
we find a statistically significant correlation (Tables 4 and 5),
which confirms that fumigation is associated to displacement inde-
pendent of the effects of violence. This is not to deny the effect of
violence on displacement, only to suggest that fumigation may
have a separate effect over and on top of violence. Furthermore
formal econometric research along the lines of Angrist and Kugler
(2008) could shed more light on the relative weights of the factors
that affect displacement and their possible interaction. One possi-
ble causal chain that needs to be further interrogated concerns a
secondary displacement effect, whereby fumigation in one area
causes displacement to a neighboring one, escalation of the vio-
lence there, leading to further displacement.
We investigated also whether there remains a correlation be-
tween aerial fumigation and displacement after controlling for
the number of people actively employed in coca, i.e. to exclude
the possibility that a higher displacement is simply the effect of
more people working in coca, and being displaced as a result of
fumigation destroying the crop. In other words, our goal is to see
whether people leave because of fumigation or because coca is
eradicated. Since there is no data available on the number of peo-
ple employed in coca in each municipality, we use the ha of coca
cultivated as a proxy for employment. Again, we find a statistically
significant correlation, suggesting that fumigation displaces also
normal residents, and not only those involved in coca cultivation
(Table 6), which is in accordance with what peasants told us in
Nariño (see below).
In conclusion there is suggestive evidence that fumigation is
associated with increased human displacement, even after taking
into account the contribution of violence and the displacement of
the labor working in the coca fields.
3.5. The uneven effects of fumigation
Is everyone in Colombia affected the same by fumigation and by
its side-effects, i.e. displacement and arrival of coca production
from the areas that were fumigated?
The first important finding is that the new coca areas where pro-
duction moves after fumigation tend to be less developed, impover-
Table 4
Partial correlation between aerial fumigation (AF) and Forced Displacement of
Population (FDP) controlled for violence (number of murders by illegal armed groups
– MIAG).
Control variable Var 1 Var 2 Partial correlation Value
MIAG 2003 AF 2003 FDP 2003 Correlation 0.075
Significance (2-tailed) 0.011
MIAG 2004 AF 2004 FDP 2004 Correlation 0.189
Significance (2-tailed) 0.000
MIAG 2005 AF 2005 FDP 2005 Correlation 0.118
Significance (2-tailed) 0.000
MIAG 2006 AF 2006 FDP 2006 Correlation 0.131
Significance (2-tailed) 0.000
MIAG 2007 AF 2007 FDP 2007 Correlation 0.170
Significance (2-tailed) 0.000
MIAG 2008 AF 2008 FDP 2008 Correlation 0.217
Significance (2-tailed) 0.000
Table 5
Partial correlation between aerial fumigation (AF) and Forced Displacement of
Population (FDP) controlled for violence (number of violent acts by illegal armed
groups – VAIA)/there was not statistical significance for 2007.
Control variable Var 1 Var 2 Partial correlation Value
VAIA 2002 AF 2002 FDP 2002 Correlation 0.122
Significance (2-tailed) 0.000
VAIA 2003 AF 2003 FDP 2003 Correlation 0.019
Significance (2-tailed) 0.535
VAIA 2004 AF 2004 FDP 2004 Correlation 0.142
Significance (2-tailed) 0.000
VAIA 2005 AF 2005 FDP 2005 Correlation 0.080
Significance (2-tailed) 0.007
VAIA 2006 AF 2006 FDP 2006 Correlation 0.073
Significance (2-tailed) 0.014
Table 6
Partial correlation between aerial fumigations (AF) and Forced Displacement of
Population (FDP) controlled for coca crops area (CA)/there was no statistical
significance for 2003 and 2008.
Control variable Var 1 Var 2 Partial correlation Value
CA 2001 AF-01 FDP-01 Correlation 0.113
Significance (2-tailed) 0.001
CA02 AF-02 FDP-02 Correlation 0.156
Significance (2-tailed) 0.000
CA04 AF-03 FDP-04 Correlation 0.190
Significance (2-tailed) 0.000
CA05 AF-04 FDP-05 Correlation 0.169
Significance (2-tailed) 0.000
CA06 AF-05 FDP-06 Correlation 0.130
Significance (2-tailed) 0.000
CA07 AF-06 FDP-07 Correlation 0.077
Significance (2-tailed) 0.014
A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx 13
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

ished areas, populated by indigenous and Afro-Colombians. 73% of
the municipalities exhibiting an increase in coca cultivation be-
tween 2001 and 2008 have a Rural Unsatisfied Basic Needs Index
(RUBN) exceeding 50% (generally considered the limit value indi-
cating impoverishment). A second index of impoverishment and
public services is the ‘‘municipal development index’’ (IMD): 83%
of municipalities that have exhibited increasing coca cultivation be-
tween 2001 and 2008 have an index value less than 50.
Our data shows also a statistically significant correlation be-
tween coca cultivation and the presence of illegal armed groups,
which are typically (though not always) those involved in the coca
business, as well as with the remoteness of an area (Table 7; the
presence of illegal armed groups is captured in indicators such as
the number of violent acts and murders by illegal armed groups).
In Nariño 17 people from the Awá indigenous community were
massacred in 2009, events that have been covered by the mass
media and were reported to us in our interviews with indigenous
leaders (ACNUR, 2009; Espectador, 2009).
Coca cultivation is correlated also well with (low) road density
and (high) level of natural cover. There is also a statistically-signif-
icant and strong inverse correlation between the percentage of the
municipal area cultivated with coca and indicators of development
(RUBN and IMD) (Table 7). All this suggests a particular geography
of the ballooning coca frontier towards remote and impoverished
areas where violence is already present (Garcés, 2005), with the
possibility of escalating levels of violence after the arrival of coca
(Angrist and Kugler, 2008), since the general income hardly in-
creases (Dávalos et al., 2009).
The communities where coca expands, possibly as a result of
fumigation, tend to be predominantly indigenous and Afro-Colom-
bian. According to our analysis, during the early 2000s, 7% of the
area of coca cultivation was found in indigenous territories, 4% in
collective territories of Afro-Colombian communities and 2% in
natural parks. By 2008 there is a significant change, as the crops
become located mainly in the collective territories of Afro-Colom-
bian communities (36% of the crops). It is estimated that there was
Table 7
Pearson’s correlations between% municipal area cultivated with coca and other variables associated.
% Coca crops area/road density 2000
Year 2001 2002 2003 2004 2005 2006 2007 2008
Pearson Correlation .093(
**
).168(
**
).178(
**
).174(
**
).193(
**
).179(
**
).193(
**
).216(
**
)
Sig. (2-tailed) 0.002 0.000 0.000 0.000 0.000 0.000 0.000 0.000
N1125 1125 1125 1125 1125 1125 1125 1125
% Coca crops area/municipal development index
Year 2001 2002 2003 2004 2005 2006 2007 2008
Pearson Correlation .101(
**
).186(
**
).194(
**
).190(
**
).207(
**
).202(
**
).235(
**
).283(
**
)
Sig. (2-tailed) 0.001 0.000 0.000 0.000 0.000 0.000 0.000 0.000
N1102 1102 1102 1102 1102 1102 1102 1102
% Coca crops area/forced displacement of population – FDP (number of people)
Year 2001 2002 2003 2004 2005 2006 2007 2008
Pearson Correlation .192(
**
) .228(
**
) .138(
**
) .144(
**
) .264(
**
) .305(
**
) .262(
**
) .378(
**
)
Sig. (2-tailed) 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
N1124 1124 1124 1124 1124 1124 1124 1124
% Coca crops area/% natural cover 00
Year 2001 2002 2003 2004 2005 2006 2007 2008
Pearson Correlation .109(
**
) .256(
**
) .246(
**
) .220(
**
) .261(
**
) .238(
**
) .254(
**
) .316(
**
)
Sig. (2-tailed) 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
N1124 1124 1124 1124 1124 1124 1124 1124
% Coca crops area/rate of Forced displaced of population
Year 2002 2003 2004 2005 2006 2007 2008
Pearson Correlation .152(
**
) .090(
**
) .094(
**
) .172(
**
) .167(
**
) .210(
**
) .275(
**
)
Sig. (2-tailed) 0.000 0.003 0.002 0.000 0.000 0.000 0.000
N1124 1124 1124 1124 1124 1124 1124
% Coca crops/murders by illegal armed groups
Year 2003 2004 2005 2006 2007 2008
Pearson Correlation .119(
**
) .108(
**
) .328(
**
) .286(
**
) .289(
**
) .259(
**
)
Sig. (2-tailed) 0.000 0.000 0.000 0.000 0.000 0.000
N1124 1124 1124 1124 1124 1124
% Coca crops area/rate murders by illegal armed groups
Year 2003 2004 2005 2006 2007 2008
Pearson Correlation 0.030 .129(
**
) .214(
**
) .291(
**
) .279(
**
) .226(
**
)
Sig. (2-tailed) 0.320 0.000 0.000 0.000 0.000 0.000
N1102 1102 1102 1102 1102 1102
% Coca crops/RUBN 05
Year 2005 2006 2007 2008
Pearson Correlation .185(
**
) .130(
**
) .174(
**
) .176(
**
)
Sig. (2-tailed) 0.000 0.000 0.000 0.000
N1116 1116 1116 1116

Correlation is significant at the 0.05 level (2-tailed).
**
Correlation is significant at the 0.01 level (2-tailed).
14 A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

an increase of the areas cultivated with coca in the collective terri-
tories of Afro-Colombian communities from 10,231 ha in 2001 to
29,076 ha in 2008. On the other hand the coca area in indigenous
territories decreased (according to the official data always) be-
tween 2001 and 2008 from 11,791 ha to 6049 ha, but following
the general pattern it has spread to new indigenous territories,
such as the indigenous territory AWA, which is discussed later.
We do not address here why coca moves to this type of regions.
Physical-geographical factors that have to do with forest cover,
remoteness vis-à-vis lack of access infrastructure are important,
but so are socio-political ones of a historical nature (Hough,
2011), socio-economic and institutional factors (Garcés, 2005; Ran-
gel, 2000; Rocha, 2000; Rubio, 2005) probably related to the low
social capital of certain regions (Thoumi, 2005b, 2005c), particu-
larly in those zones where low state capacity and the presence of
terror groups prevail (Oslender, 2008).
The same type of statistically significant correlations is ob-
served between fumigation, poverty and remoteness indicators;
this makes sense, given that fumigations concentrate to the areas
where coca is cultivated and coca cultivation also correlates with
these indicators (Table 8). There is therefore an association be-
tween the extent an area is fumigated and low levels of rural and
municipal development and high levels of rural population. On
the other hand, looking at the territorial distribution of coca
according to the legal status of a territory, we find that between
2000 and 2008 coca production increased only in collective territo-
ries of Afro-Colombian communities (Table 9), which are suffering
disproportionately from the effects of the war on drugs (Walsh
et al., 2008b), something confirmed also by our in situ observa-
tions. Afro-Colombian groups have denounced on several occasions
the fumigation policy (OAIPC, 2010), for example in departments
such as Nariño (one of the departments with the largest expansion
of fumigations in the study period – see Table 10).
A further question is whether there is intentional discrimination
by the government on its fumigation targets, i.e. whether Afro-
Colombian and indigenous areas are more likely to be fumigated
than those populated by whites, other factors equal. Our data does
not suggest so, since there is no remaining correlation between
area fumigated and the area occupied by indigenous territories
and Afro-Colombian communities if we control for the extent of
coca cultivations in the municipality. In other words, the areas
mostly fumigated are those that have the most coca. We confirm
together with (O’Shaughnessy and Branford, 2005) that these are
areas of poor peasants (‘‘campesinos’’), often of indigenous or
Afro-Colombian communities, who therefore suffer disproportion-
ately more from the war on drugs, but we do not find evidence of
selective targeting. Nonetheless, this is still a ‘‘war on the poor’’,
since it is the poor that live in the areas where the coca frontier
moves and the ones who suffer the impacts of both the coca trade
and the chemicals that are supposed to stop it (O’Shaughnessy and
Branford, 2005).
In conclusion: fumigations and coca cultivations are dispropor-
tionately concentrated in impoverished areas of peasant, indige-
nous and Afro-Colombian communities. There is probably no
selective fumigation targeting of such communities by the govern-
ment, but this should not divert us from the basic fact that at the
end it is these communities that suffer the most from the indis-
criminate chemical ‘‘war on drugs’’.
Table 8
Pearson’s correlations between aerial fumigation and other variables associated.
Aerial spraying/complaints related to impacts by aerial spraying
Year 2001 2002 2003 2004 2005 2006 2007 2008
Pearson Correlation .661(
**
) .507(
**
) .301(
**
) .294(
**
) .567(
**
) .287(
**
) .402(
**
) .510(
**
)
Sig. (2-tailed) 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000
N1124 1124 1124 1124 1124 1124 1124 1124
Aerial aspersion/municipal development index
Year 2001 2002 2003 2004 2005 2006 2007 2008
Pearson Correlation .109(
**
).096(
**
).122(
**
).154(
**
).095(
**
).185(
**
).159(
**
).240(
**
)
Sig. (2-tailed) 0.000 0.001 0.000 0.000 0.002 0.000 0.000 0.000
N1102 1102 1102 1102 1102 1102 1102 1102
Aerial aspersion/murders carried out by illegal armed groups
Year 2003 2004 2005 2006 2007 2008
Pearson Correlation .110(
**
) .151(
**
) .140(
**
) .365(
**
) .347(
**
) .258(
**
)
Sig. (2-tailed) 0.000 0.000 0.000 0.000 0.000 0.000
N1124 1124 1124 1124 1124 1124
Aerial aspersion/RUBN 2005
Year 2005 2006 2007 2008
Pearson Correlation .066(
*
) .131(
**
) .124(
**
) .153(
**
)
Sig. (2-tailed) 0.028 0.000 0.000 0.000
N1116 1116 1116 1116
Aerial aspersion/rural population 2005
Year 2005 2006 2007 2008
Pearson Correlation .136(
**
) .104(
**
) .107(
**
) .089(
**
)
Sig. (2-tailed) 0.000 0.001 0.000 0.003
N1112 1112 1112 1112
*
Correlation is significant at the 0.05 level (2-tailed).
**
Correlation is significant at the 0.01 level (2-tailed).
Table 9
Area cultivated with coca in each of the main territorial divisions according to legal
status.
Legal status of the territory 2000 2008
Collective territories of Afro-Colombian communities 3429 15,032
Natural National Park 3877 2691
Forest Reserve 40,919 19,007
Indigenous territories 11,876 5636
Subtracted area from forest reserve 82,909 16,450
A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx 15
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

4. Discussion
Why does aerial fumigation fail? Here we follow this body of lit-
erature which suggests that centrally managed social plans often
misfire, when they impose schematic visions that do violence to
complex local interdependencies that are not fully understood
(Scott, 1998). Norgaard (1994) for example, investigated the failure
of State development programs in the Brazilian Amazon and ar-
gued that ecological conditions posed obstacles and huge transac-
tion costs to the development approaches that the Brazilian State
imported from other parts of the world. In particular he showed
how the productive practices of local groups were well adapted
through a historical trial and error process to the ecosystem fea-
tures of the rainforest, ensuring that small surpluses could be
drawn with little transaction costs. Our case-study here, which
concerns not a developmental intervention but an ‘‘anti-crime’’
State project of improvement, confirms this thesis of Scott and
Norgaard in a very different context. We argue that State policies
of fumigation fail to see the particular ecological economy of coca
production in the Colombian territory, and hence fail to under-
stand why and how a policy of fumigation is likely to backfire.
The illegal nature of the coca business requires remoteness and
ability to hide the plantations. Tropical forests provide ideal envi-
ronments for growing coca. One reason is that their bio-physical
characteristics are favorable to the growth of the crop and to high
yields. Equally important however is that access to tropical forests
is limited, as it is very difficult to develop road infrastructure there.
Their remoteness renders them beyond direct central State control,
allowing criminal organizations to hide and avoid persecution
(Díaz and Sánchez, 2004; UNODC, 2011). It is therefore the same
factors of remoteness vis-à-vis the lack of modern state-based
development that render forests both primary biodiversity and
conservation hotspots (since human activity has been historically
limited) and ideal ‘‘habitats’’ for coca production.
4
From the outset
therefore, any policy which has as a result the displacement of pro-
duction, without being able to control its relocation elsewhere, is
likely to cause more deforestation, as illegal groups are likely to
move to new patches of remote tropical forest.
For those invested in the coca business, the tropical forests
serve multiple functions: stock, shelter and territory (Díaz and Sán-
chez, 2004; Thoumi, 2005d; UNODC, 2006a). Despite distance from
urban areas, the abundance and diversity of hydrological resources
and flora and fauna can sustain both production and the daily
needs of the armed groups that battle for control over territory
for coca cultivation. The tremendous surplus generated by the
business makes it possible for the criminal organizations to finance
and sustain lavish settlements for themselves (and livable for the
workers), even if located very distant from markets. Transport
and connection are secured by fluvial or aerial transport (Le Billon,
2001) explains how this peculiar socio-environmental geography
of products like coca, which are produced diffusively in the
territory (i.e. they are not localized resources, such as mines) and
require remoteness, go hand-and-glove with ‘‘war-lordism’’, i.e.
armed illegal groups controlling production and engaging in war
with the distant central government. From a government perspec-
tive, he explains, fighting war-lords in remote jungles requires ris-
ky ground engagement. Aerial fumigation emerges then as a risk-
less war from control centers in the cities, yet, it is one of question-
able effectiveness.
Why does coca production move though to poor areas and par-
ticularly areas where indigenous and Afro-Colombian groups re-
side? For historical reasons, which are well covered in other
works (Álvarez, 2001; Angrist and Kugler, 2008; Fajardo, 2002;
Ibáñez and Moya, 2007; Oslender, 2008; Thoumi, 2005a), the areas
where these groups reside are also the least-developed (in eco-
nomic terms) in the Colombian territory and the most remote.
Not only they provide ideal hiding and growing locations and are
far from the range of intervention of the military, but also, other
factors equal, it is easier for the illegal groups to recruit labor there.
The majority of the labor in coca production is allocated to cul-
tivation. Some of the wage laborers not only cultivate but also pro-
cess the coca leafs. In addition there is a floating population of day
laborers who sell periodically their labor on different parts of the
chain depending on seasonal production cycles. Wages are gener-
ally higher than wages in the labor market (Ibañez, 2010), although
this is not the significant difference. It is the stability and security
of income and employment that coca offers that is most appealing
to producers. This relative stable profit is tempting enough to com-
pensate for the personal and social disapproval that coca cultiva-
tion generates (Ibanez and Carlsson, 2010; Ibañez, 2010). Again,
supply-side policies of distant engagement, such as fumigation,
do little to change these dynamics. By destroying legal together
with illegal crops, they retain coca production as an attractive live-
lihood option for poor peasants. As more and more people become
destitute as a result of the fumigations and the terror of the illegal
groups, the supply of mobile coca laborers increases, making coca
production more responsive and adaptable to fumigation, labor
and production moving back and forth from fumigated areas with
more ease.
There are different policy options, more fit to the complex so-
cio-ecology of Colombia’s tropical forests and the ecological econ-
omy of the coca industry. Ramirez (2011) for example documents
how in Putumayo the most effective way of eradicating coca was
by hand, rather than by plane. Close engagement reduces the ben-
efits of remoteness and hiding in the tropical forest, whereas it al-
lows a more selective targeting of coca cultivations, without
affecting negatively other agricultural activities. A comparative
analysis of the relative effects of manual and aerial eradication of
coca in Colombia is an important object for further study. Another
policy option is the investment in alternative modes of develop-
ment (or alternatives to State-led, Western-type development),
responding to the needs of local population, with poverty reduc-
tion and development of local public infrastructure (Dion and
Russler, 2008). Local development can change the choice domain
for peasants, and make coca production an unattractive alternative,
reducing the labor supply for the illegal groups and making pro-
duction more expensive and less profitable. Still, supply-side poli-
cies alone are not likely to be effective, as long as there are no
policies to curb global demand for processed coca (Laffiteau,
2010). The costs of labor and the production process in general,
Table 10
Expansion of the aerial fumigations 2001–2008 in Colombia and Nariño (ha fumigated).
Aerial Spraying/año 2000 2002 2004 2006 2008
Aerial spraying in Colombia (ha) 58,074 130,364 136,551 171,754 133,496
Aerial spraying in Nariño (ha) 6,442 17,962 31,307 59,865 54,050
% of the total area sprayed in Colombia corresponding to Nariño 11 14 23 35 40
4
Colombia is in the top 12 countries with greatest biodiversity in the world Myers
et al. (2000). With a land area of only a 0.7% of the planet’s surface, Colombia hosts
about 10% of the fauna and flora of the world. Two of the world’s most important
biodiversity hotspots are in Colombia: the tropical Andes and the Chocó Humid
Forests Myers et al. (2000).
16 A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

are very small compared to the profits, which in their majority ac-
crue at later stages of the commodity chain, i.e. in trafficking. De-
mand-side policies are likely to have a much stronger effect on
the benefits of the trade than supply-side ones.
Why then is an ineffective policy, this of fumigation, sustained
in the presence of better alternatives? This is an important ques-
tion, but one that is beyond the purposes of this article. Others
have studied Colombia’s anti-drug policy, in the context of its geo-
political relations with the US and the dynamics of globalized cap-
italism (Corva, 2008; Crandall, 2008; Guizado, 2006; Thoumi,
2005a, 2005b). Our goal here has been more modest and consisted
of developing a spatialized information base for evaluating the im-
pacts of aerial fumigation and informing understandings of why
and how the policy has been failing.
5. Policy implications and conclusions
This article offered new evidence on the socio-environmental
consequences of Colombia’s war on drugs, and more specifically,
its fumigation policy. Fumigations have diffused the frontier of
coca cultivation, expanding deforestation to some of the world’s
most important biodiversity hotspots. The potential causal link
suggested by our research is important: it is not coca production
alone that causes the deforestation; it is the fumigation that is con-
tinuously pushing it to new areas. More and more people are being
displaced, particularly from the more vulnerable segments of the
population, including Afro-Colombian descendants. Even if fumiga-
tions have been reducing the cultivated area, which is question-
able, their goal of total eradication is not feasible; illegal groups
have easily adapted and responded to fumigation with fast reloca-
tion, forest clearance and production anew. While the intention of
the fumigation policy may have been to make coca cultivation too
costly to maintain, illegal groups have managed to shift the cost to
producers and the local people, maintaining the lucrative cocaine
trade going on. The costs of this ineffective war on drugs are dis-
proportionately distributed along lines of class (income), race
and ethnicity. Colombia continues to receive massive amounts of
US aid to wage this chemical air war on drugs. The policy implica-
tions of our study for Colombia and beyond are clear: any govern-
ment that attempts to stamp out coca production through aerial
fumigations should think twice about its effectiveness and its
side-effects.
In essence the problem at stake is one of (environmental and so-
cial) justice. Whereas the State and the illegal organizations may
satisfy some of their purposes with the existing status quo, the lo-
cal populations and the forest upon which they depend for their
livelihood lose. Correcting this grave injustice and ending the inef-
fective fumigation policy is not easy as there are strong political-
economic forces and interests at play that we did not address here.
Our goal was more modest and was to reinforce in a more rigorous,
integrated and scale-refined manner the documentation of the so-
cial and environmental effects of the war on drugs upon people
and forests. The hope is that such documentation will contribute
to building-up the pressure for a real public debate on the social
and environmental costs of the policy, and provide fodder to those
who are arguing for alternative approaches and for justice to be
given.
Acknowledgment
We thank the Programme Alßan for Financial Support. Con-
structive comments of Julianne Hazlewood, Marco Armiero and
Joan Martinez Alier. We would like to thank Milton Romero for
your guidance on spatial analyses and Francisco Thoumi for your
guidance at the beginning of the research. Plinio Perez and the
Awá indigenous community for their help in Nariño’s department
and the Integrated Illicit Crops Monitoring System (SIMCI in Span-
ish) for their support and supply of spatial information on coca
crops. Finally we thank the project CSO 2010 21979 on Social
Metabolism and Environmental Conflicts from the Spanish Minis-
try of Science and Innovation and to Blanca Edilma for your sup-
port all these years.
References
ACNUR, 2009. Diecisiete indígenas Awá asesinados en Nariño. Alto Comisionado de
las Naciones Unidas para los Refugiados (ACNUR), Bogotà. <www.acnur.org/
index.php?id_pag=8303> (10.02.09).
Álvarez, M., 2001. Could Peace Be Worse Than War for Colombia’s Forests? The
Environmentalist. Kluwer Academic Publishers – Manufactured in The
Netherlands, The Netherlands, pp. 305–315.
Álvarez, M., 2007. Environmental damage from illicit drug crops in Colombia. In: De
Jong, W.D., Donovan, Deanna, Abe, Ken-ichi (Eds.), Extreme Conflict and
Tropical Forests. Springer, Dordrecht, The Netherlands, p. 184.
Angrist, J.D., Kugler, A.D., 2008. Rural windfall or a new resource curse? Coca,
income, and civil conflict in Colombia. The Review of Economics and Statistics/
MIT Press 90 (2).
Anselin, L., 2005. Exploring Spatial Data with GeoDa: A Workbook. Center for
Spatially Integrated Social Science/Spatial Analysis Laboratory – Department of
Geography – University of Illinois, Urbana–Champaign.
Anselin, L., Syabri, I., Smirnov, O., 2002. Visualizing Multivariate Spatial Correlation
with Dynamically Linked Windows. University of Illinois, Urbana–Champaign,
Urbana, IL.
Autoridades Indígenas AWÁ – UNIPA, 2011. Fumigaciones en el Resguardo Indígena
del Gran Sábalo. Observatorio Pacífico y Territorio, Nariño. <http://
www.pacificocolombia.org/novedades/fumigaciones-resguardo-indigena-gran-
sabalo/390> (03.03.11).
Ávila, R., Bravo, E., Breilh, J., Campaña, A., Paz-y-Miño, C., Peñaherrera, L., Valencia, J.,
2007. El sistema de aspersiones aéreas del plan Colombia y sus impactos sobre
el ecosistema y la salud en la frontera ecuatoriana, Quito, p. 150.
Bernal, H., 2007. Generadores de impacto ambiental de los cultivos ilícitos y de la
extracción y refinamiento de alcaloides. Subdirección Estratégica y de
Investigaciones – Dirección Nacional de Estupefacientes, Bogotá, p. 22.
Bernal, M.H., Solomon, K.R., Carrasquilla, G., 2009. Toxicity of formulated glyphosate
(glyphos) and cosmo-flux to larval and juvenile Colombian frogs 2. Field and
laboratory microcosm acute toxicity. Journal of Toxicology and Environmental
Health, Part A: Current Issues 72 (15), 966–973.
Bolognesi, C., Carrasquilla, G., Volpi, S., Solomon, K.R., Marshall, E.J.P., 2009.
Biomonitoring of genotoxic risk in agricultural workers from five Colombian
regions: association to occupational exposure to glyphosate. Journal of
Toxicology and Environmental Health, Part A: Current Issues 72 (15), 986–997.
Bradley, A.V., Millington, A.C., 2008. Coca and colonists: quantifying and explaining
forest clearance under coca and anti-narcotics policy regimes. Ecology and
Society 13 (1), 31.
Brain, R.A., Solomon, K.R., 2009. Comparison of the hazards posed to amphibians by
the glyphosate spray control program versus the chemical and physical
activities of coca production in Colombia. Journal of Toxicology and
Environmental Health, Part A: Current Issues 72 (15), 937–948.
Cabieses, H., 1999. Drogas, coca y desarrollo sostenible, Iniciativa Fondo Perú –
Canadá 10 años. Fondo Perú – Canadá, Lima – Perú.
CODHES, 2008. Tapando el sol con las manos Informe sobre desplazamiento
forzado, conflicto armado y derechos humanos, enero-junio de 2008, Boletín
informativo de la Consultoría para los Derechos Humanos y el Desplazamiento.
Codhes – Consultoría para los Derechos Humanos y el Desplazamiento, Bogotá,
Colombia, p. 20.
CODHES, 2009. Víctimas emergentes – Desplazamiento, derechos humanos y
conflicto armado en 2008, Boletín informativo de la Consultoría para los
Derechos Humanos y el Desplazamiento. Codhes – Consultoría para los
Derechos Humanos y el Desplazamiento, Bogotá, Colombia, p. 15.
CODHES, 2011. Campesinos e indígenas de Cumaribo estarían en peligro de
desplazamiento por fumigaciones aéreas. Consultoria para los derechos
humanos y el desplazamiento (CODHES), Bogotá.
Corva, D., 2008. Neoliberal globalization and the war on drugs: transnationalizing
illiberal governance in the Americas. Political Geography 27 (2), 176–193.
Crandall, R., 2008. Driven by Drugs: US Policy toward Colombia, second ed. Lynne
Rienner Publishers.
Dávalos, L., Bejarano, A.C., Correa, H.L., 2009. Disabusing cocaine: pervasive myths
and enduring realities of a globalised commodity. International Journal of Drug
Policy 20 (5), 381–386.
Dávalos, L., Bejarano, A., Hall, M., Correa, H.L., Corthals, A., Espejo, O.J., 2011. Forests
and drugs: coca-driven deforestation in tropical biodiversity hotspots.
Environmental Science & Technology 45 (4), 1219–1227.
Defensoria del Pueblo, 2007. Informe defensorial sobre la situación de derechos
humanos y derecho internacional humanitario del pueblo indigena AWA
departamento de Nariño Colombia. Defensoria del Pueblo, República de
Colombia, Bogotá.
A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx 17
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

Defensoria del Pueblo, 2009. Base de datos de quejas y denuncia de la poblacion
afectada por las fumigaciones aereas. Defensoria del Pueblo, República de
Colombia, Bogota.
Díaz, A.M., Sánchez, F., 2004. Geography of Illicit Crops (Coca Leaf) and Armed
Conflict in Colombia. Center of Studies for Economic Development, Department
of Economics. Universidad de los Andes.
Dion, M.L., Russler, C., 2008. Eradication efforts, the state, displacement and
poverty: explaining coca cultivation in Colombia during plan Colombia. Journal
of Latin American Studies 40 (03), 399–421.
El Espectador, 2008. Denuncian la muerte de 25 niños por fumigaciones aéreas, El
Espectador, Bogotà (04.08.08).
Eslava, P., Ramírez, W., Rondón, I., 2007. Sobre los efectos del glifosato y sus
mezclas: impacto en peces nativos. Universidad de los llanos, Villavicencio –
Colombia.
Espectador, E., 2009. Defensoría del Pueblo habría advertido masacre en comunidad
Awa. Nariño, El Espectador, Bogotá.
Fajardo, D., 2002. Para sembrar la paz hay que aflojar la tierra. Instituto de Estudios
Ambientales (IDEA) – Universidad Nacional de Colombia, Bogotá – Colombia.
GAO, 2008. Plan Colombia: Drug Reduction Goals Were Not Fully Met, but Security
Has Improved. U.S. Agencies Need More Detailed Plans for Reducing Assistance,
Washington.
Garcés, L., 2005. Colombia: the link between drugs and terror. Journal of Drug Issues
35 (1), 83–105.
González, S., 2006. El Programa de Erradicación de Cultivos Ilícitos Mediante
Aspersión Aérea de Glifosato: hacia la clarificación de la política y su debate.
Monitoring and Research Center on Drugs and Crime. Facultad de Economía.
Universidad del Rosario, Bogotá, Colombia, p. 43.
Guáqueta, A., 2005. Change and continuity in U.S.-Colombian relations and the war
against drugs. Journal of Drug Issues 35 (1), 27–56.
Guizado, Á., 2006. Narcotráfico: Europa, Estados Unidos. Universidad de los Andes –
Facultad de ciencias sociales (CESO), Bogotá, Colombia, América Latina.
Guridi, L., 2002. Crisis económica, droga y desarrollo alternativo en la región andina:
el caso de Bolivia, 1985–1997. Universidad del País Vasco, servicio editorial
Bilbao.
Haney, R.L., Senseman, S.A., Hons, F.M., Zuberer, D.A., 2000. Effect of glyphosate on
soil microbial activity and biomass. Weed Science 48 (1), 89–93.
Hewitt, A.J., Solomon, K.R., Marshall, E.J.P., 2009. Spray droplet size, drift potential,
and risks to nontarget organisms from aerially applied glyphosate for coca
control in Colombia. Journal of Toxicology and Environmental Health, Part A:
Current Issues 72 (15), 921–929.
Hough, P.A., 2011. Disarticulations and commodity chains: cattle, coca and capital
accumulation along Colombia’s agricultural frontier. Environment and Planning
A 43 (5).
Ibañez, M., 2010. Who Crops Coca and Why? The Case of Colombian Farmers.
Discussion Papers, Courant Research Centre PEG.
Ibanez, M., Carlsson, F., 2010. A survey-based choice experiment on coca cultivation.
Journal of Development Economics 93 (2), 249–263.
Ibáñez, A.M., Moya, A., 2007. La población desplazada en Colombia: Examen de sus
condiciones socioeconómicas y análisis de las políticas actuales. Departamento
Nacional de Planeación Bogotá, Colombia.
Ibáñez, A.M., Vélez, C.E., 2008. Civil conflict and forced migration: the micro
determinants and welfare losses of displacement in Colombia. World
Development 36 (4), 659–676.
IDEA, 2005. Observaciones al ‘‘Estudio de los efectos del programa de Erradicación
de Cultivos Ilícitos mediante la aspersión aérea con el herbicida Glifosato
(PECIG) y de los cultivos ilícitos en la salud humana y en el medio ambiente’’.
Instituto De Estudios Ambientales (IDEA) – Universidad Nacional de Colombia,
Bogotá – Colombia, p. 35.
IDEAM, IGAC, IAvH, Invemar, Sinchi, IIAP, 2007. Ecosistemas continentales,
costeros y marinos de Colombia. Instituto de Hidrología, Meteorología y
Estudios Ambientales (IDEAM), Instituto Geográfico Agustín Codazzi (IGAC),
Instituto de Investigación de Recursos Biológicos Alexander von Humboldt
(IAvH), Instituto de Investigaciones Ambientales del Pacífico Jhon von
Neumann (IIAP), Instituto de Investigaciones Marinas y Costeras José Benito
Vives De Andréis (INVEMAR) e Instituto Amazónico de Investigaciones
Científicas (Sinchi), Bogotá, Colombia.
Kennedy, R., Stefani, F., 2009. Acerca del sector rural – el glifosato. Fundación de
Investigaciones Sociales y Políticas, Buenos Aires.
La Nación.com.co, 2010. Nuevo S.O.S. por fumigaciones. La Nación – Florencia/
Reporter: Alexa Ramirez. http://www.lanacion.com.co/2010/11/03/nuevo-s-o-
s-por-fumigaciones/ (03.11.10).
Laffiteau, C., 2010. The Balloon Effect: The Failure of Supply Side Strategies in the
War on Drugs. The European Consortium for Political Research (ECPR).
Le Billon, P., 2001. The political ecology of war: natural resources and armed
conflicts. Political Geography 20 (5), 561–584.
Martinez-Alier, J., Kallis, G., Veuthey, S., Walter, M., Temper, L., 2010. Social
metabolism, ecological distribution conflicts, and valuation languages.
Ecological Economics 70 (2), 153–158.
Messina, J.P., Delamater, P.L., 2006. Defoliation and the war on drugs in Putumayo,
Colombia. International Journal of Remote Sensing 27 (1), 121–128.
Moreno-Sanchez, R., Kraybill, D.S., Thompson, S.R., 2003. An econometric analysis of
coca eradication policy in colombia. World Development 31 (2), 375–383.
Myers, N., Mittermeier, R.A., Mittermeier, C.G., da Fonseca, G.A.B., Kent, J., 2000.
Biodiversity hotspots for conservation priorities. Nature 403 (6772), 853–858.
Nivia, E., 2001a. Efectos sobre la salud y el ambiente de herbicidas que contienen
glifosato. Boletín electrónico de la UITA, Montevideo, Uruguay.
Nivia, E., 2001b. Las fumigaciones aéreas sobre cultivos ilícitos si son peligrosas –
Algunas aproximaciones, Conferencia ‘‘Las Guerras en Colombia: Drogas, Armas
y Petróleo’’ ‘‘The Wars in Colombia: Drugs, Guns and Oil’’. Instituto Hemisférico
de las Américas, Universidad de California, Davis.
Norgaard, R., 1994. Development Betrayed: The End of Progress and a
Coevolutionary Revisioning of the Future. Routledge.
OAIPC, 2010. Declaración de Buenaventura: Declaración de organizaciones
afrocolombianas e indígenas del Pacífico ecoportal.net. OAIPC,
(Organizaciones afrocolombianas e indígenas del Pacífico Colombiano).
Ecoportal.net, Buenaventura, Colombia.
O’Shaughnessy, H., Branford, S., 2005. Chemical Warfare in Colombia: The Costs of
Coca Fumigation. Latin America Bureau, London.
Oslender, U., 2008. Another history of violence: The production of ‘‘Geographies of
Terror’’ in Colombia’s Pacific Coast Region. Latin American Perspectives 35 (5),
77–102.
Oslender, U., 2010. Fumigaciones en el área de Guapi, Costa Caucana – un crimen
contra la humanidad y la biodiversidad – Letter addressed to the President of
Colombia (Alvaro Uribe), demand an immediate stop to the fumigations. The
signers are Dr. Ulrich Oslender and othe 50 academics that have worked in the
Colombian Pacific Region, Glasgow.
Paredes, M., Correa, H., 2007. Coca, cocaína y medioambiente: treinta años de
historia. In: Calvani, S. (Ed.), La coca, pasado y presente, mitos y realidades.
Ediciones Aurora, Bogota, Colombia, pp. 103–124.
Policia Nacional de Colombia – Dirección de Antinarcoticos, D., 2010. Base de datos
de quejas y denuncias de los afectados por el Programa de Aspersion Aerea de
Cultivos Ilicitos con el herbicida glifosato (2001–2008), Bogotá.
Ramirez, M.C., 2011. Between the Guerrillas and the State: The Cocalero Movement,
Citizenship, and Identity in the Colombian Amazon. Duke University Press,
Durham and London.
Rangel, A., 2000. Parasites and predators: guerrillas and the insurrection economy
of Colombia. Journal of International Affairs 53 (2), 577.
Rocha, R., 2000. La economía colombiana tras 25 años de narcotráfico. Siglo del
Hombre, UNDCP, Bogotá, Colombia.
Rubio, M., 2005. Illegal armed groups and local politics in Colombia. Journal of Drug
Issues 35 (1), 107–130.
Saavedra, L.Á., 2009. Awas Unite to Fend Off Threats. Latinamerica Press.
Salisbury, D.S., Fagan, C., 2011. Coca and conservation: cultivation, eradication, and
trafficking in the Amazon borderlands. GeoJournal. http://dx.doi.org/10.1007/
s10708-011-9430-x.
Scott, J., 1998. Seeing Like a State How Certain Schemes to Improve the Human
Condition Have Failed. Yale University Press.
Sherret, L., 2005. Futility in action: coca fumigation in Colombia. Journal of Drug
Issues 35 (1), 151–168.
Social, Accion, 2009. Caracterización de las Reservas Forestales de Ley 2/59.
Consultorías e investigaciones/Proyecto Protección de Tierras y Patrimonio de la
Población Desplazada – Acción Social, Bogotá.
Solomon, K., Anadón, A., Luiz, C.A., Marshall, J., Sanin, L.H., 2005a. Estudio de los
efectos del Programa de Erradicación de Cultivos Ilícitos mediante la aspersión
aérea con el herbicida Glifosato (PECIG) y de los cultivos ilícitos en la salud
humana y en el medio ambiente. Comisión Interamericana para el Control del
Abuso de Drogas (CICAD) – División de la Organización de los Estados
Americanos (OEA).
Solomon, R.K., Anadón, A., Cerdeira, A.L., Marshall, J., Sanin, L.H., 2005b.
Environmental and Human Health Assessment of the Aerial Spray Program
for Coca and Poppy Control in Colombia. A Report Prepared for the Inter-
American Drug Abuse Control Commission (CICAD) Section of the Organization
of American States (OAS).
TeleSur_TV, 2011. Cultivos colombianos afectados por fumigaciones aéreas.
TeleSur_TV Colombia/Reporter Vladimir Carrillo, Barbacoas – Nariño –
Colombia. <http://multimedia.telesurtv.net/media/telesur.video.web/telesur-
web/#!es/video/cultivos-colombianos-afectados-por-fumigaciones-aereas/>
(08.08.11).
TeleSur_TV, 2011a. Campesinos colombianos afectados por glifosato. TeleSur_TV
Colombia/Reporter Vladimir Carrillo. <http://multimedia.telesurtv.net/media/
telesur.video.web/telesur-web/#!es/video/campesinos-colombianos-afectados-
por-glifosato/> (17.05.11).
The Economist, 2001. How governments try—and fail—to stem the flow of drugs.
The Economist.
Thoumi, F., 2005a. The colombian competitive advantage in illegal drugs: the role of
policies and institutional changes. Journal of Drug Issues 35 (1), 7–25.
Thoumi, F., 2005b. The Colombian competitive advantage in illegal drugs: the role
of policies and institutional changes. Journal of Drug Issues 35 (1).
Thoumi, F., 2005c. Ventajas competitivas ilegales, el desarrollo de la industria de
drogas ilegales y el fracaso de las políticas contra las drogas en Afganistán y
Colombia, Facultad de Economía. Centro de Estudios y Observatorio de Drogas y
Delito ed. Centro de Estudios y Observatorio de Drogas y Delito – Facultad de
Economía. Universidad del Rosario, Bogotá, Colombia, p. 17.
Thoumi, F., 2005d. The causes of illegal drug industry growth in the Andes, anti-
drug policies and their effectiveness, Centro Editorial Universidad del Rosario
ed. Centro de Estudios y Observatorio de Drogas y Delito – Facultad de
Economía. Universidad del Rosario, Bogotá, Colombia, p. 57.
United States Department of State, 2010. International Narcotics Control Strategy –
Report Drug and Chemical Control. Department of State – Bureau for
International Narcotics and Law Enforcement Affairs, United States.
UNODC, 2006a. Características Agroculturales de los Cultivos de Coca en Colombia.
United Nations Office on Drugs and Crime – UNODC – Colombia, Bogotá.
18 A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009

UNODC, 2006b. Colombia Coca Cultivation Survey. United Nations Office on Drugs
and Crime – UNODC Bogota, Colombia.
UNODC, 2008a. World Drug Report 2008. United Nations Office on Drugs and Crime
– UNODC, Vienna, Austria, p. 306.
UNODC, 2008b. Colombia Coca Cultivation Survey. United Nations Office on Drugs
and Crime – UNODC, Bogota, Colombia, p. 100.
UNODC, 2009. Colombia Coca Cultivation Survey. United Nations Office on Drugs
and Crime – UNODC, Bogota, Colombia, p. 108.
UNODC, 2010. World Drug Report 2010. United Nations Office on Drugs and Crime –
UNODC, Vienna, Austria, p. 309.
UNODC, D., 2011. Persistencia y productividad de la coca en la región pacífico.
United Nations Office on Drugs and Crime – UNODC and Dirección Nacional de
Estupefacientes (DNE), Bogotá, Colombia.
Vargas, R., 2004. Fumigaciones y política de drogas en Colombia: ¿fin del círculo
vicioso o un fracaso estratégico? In: Cárdenas, M., Rodriguez, M. (Eds.), Guerra,
sociedad y medio ambiente. Foro Nacional Ambiental, Bogotá, pp. 353–395.
Varona, M., Henao, G., Díaz, S., Murcia, A., Rodríguez, N., Álvarez, V., 2009.
Evaluación de los efectos del glifosato y otros plaguicidas en la salud humana en
zonas objeto del programa de erradicación de cultivos ilícitos. Biomédica 29, 19.
Walsh, J., Sánchez, J., Salinas, Y., 2008a. Chemical Reactions Fumigation: Spreading
Coca and Threatening Colombia’s Ecological and Cultural Diversity. Washington
Office on Latin America (WOLA), Washington.
Walsh, J., Sánchez, J., Salinas, Y., 2008b. La aspersión de cultivos de uso ilícito en
Colombia. Washington Office on Latin America (WOLA), Instituto de Estudios
para el Desarrollo y la Paz (Indepaz), Bogotá.
A. Rincón-Ruiz, G. Kallis / Geoforum xxx (2013) xxx–xxx 19
Please cite this article in press as: Rincón-Ruiz, A., Kallis, G. Caught in the middle, Colombia’s war on drugs and its effects on forest and people. Geoforum
(2013), http://dx.doi.org/10.1016/j.geoforum.2012.12.009