October 2016
This publication was produced at the request of the United States Agency for International Development (USAID). 
It was prepared independently by Jairo Núñez, Daniel Wiesner, Keri Culver, Roger Rasnake, Centro Nacional de 
Consultoría and Management Systems International. 
EVALUATION
Midterm Impact Evaluation of the Consolidation 
and Enhanced Livelihood Initiative
General Report
Midterm Impact Evaluation
of the Consolidation and 
Enhanced Livelihood 
Initiative
GENERAL REPORT
Management Systems International
A Tetra Tech Company
200 12th Street South
Suite 1200
Arlington, VA 22202, USA
www.msiworldwide.com
October 2016 
Contracted under AID-514-C-13-00003 
USAID/Colombia Evaluation and Analysis for Learning (EVAL)
DISCLAIMER
The author’s views expressed in this publication do not necessarily reflect the views of the United States 
Agency for International Development or the United States Government.
CONTENTS
Acronyms ...................................................................................................................................................................... 1
Abstract.......................................................................................................................................................................... 3
Executive summary ....................................................................................................................................................... 5
Introduction............................................................................................................................................................... 5
Economic development............................................................................................................................................. 7
Social development .................................................................................................................................................13
Governance and institutions....................................................................................................................................17
Security ...................................................................................................................................................................20
Recommendations...................................................................................................................................................24
Introduction ................................................................................................................................................................. 25
Evaluation purpose & evaluation questions.................................................................................................................29
Evaluation rationale.................................................................................................................................................29
Evaluation questions ...............................................................................................................................................30
Evaluation design ........................................................................................................................................................32
How to read this report............................................................................................................................................32
Methodology and design .........................................................................................................................................33
Quantitative methods..........................................................................................................................................35
Qualitative methods............................................................................................................................................40
Limitations ..............................................................................................................................................................41
Economic development ...............................................................................................................................................46
Findings and conclusions........................................................................................................................................46
Direct CELI beneficiaries...................................................................................................................................46
Rural economic opportunities: View from the grassroots...................................................................................49
Assistance for productive projects...................................................................................................................... 55
Markets and trade................................................................................................................................................57
Credit, savings and financial services.................................................................................................................60
Transportation infrastructure...............................................................................................................................63
Land ....................................................................................................................................................................65
Recommendations...................................................................................................................................................68
Social development......................................................................................................................................................71
Findings and conclusions........................................................................................................................................73
The Index of Organizational Competencies (ICO) .............................................................................................73
Social capital: Trust ............................................................................................................................................76
Social capital: Participation in associations........................................................................................................81
The particular role of the Juntas de Acción Comunal (JACs).............................................................................83
Associations and productive projects..................................................................................................................86
Women’s and victims’ associations....................................................................................................................90
Cultural and other associations...........................................................................................................................91
Recommendations...................................................................................................................................................93
Governance and institutions ........................................................................................................................................96
Findings and conclusions........................................................................................................................................97
Participation, transparency, accountability and trust ..........................................................................................97
Service delivery ................................................................................................................................................102
Recommendations.................................................................................................................................................107
Security......................................................................................................................................................................110
Findings and conclusions......................................................................................................................................111
Synthetic index of insecurity ............................................................................................................................111
Coca cultivation ................................................................................................................................................113
Terrorism and high-impact crimes....................................................................................................................115
Other crimes......................................................................................................................................................117
Security- and justice-related capacities.............................................................................................................119
Military capacities.............................................................................................................................................119
Police presence .................................................................................................................................................120
Availability of justice services..........................................................................................................................121
Recommendations.................................................................................................................................................122
TABLE OF FIGURES
Figure 1. Impacts on perceptions of the economic situation, direct beneficiaries .........................................................7
Figure 2. Impacts on reasons they associate with others, direct beneficiaries...............................................................8
Figure 3. Impacts on land issues, direct beneficiaries versus all CELI..........................................................................8
Figure 4. CELI performance, increasing sales...............................................................................................................9
Figure 5. Trends in economic situation perceptions, CELI municipalities..................................................................10
Figure 6. Impacts on perceptions of problems in production and sales, CELI and PNCRT........................................11
Figure 7. Trends in perceptions of problems in production and sales, CELI and PNCRT ..........................................11
Figure 8. CELI Performance: Average improvements by category, CELI N/S...........................................................14
Figure 9. Impacts on the Social Capital Index, CELI and PNCRT..............................................................................14
Figure 10. Impacts on participation in community activities, CELI and PNCRT .......................................................15
Figure 11. Impacts on participation in producer associations, CELI and PNCRT ......................................................16
Figure 12. CELI Performance: Accountability Index, by region.................................................................................17
Figure 13. Secondary data: Integral Performance Index, 2012 (left) and 2014 (right)................................................19
Figure 14. CELI Performance: Governance Capacity Index, total and regional .........................................................19
Figure 15. Secondary data: Insecurity Index ...............................................................................................................21
Figure 16. Trends in perceptions about illicit crops in their communities, CELI........................................................22
Figure 17. Secondary data: Average number of attacks by guerrilla groups, per municipality...................................23
Figure 18. Secondary data: Police presence ................................................................................................................24
Figure 1.1. The three pillars of the PNCRT and their CELI peers...............................................................................26
Figure 1.2. Evaluation questions ................................................................................................................................. 30
Figure 2.1. How to read and interpret the statistical data in the report........................................................................32
Figure 2.2. Sample by sex of respondent.....................................................................................................................33
Figure 2.3. Sample by ethnicity...................................................................................................................................34
Figure 2.4. Sample by poverty proxies........................................................................................................................34
Figure 2.5. Veredas by security threat level ................................................................................................................35
Figure 2.6. Consolidation Index pillars and their components ....................................................................................37
Figure 2.7. Secondary data: Consolidation Index overall scores.................................................................................38
Figure 2.8. Secondary data: Consolidation Index scores for territorial institutionalization.........................................39
Figure 2.9. Secondary data: Consolidation Index scores for good government, citizen participation.........................39
Figure 2.10. Secondary data: Consolidation Index scores for regional integration .....................................................39
Figure 3.1. Impacts on perceptions of the economic situation, direct beneficiaries ....................................................47
Figure 3.2. CELI Performance: Increasing sales.........................................................................................................47
Figure 3.3. Impacts on reasons they associate with others, direct beneficiaries..........................................................48
Figure 3.4. Impacts on credit and financial services, direct beneficiaries versus all CELI..........................................48
Figure 3.5. Impacts on land issues, direct beneficiaries versus all CELI.....................................................................49
Figure 3.6. Trends in economic situation perceptions, CELI municipalities...............................................................50
Figure 3.7. Impacts on perceptions of economic and food security variables, CELI and PNCRT..............................51
Figure 3.8. Impacts on perceptions of serious economic problems, CELI and PNCRT..............................................51
Figure 3.9. Secondary data: Consolidation Index on regional integration, trends 2011-2014.....................................52
Figure 3.10. Trends in reported monthly household expenditure, CELI and PNCRT.................................................54
Figure 3.11. CELI Performance: Estimated monthly household expenditure, total and regional................................54
Figure 3.12. Impacts on perceptions of problems in production and sales, CELI and PNCRT...................................58
Figure 3.13. Trends in perceptions of problems in production and sales, CELI and PNCRT .....................................58
Figure 3.14. CELI Performance: New use of financial services..................................................................................61
Figure 4.1. Secondary data: Consolidation Index scores for social development........................................................72
Figure 4.2. CELI Performance: Average change in ICO score, CELI Central ............................................................74
Figure 4.3. CELI Performance: Average change in ICO-VEO scores, CELI Norte/Sur.............................................74
Figure 4.4. CELI Performance: Average changes by ICO-VEO category, CELI N/S ................................................75
Figure 4.5. CELI Performance: Social Capital Index, total and regional ....................................................................76
Figure 4.6. Impacts on the Social Capital Index, CELI and PNCRT...........................................................................77
Figure 4.7. Impacts on trust between peers, CELI and PNCRT ..................................................................................77
Figure 4.8. Impacts on opinions of the personería; Trends by percent change ...........................................................78
Figure 4.9. Impacts on perceptions of the police and army, CELI and PNCRT..........................................................79
Figure 4.10. Impacts on resolving conflicts in the community, CELI and PNCRT ....................................................80
Figure 4.11. Impacts on perceptions of mayor’s office performance, CELI and PNCRT...........................................80
Figure 4.12. Impacts on participation in community activities, CELI and PNCRT ....................................................82
Figure 4.13. Impacts on participation in the JACs, CELI and PNCRT .......................................................................85
Figure 4.14. Impacts on participation in producer associations, CELI and PNCRT ...................................................87
Figure 4.15. Impacts on reasons they associate, CELI and PNCRT............................................................................88
Figure 4.16. Impacts on participation in victims’ and women’s groups, CELI and PNCRT.......................................91
Figure 5.1. Impacts on the Participation Index, CELI and PNCRT.............................................................................97
Figure 5.2. CELI Performance: Accountability index, total and regional ...................................................................99
Figure 5.3. Impacts on perceptions of problems with municipal administration, CELI and PNCRT........................100
Figure 5.4. Impacts on citizens’ perceptions of public management, CELI and PNCRT..........................................101
Figure 5.5. Secondary data: Integral Performance Index for the four groups of municipalities................................103
Figure 5.6. Secondary data: Integral Performance Index, 2012 (left) and 2014 (right).............................................104
Figure 5.7. CELI Performance: Governance capacity index, total and regional........................................................105
Figure 5.8. Impacts on perceptions of social services, CELI and PNCRT ................................................................106
Figure 6.1. Secondary data: Insecurity Index ............................................................................................................112
Figure 6.2. Secondary data: Evolution of coca cultivation........................................................................................113
Figure 6.3. Trends in reports that there are ICs in the village, CELI and PNCRT ....................................................114
Figure 6.4. Trends in perceptions about illicit crops in their communities, CELI.....................................................115
Figure 6.5. Secondary data: Average number of attacks by guerrilla groups, per municipality................................116
Figure 6.6. Trends in proportion perceiving security problems, CELI......................................................................117
Figure 6.7. Trends in reports of illegal mining, CELI and PNCRT...........................................................................118
Figure 6.8. Secondary data: Police presence .............................................................................................................121
1 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
ACRONYMS
ARD Associates for Rural Development
BACRIM Criminal bands, Bandas Criminales (in the service of the narcotics trade)
CAR Regional Autonomous Corporations, Corporaciones Regionales Autónomas
CDCS Country Development Cooperation Strategy
CELI Consolidation and Enhanced Livelihood Initiative
CELI Central CELI program in the Central region
CELI MdM CELI program in the Montes de María region
CELI Norte/Sur CELI program in the North and South regions
CERAC Conflict Analysis Resource Center, Centro de Recursos para el Análisis de Conflictos
CNC National Consulting Center, Centro Nacional de Consultoría
CNC-DoD Crime and Narcotics Center, U.S. Department of Defense
CSDI Colombia Strategic Development Initiative
CSO Civil Society Organization
DNP National Planning Department, Departamento Nacional de Planeación
DO Development objective
DoD Department of Defense (U.S.)
DPS Department for Social Prosperity, Departamento para la Prosperidad Social
ELN National Liberation Army, Ejército de Liberación Nacional
EVIDINCE Research effort associated with the Empirical Studies of Conflict (esoc.princeton.edu), a 
multi-institutional collaboration studying conflict and insurgency at the sub-national level
FARC Revolutionary Armed Forces of Colombia, Fuerzas Armadas Revolucionarias de Colombia
FHH Female head of household
GLAC Local savings and loans groups, grupos locales de ahorros y crédito
GOC Government of Colombia
HH Household
IAG Illegal Armed Group
IC Illicit crops
ICO Index of Organizational Competencies, Índice de competencias organizacionales
IDI Integral Performance Index, Índice de desempeño integral
INCODER Colombian Institute of Rural Development, Instituto Colombiano de Desarrollo Rural
JAC Community Action Council, Junta de Acción Comunal
LAPOP Latin American Public Opinion Project
MADR Ministry of Agriculture and Rural Development, Ministerio de Agricultura y Desarrollo 
Rural
MdM Montes de María
2 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
MHH Male head of household
MSI Management Systems International, Inc. 
PAR Regional Action Plan, Plan de acción regional
PMP Performance management plan
PNCRT National Territorial Consolidation and Reconstruction Policy, Política Nacional de 
Consolidación y Reconstrucción Territorial
SENA National Learning Service, Servicio Nacional de Aprendizaje
SGP General Participation System, Sistema General de Participación
SIJIN Judicial police, Seccional de Investigación Judicial
SIMCI Integrated Illicit Crops Monitoring System, Sistema Integrado de Monitoreo de Cultivos 
Ilícitos
SISBEN Identification and Classification system for potential social program Beneficiaries, Sistema 
de Identificación y Clasificación de potenciales Beneficiarios para programas sociales
TA Technical Assistance
UACT Territorial Consolidation Administrative Unit, Unidad Administrativa para la 
Consolidación Territorial
UARIV Victims’ Integral Attention and Reparation Unit, Unidad para la Atención y Reparación 
Integral a las Víctimas
UMATA Municipal Agricultural Extension Units, Unidades Municipales de Asistencia Técnica 
Agropecuaria
UNODC United Nations Office on Drugs and Crime
URT Land Restitution Unit, Unidad de Restitución de Tierras
USAID United States Agency for International Development
VEO Verification of Organizational Status, Valorización del Estado Organizacional
3 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
ABSTRACT 
USAID/Colombia designed the Consolidation and Enhanced Livelihood Initiative (CELI) to 
support the Government of Colombia (GOC) in resuming its institutional presence in areas of the 
country under the influence of insurgent groups engaged in the production of illegal crops, drug 
trafficking, kidnapping and terrorist activities. 
This impact evaluation compares the status or conditions of beneficiaries and their households 
before and after their exposure to CELI interventions, isolating external factors that could 
influence changes in these conditions. The evaluation team conducted 19,946 rural household 
surveys across in 32 CELI municipalities as well as in 16 municipalities from three regions 
(Catatumbo, Cauca/Valle and Putumayo) in which the GOC intervened through its National 
Policy for Territorial Consolidation and Reconstruction (PNCRT) activities.
Economic Development 
Nationwide rural development challenges and infrastructure deficiencies continue to affect CELI 
and PNCRT municipalities, and most respondents report a decline in their economic conditions. 
Poverty is present at the household, individual and community level, as market failings and lack 
of road access severely constrain growth. Access to credit remains tenuous across the regions, 
and land tenancy has made few gains. Income has increased, roughly at the level of (urban) 
inflation, but data suggest worsening food security. CELI beneficiaries report general 
improvements in economic conditions. The challenge of sustainably transporting products to 
markets remains.
Social Development 
The CELIs documented increases in organizational competency scores – especially in business 
initiatives and services – and surpassed program targets. From citizen perspectives, however, 
there were no increases in the social capital index and a statistically significant decrease in the 
trust component. Trust in state institutions fell across the board except for the Ombudsman 
offices. Participation in the Juntas de Acción Comunal (JACs) declined, but participants who 
stayed played a greater role in decision-making. Surprisingly, the survey revealed a decline in 
CELI municipalities’ involvement with producers associations. In both CELI and control 
municipalities, respondents attested to an increase in productive capacities. Participation in 
women´s organizations and victims’ associations increased, and ethnically oriented cultural 
associations strengthened a sense of identity and built social capital.
Governance and Institutions
Citizen participation at the ballot box remains relatively high, but electoral challenges such as 
vote-buying continues. Citizen oversight is minimal, but CELI’s work with the Ombudsman (in 
at least one region) strengthened the registry of veedurías. Service delivery remains a challenge, 
with citizen perceiving poor access and low quality. Reports of human rights violations by the 
military decreased but reports of police violations increased in PNCRT municipalities. 
4 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Security
The data show fewer crimes overall and improved perceptions of security in CELI and PNCRT 
municipalities. However, extortion is rising and coca cultivation has doubled nationally. 
Respondents in CELI and PNCRT zones show more tolerant opinions about the growth and sale 
of coca. Military and police force levels have changed and justice services continue to be absent 
in many of the areas under study. These characteristics leave the zones unstable as before, with a 
lack of state presence that continues to put citizens at risk.
5 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
CSDI and PNCRT zones in Colombia
Source: 2013, CSDI Impact Evaluation Baseline Report 
prepared by DevTech Systems, Inc. and Econometría, S.A.
EXECUTIVE SUMMARY
Introduction
After decades of conflict in rural areas with illegal armed groups (IAGs) and with increasing 
success in limiting their activities, the Government of Colombia (GOC) launched the National 
Consolidation and Territorial Reconstruction Policy (PNCRT) in 2009 to increase its presence 
and to reduce inequity in the historically marginalized conflict zones where IAGs and coca 
production coexist. To support this effort, the U.S. Government created the Colombia Strategic 
Development Initiative (CSDI) as an integrated, geographically-focused strategy with significant 
investments in security, anti-narcotics, access to justice, and coca eradication.
For its part, USAID/Colombia created the Colombia 
Enhanced Livelihood Initiative (CELI) to assist in the 
GOC’s goals of consolidating security gains, reducing 
coca cultivation and the influence of armed groups, 
and expanding state services. USAID selected three 
organizations to work in four regions that were under 
the influence of insurgent groups engaged in the 
production of illegal crops, drug trafficking, 
kidnapping and terrorist activities. CELI Norte/Sur 
(N/S) is implemented by Chemonics in Bajo Cauca, 
Sur de Córdoba and Tumaco; CELI Central, operated 
by ARD Tetra Tech, works in Cauca, Caquetá, Meta 
and Tolima; and CELI Montes de María (MdM) was 
managed by Global Communities in four 
municipalities in that region. At the time of the 
evaluation, CELI N/S and CELI Central were 
ongoing, while CELI MdM completed its work in 
early 2015. 
The three CELIs were intended from the beginning to 
be the subject of a rigorous impact evaluation 
consistent with USAID’s agency-wide Evaluation 
Policy, and in 2012 USAID charged its monitoring 
and evaluation (M&E) contractor DevTech Systems, 
Inc., to design an impact evaluation of the set of CSDI interventions and collect the baseline 
survey data. DevTech worked with Econometría, S.A., to carry out this work.1 The evaluation 
 
1 DevTech Systems, Inc. and Econometría S.A. 2013. “Impact Evaluation Methodology of the USAID Strategy in CSDI Zones - Final 
Methodology Report.”
6 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
design included two subsequent rounds of evaluations: a midterm evaluation, and a final 
evaluation following completion of the program. 
The current independent midline impact evaluation of the results of CSDI efforts took place after 
three years of program implementation for the Norte/Sur and Central corridors and four years for 
Montes de María. USAID charged its Evaluation and Analysis for Learning (EVAL) project, a 
contract with Management Systems International, Inc. (MSI), to carry out the midterm measure 
of the evaluation and to report on the CELI program impacts to date. This report represents that 
effort, based on the baseline evaluation design, additional qualitative fieldwork, interviews with 
the CELI teams, and an extensive survey of participants. 
The evaluation took place in the 33 municipalities targeted by USAID and CSDI as well as in 
162 municipalities from three regions (Catatumbo, Cauca/Valle and Putumayo) in which the 
GOC intervened through its PNCRT activities. The midline, structured around the baseline 
completed by DevTech Systems, Inc., examined changes since baseline across the institutional, 
social, and economic indicators related to the interventions. 
The multi-faceted nature of the security and development problems required a coordinated 
response, as the GOC also planned. The joint CSDI activities were designed to address the 
multiple challenges remaining across consolidation zones, including continued presence of 
armed groups and the illicit narcotics trade, enduring poverty, and lack of rule of law. 
Under the assumption that the government would serve as the main builder and executor of 
consolidation, the CELIs developed a strategy for institutional, social, and economic 
development to support, promote, and catalyze the GOC’s PNCRT. All three CELIs included the 
following components to reach the objectives listed in the results framework:
• Support for productive activities and agricultural value chains (technical assistance and 
training in farming, processing, and marketing) and for youth entrepreneurial efforts in 
developing skills in business and commerce; 
• Support for development of credit and financial institutions and programs;
• Infrastructure support in building social goods (schools, roads, health posts, improved 
sanitation, electricity) and productive infrastructure (processing plants, warehouses, and 
marketing centers);
• Assistance in land issues for communities and individuals, including the formalization of 
land titles and private property rights, claims for restitution, regularizing the status of 
public lands, and dealing with special rights reserved for ethnic groups; 
• Social capital and civil society development, with training and support for Community 
Action Councils (the Juntas de Acción Comunal or JACs), producers associations, 
community councils in indigenous and Afro-Colombian areas, and others; and 
 
2 18 municipalities were sampled, but security concerns at baseline excluded El Tarra (Norte de Santander) and Toribio (Cauca).
7 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
• Governance support to strengthen local and regional governments, such as through 
training in planning, budgeting, taxes, royalties, project formulation and management, 
and accountability. 
Economic development
CELI projects promoted rural integration and economic development. To understand the degree 
to which their goals were met and impacts felt, the evaluation gathered information from CELI, 
PNCRT and control zones; from direct and indirect beneficiaries, control groups, and community 
representatives such as personeros, the alcaldías, members and leaders of JACs, and civil society 
leaders. 
Nationwide rural development challenges and infrastructure deficiencies continue to affect CELI 
and PNCRT municipalities deeply, and most respondents in all studied areas report a significant 
decline in their economic conditions since baseline. Poverty is not only at the household or 
individual level, but also at the level of communities, as the municipal and departmental 
conditions (market failings and lack of road access) severely constrain growth. Access to credit 
remains tenuous across the regions, and land tenancy has made few gains. Income has increased, 
roughly at the level of (urban) inflation, but survey and focus group data suggest worsening food 
security. The bright spot amidst this backdrop are direct CELI beneficiaries, who report 
improvements in economic conditions generally and in a set of key variables particularly. The 
challenge of sustainably transporting products to markets remains. 
Direct beneficiaries 
Respondents who had participated in CELI productive projects showed important positive 
effects. First, when asked about whether their economic situation was good or very good, there 
was a positive impact of approximately 8pp (Figure 1).3 There was also an impact of 6pp fewer 
respondents saying that their productive activities generated very little money.
Figure 1. Impacts on perceptions of the economic situation, direct beneficiaries
Variable Direct CELI 
beneficiaries
Report that the economic situation is good or very good 8pp**
Their productive activities generate very little money -6pp**
Source: Authors’ calculations. *** At 99% confidence, ** at 95%, * at 90%.4
 
3 See the Limitations section in the full report and its Annex 3 for a full description of direct beneficiaries.
4 Throughout our reports, the data source is the evaluation’s survey data, unless otherwise noted. Standard significance notations apply: *** for 
99% confidence, ** for 95%, * for 90%. Impact estimation models are found in Annex 3 to the main report, and data in the text are tabled in its 
Annex 10.
8 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
CELI direct beneficiaries reported a 9pp positive impact when they were asked if they received 
support for their productive projects. They were significantly more likely than other respondents 
to associate with others to buy inputs, to process and sell products, to organize transportation, 
and to receive training (Figure 2). However, the impact analysis showed no impacts, positive or 
negative, on direct beneficiaries’ responses about their perceptions of food security and poverty. 
Figure 2. Impacts on reasons they associate with others, direct beneficiaries
Variable
Direct CELI 
beneficiaries
To buy agricultural inputs 6pp** approx.5
To process their products 7.5pp**
To sell their products 10pp**
To transport their products to market 9pp**
To received shared training or technical assistance 13pp**
To have a shared representation as a group 7pp*
To present group projects and seek financing 8pp*
To seek credit 5pp*
There were no impacts for direct beneficiaries in having a parcel of land for productive projects, 
but they did have a higher average number of hectares of land in usufruct with title, at more than 
double the hectarage found among the wider CELI population (Figure 3). There was a positive 
impact on having land in usufruct without title, similar to that of the other CELI beneficiaries.
Figure 3. Impacts on land issues, direct beneficiaries versus all CELI
Variable Direct CELI 
beneficiaries
All CELI zone 
residents
Average hectares of land in usufruct with title 14 to 17 HAs** 7.8 has**
Do you have land in usufruct without title? 2.8pp** 2.8pp**
Did you get your land through restitution? 3.5pp* 0.1pp
Did you get your land through agrarian reform? 1.5pp* 0.9pp*
The CELI projects track as one of their main indicators of performance the incremental or added 
value for key products that they support in productive projects. In the most recent reports, this 
incremental value substantially exceeded the goals set by the CELIs and would significantly help 
the supported associations and productive projects. (Figure 4) 
 
5 Figures are approximate for the direct beneficiaries because there are a range of responses based on which definition of direct beneficiaries is 
used. See the Limitations section and Annex 3 for full details.
9 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 4. CELI performance, increasing sales
CELI USD % of goal
Central $1,723,467 121%
North-South $16,735,470 358%
Source: Quarterly and annual CELI reports; see Annex 5
N.B. CELI Montes de María did not report on this indicator, adopted as that CELI was nearing its end
Regarding credit and financial services, direct beneficiaries were more likely to report having 
received credit, at nearly 11pp greater than at baseline, while the full CELI sample reported 
7.6pp positive impact on having received credit to support their projects. There was also a 
significant 8pp increase in responses indicating they had current debt. 
Overall, direct beneficiaries give the CELIs positive grades in their survey responses. For 
example, when asked to rate the degree to which the projects had met their expectations on a 
scale of 1 to 5 (where 1 indicates that expectations for the program were not fulfilled and 5 
indicates that they were completely fulfilled), direct beneficiaries gave an average grade of 3.7. 
A full 96% of direct beneficiaries affirm that they would recommend participating in a Colombia 
Responde6
 program to family and friends. 
CELI direct beneficiaries and the rural economy. Given the panorama discussed below, CELI 
direct beneficiaries contrast with the rest of the samples in pointing to positive trends in a range 
of economic possibilities and the perception of the problems they face, such as the lack of 
opportunities and the low income provided by their productive activities. Female-headed 
households form an important part of this optimism. The beneficiaries appreciate the technical 
assistance (TA) that accompanies the CELI programs and other state institutions (e.g., SENA), 
and the support promotes assets and capital, which they will be able to enjoy in the long term. 
Wider CELI and PNCRT samples
Perceptions of the rural economy. The positive impacts reported by direct beneficiaries are 
found in an economic context that appears ever more challenging, and CELI municipalities (and 
the rural areas in general) continue to face conditions that keep the majority of households in 
poverty. Across the full sample, in control groups and in CELI and PNCRT municipalities, 
citizens say that their economic situation has deteriorated since the baseline. This is true across 
municipalities, across various survey questions, and by notable percentage changes from baseline 
to midline. Since control respondents show the same trends, no impact of the interventions was 
found. Rather, this trend is likely related to a series of factors beyond the reach of any specific 
intervention which, for rural populations, constitute a poverty trap almost impossible to escape. 
These declines apparent in survey results reveal the plight of households which depend on the 
income of one person, live in precarious housing with limited access to land, and constitute 
contextual conditions not attributable to any intervention. (Figure 5) 
 
6 Colombia Responde is the name by which the CELIs are known in the communities of implementation.
10 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
We learned to establish rubber and cocoa 
as crops. We didn’t know how to graft a 
plant for it to be more productive, now 
we have the technical methods. We can 
improve our parcels by our own means, 
with citrus plants, cocoa, with all these 
products. Their workshops have given us 
knowhow that has enriched us socially
and culturally. The most important thing is 
that our vereda has a different perspective 
of what the countryside can be. 
(Producer, Antioquia, CELI) 
33% 29% 38%
15% 18%
35%
77%
31%
0%
20%
40%
60%
80%
100%
The economic situation
is good or very good
Last week the
respondent ran out of
food
The family is suffering
hunger
The conditions in which
they live are of extreme
poverty
CELI Baseline CELI Midline
Figure 5. Trends in economic situation perceptions, CELI municipalities
Despite the difficulties in the consolidation zones, qualitative data collection revealed that when 
smallholders were asked about their views on their region and its potential, a different and more 
positive picture emerged. This is an indication of the ownership inhabitants feel regarding their 
territories, and, in some cases, a community-wide rejection of illegal activities7 in favor of a 
culture of legality. However, in sum, the results indicate that the economic situation in CELI and 
PNCRT regions parallels that found in rural areas across the country, with the inhabitants of 
remote rural zones faced with the same hardships and needs in surviving day to day, facing 
difficulties even in terms of basic food security. 
Support for productive projects. CELI interventions aim to support sustainable and licit 
income generation. The CELIs made great investments to support smallholder associations, to 
improve the land, to provide technical assistance and inputs for production, and to support 
commercialization. Support for specific crops included a focus on quality, better prices for 
products, organizational development, and particular project support. The CELIs provided 
technical assistance to improve productivity and quality and 
emphasized organizational training and capacity building 
for producer associations. The intervention components 
attempted to counteract the deficiencies that producers 
faced in their regions. Productive projects in CELI 
programs showed evidence of economic changes in the 
households. The projects aim to create long-term value that 
helps producers provide for themselves through their 
productive projects and to strengthen savings and 
investment, as they oblige the beneficiaries to work and 
look after their crops with a long-term perspective, with all 
the costs and sacrifices that this implies. 
 
7 Please see the sections on illicit activities, security, and their contribution to structural poverty, later in the report. 
11 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Marketing and trade for rural producers. TA and marketing support help counteract problems 
related to low profitability. For the most part, only local and regional markets exist in the 
consolidation zones, and this lack of broader markets seriously hampers the tasks of the CELI 
and of the GOC. Deficient and partially depressed markets, pressure from armed groups, and the 
lack of access roads all diminish the effectiveness of productive projects in counteracting the 
incentives to plant illicit crops. Such structural problems – which are the GOC’s direct 
responsibility – diminish the ability of the CELI and PNCRT programs to produce effects that 
are sustainable in time.
In the survey, a series of items enquired about the problems experienced in production and 
marketing products. There were positive impacts on three of these questions in CELI and 
PNCRT municipalities (Figure 6); that is, negative perceptions diminished over time: 
Figure 6. Impacts on perceptions of problems in production and sales, CELI and PNCRT
Variable CELI PNCRT
Abuse by intermediaries -0.076* -0.149
Very low prices for the products -0.062*** -0.094*
Lack of water -0.021 0.100*
The impacts listed in the table represent significant reductions in the perception of problems 
among the respondents, relative to control sites. Though impacts were only found for these three 
questions, the trend over time was a reduction in CELI, PNCRT and control municipalities 
across a range of such problems, though these did not always show significant impact. (Figure 7) 
Figure 7. Trends in perceptions of problems in production and sales, CELI and PNCRT 
Credit for rural producers. Although CELI interventions did not directly provide credit, some 
participating associations created rotating funds to support their members. The revolving funds 
have proven valuable for small groups of neighbors and acquaintances. Implementation costs are 
low and there are important economic benefits to the group. However, according to the
household survey, formal credit availability has fallen in both CELI and PNCRT municipalities. 
12 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Informal credits contained in trading or illegal gota a gota (“drop by drop”) loans continue, 
filling a gap when the broader economy fails to provide viable alternatives. Community credit or 
loans which are a product of programs such as CELI provide an alternative to informality, and 
information gathered in focus groups indicates that revolving savings and credit funds have 
served as important instruments to support capitalization in rural areas. 
Rural roads. The Consolidation zones lack access roads that connect to the country’s most 
dynamic and populated markets.8 The lack of reliable transportation infrastructure means that 
local producers cannot access markets where they can gain prices that are higher than costs. The 
data show very negative perceptions and experiences with road quality. Local market 
deficiencies are closely related to the state of the roads. Building and repairing roads was not 
originally one of CELI’s central axes, but such activities have increased in recent years in a 
broad approach to resolve bottlenecks to commerce. The population recognizes the direct benefit 
of CELI interventions in roads and infrastructure. Local level users, associations and Juntas de 
Acción Comunal (JACs)9
 have taken ownership of road projects. 
Land rights and access. An important element in the CELI vision is to support land 
formalization. This includes helping rural residents get formal titles, supporting displaced 
families and communities to return to their lands, and supporting governmental institutions 
which, ultimately, grant land tenure rights. The continuing informality in ownership rights 
reveals the limitations of the State. The communities deal with this problem through precarious 
ownership rights, based on fragile documents such as sales letters. The absence of formal 
ownership titles renders such documents invalid for use in guaranteeing credit and discourages 
the development of productive projects in the long term. 
Each of the CELIs have provided institutional support to state institutions involved in the land 
claims, while facing the obstacles that such institutions have been weak, without clear and 
consistent processes. The CELIs have maintained records pertaining to restitution cases and the 
formalization of property ownership. They have, in large part, reached or exceeded their goals. 
The advances in the titling procedures in the CELI and PNCRT municipalities continue to be 
limited in the face of the huge scale of the problem. Although the CELIs can support such 
processes, it is ultimately a matter the State must resolve. 
Recommendations
1. Set realistic expectations for impact at a level that implementing partners can affect. The 
impact indicators for CELI were cast at a high level of generality in the results framework, 
relative to the investment and interventions, the two years of implementation, and the 
problematic contexts of operations in conflict zones. Given that the economic situation reflects 
 
8 Data from the Ministry of Transport. In departments in which the CELI works, reduced coverage is confirmed (8-16% of the area), reduced 
paved area (<7%), and poor quality (1-13% in “good” state according to the Ministry’s classification).
9 JACs are the principal institutional (and constitutionally recognized) entity of community governance at the local level.
13 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
regions’ critical structural deficiencies, it is important to not consider lack of wider impact as 
representing a failure in programming. 
2. Make simultaneous improvements in production and marketing. Increased production must 
go hand-in-hand with market improvements and private sector alliances. Small, focused pilots 
that treat these interconnected structural failings holistically, by donors or the GOC, are 
preferable to diluting the effects by intervening across broad geographic areas with expansive 
mandates. 
3. Support wider markets. Support commercialization designed to improve both local markets 
and those that export goods to other regions, which may provide a solution to local 
overproduction and low prices. Programs must bring producers and buyers together in the 
productive process to work with broader marketing efforts to improve quality. Involving banks to 
increase access to working capital and providing better market information, such as through 
mobile phone access to up-to-date market information on local and regional conditions, can help. 
4. Plan for sustainability in productive projects with delayed returns. CELI and USAID should 
ensure that program support for long-term productive projects also support food security. 
Productive projects must combine short- and long-term strategies to ensure food security, cash 
flow and project sustainability. 
5. Reach Female Heads of Household (FHHs) with tailored interventions. . Given that FHHs 
and MHHs learn and use their learning in different ways, models should reflect this difference 
and provide different services to take advantage of the characteristics and behavior of each. 
6. Maintain activities to expand responsible use of financial services. Revolving funds are a 
suitable instrument to deal with communities’ borrowing needs but they must be protected 
against the risk of non-payment. To strengthen governance principles to guarantee sustainability, 
a system that monitors the funds, sets up discussion fora and experience exchanges, creates 
incentives for performance and draws attention to successful cases is necessary. 
7. Continue focused work on roads. The current advances made by CELI N/S and CELI Central 
in mobilizing communities for tertiary road maintenance provide a space for experimentation for 
community involvement in this arena in the future. The CELI programs should continue the 
successful support they provide to the JACs in institutional strengthening and participation to 
undertake community infrastructure efforts. 
Social development
Strengthening civil society and social development is considered fundamental to the 
consolidation of conflict-affected areas in Colombia. Social development is understood as the 
population’s participation in civil society, trust between peers, and trust between citizens and the 
state. The CELI interventions are intended to develop and strengthen social capital in target 
populations as a tool for territorial consolidation. 
Organizational capacity. The CELIs documented substantial and important increases in 
organizational competency scores, especially in business initiatives and services, surpassing 
program targets. An outward focus on markets and services appears to be important to 
sustainability, balanced with support to organizations’ internal management and other features. 
(Figure 8) 
14 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 8. CELI Performance: Average improvements by category, CELI N/S 
Social capital: Trust. From citizen perspectives, there were no increases in the social capital 
index and a slight statistically significant decrease in the trust component, led by a greater drop 
among FHHs. Trust in 25 state institutions fell across the board in all groups surveyed (CELI, 
PNCRT, and controls), except trust in the personería. As the effect is generalized, it is not 
attributed to any intervention; rather, the deterioration likely comes from sustained, external 
events (absence of state presence, lack of services, scandals, etc.). (Figure 9) 
Figure 9. Impacts on the Social Capital Index, CELI and PNCRT
Variable CELI PNCRT
Social Capital Index -0.24 -1.16
Trust component -0.99** -1.00
Participation component 0.74** -0.16
Source: Authors’ calculations. *** At 99% confidence, ** at 95%, * at 90%.10
Qualitative data challenged the depth of this decline in trust. With some exceptions, trust in the 
institutions is a constant among respondents in focus groups and interviews. They report that 
trust and solidarity do exist between neighbors and that this has facilitated the development of 
social groups. In regions where armed groups operate, trust between peers not only improves 
quality of life but also permits the creation of associations and the strengthening of the JACs. 
This finding is transversal inasmuch as it is reflected in all kinds of efforts to organize, whether 
productive associations, victims’ organizations, cultural or youth groups, or others. 
The difference between the survey and qualitative interviews may be explained by the fact that 
the surveys represent an average opinion of the population while the qualitative information 
comes from a purposively selected sample of active, participating people. 
 
10 Throughout, the data source is the evaluation’s survey data, unless otherwise noted. Standard significance notations apply: *** 
for 99% confidence, ** for 95%, * for 90%. Impact estimation models are in Annex 3, and data in the text are tabled in Annex 10. 
15 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Social capital: Participation in associations. The survey data show that households participated 
more in some – but not all – types of community organizations. The impacts were slightly more 
significant in CELI municipalities, and among MHHs (but not FHHs). (Figure 10) 
Figure 10. Impacts on participation in community activities, CELI and PNCRT
Variable CELI PNCRT
Oversight groups (veedurías) 0.01*** 0.004
Community meetings in general 0.07** -0.003
Voluntary groups 0.06*** 0.05*
Sporting or cultural groups 0.03*** 0.03*
Security groups 0.005*** 0.005
Youth groups 0.007** 0.004
Women’s groups 0.02* 0.04***
Yet in spite of the overall positive impact, fewer households participated in JACs at midline: 
from 74% at baseline to 57% at midline in the CELI zones, and from 88% to 64% in the PNCRT 
municipalities. A variety of reasons, from the time requirements for participation to the 
perception that local politicians had co-opted the JACs for their own purposes, were cited in 
explanation. 
Nevertheless, each CELI developed strong commitments to work with the JACs, including 
directly contracting with them to carry out community projects and supporting their institutional 
strengthening. Between CELI N/S and CELI Central, dozens of JACs have taken on 
responsibility for managing and fulfilling the contractual terms of direct grants, providing these 
civic organizations with important experience in community management and infrastructure 
construction and maintenance. 
Despite the fact that on occasions these entities are considered to have been co-opted, they 
continue to represent the gateway to state services and programs. Where the security situation 
has allowed, and where the JACs have the confidence of the population, this kind of social 
organization is central to political life and to participation in the veredas. In these cases the JACs 
play a central role in providing access to, and enjoyment of, the services offered by the state in 
terms of political participation, infrastructure improvements, and social control. In the veredas of 
geographically distant municipalities the JACs are the principal spaces in which disputes 
between neighbors are resolved, through direct negotiation. By having this local resource, minor 
social conflicts are resolved at the vereda level, without the need to take cases to municipal or 
national bodies. 
16 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Associations and productive projects 
The CELI projects have supported producer associations in several ways: technical assistance 
and training; support to develop marketing and commercialization (also often involving 
training); and assistance with financial services to support production and marketing. As the 
implementer of CELI Central put it, “the focus has increasingly been on empowering and 
mobilizing smallholder producer associations, and increasing private sector investment in rural 
communities to create economic opportunities for community stakeholders within the context of 
licit markets and the rule of law.”11
Surprisingly, the survey revealed an apparent decline in CELI municipalities in involvement with 
producers associations. Respondents reported a significant 7pp reduction in their participation in 
producer associations in the CELI municipalities, whereas in the PNCRT municipalities, no 
significant variation occurred (Figure 11). On the other hand, those who did participate reported 
greater involvement in decision-making throughout all survey groups. In both CELI and control 
municipalities, participating respondents attested to a significant increase in their productive 
capacities across several dimensions. For CELI municipalities, participants affirmed joining 
producers associations because of the support they provided in processing products, sales, 
transport, and training. On the other hand, respondents recognized that some producer 
associations were created largely to secure benefits from government and donor funding, with 
little reference to their sustainability.
Figure 11. Impacts on participation in producer associations, CELI and PNCRT
Variable CELI PNCRT
Household belongs to or participates in producer or marketing associations -0.068*** -0.009
Women’s, victims’, and cultural associations 
Participation in women’s organizations and victims’ associations has 
increased. The social capital constructed as a result of the work of victims’ 
associations has permitted the creation and strengthening of other kinds of 
organizations, encouraging an expansion of their aims, fomenting a sense 
of belonging and identifying new resources. Ethnically-oriented cultural 
associations were found to strengthen a sense of identity and build social 
capital and cohesion. This has enabled the reconstruction of community 
values damaged by the armed conflict. Returning to activities such as 
dance, music, or traditional artisan production has permitted communities 
to create a cultural identity in which they are able to recognize themselves 
and feel pride in their traditions; it has also acted as a form of resistance 
against armed actors. 
 
11 CELI Central Quarterly Report Q1 FY2016, p. 5.
We created a project called 
the Son de Negros. We 
focused on the recruitment 
carried out by illegal armed 
groups and we are going to 
include this in our danza negra, 
an African dance that we have 
been dancing since the time of 
the palenque of María la Baja. 
There is a sense of connection 
with the mountains, with my 
grandparents. I remember that 
my grandfather was black and 
was a drummer in the danza 
negra. (Bolívar, CELI)
17 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Recommendations 
1. The GOC must ensure security. The intervention zones must be secure before interventions 
on trust will have any effect. 
2. Support JAC legitimacy. Leaders should continue to be trained in board election processes 
and community participation tools to counteract loss of legitimacy. Monitoring the quality of 
community associations such as the JACs, adapting tools to meet organizational and community 
expectations for these groups, and ensuring that programs involve the JACs, communities, the 
participation of victims, women, and other marginalized groups, as well as local government 
should be priorities. 
3. Loosen the requirement that beneficiaries must be grouped into associations. In future 
programs, the requirement for obligatory membership in an association to receive benefits should 
be reviewed to determine how best to structure incentives for participation. Otherwise, 
associations may have little power to consolidate and become self-sustaining. 
4. Increase collaboration with associations whose objectives may not be focused on productive 
activities. Pre-existing social organizations with aims that are not productive are also candidates 
for interventions like CELI because of their high levels of social capital, and they are increasing 
in number in the zones of intervention. Such organizations should also be able to gain help to 
become autonomous and financially sustainable. 
Governance and institutions
Participation, accountability and trust in public management
Citizen participation at the ballot box (nearly 90%) and in political organizations remains high 
and increased in the last presidential election across all groups interviewed, but electoral 
challenges such as vote buying and other clientelistic practices persist. Combining survey 
questions in an Accountability Index shows declines since baseline across the regions under 
study. Changes were uniformly negative for CELI municipalities in how the public rates the 
accountability of municipal governance, with a drop in score between 5% and 28% (Figure 12). 
In specific problem areas, the only notable drop in perceived problems in CELI municipalities 
was in the delivery of public services (water, electricity, sewer, etc.). Citizen oversight is low and 
has only increased slightly.
Figure 12. CELI Performance: Accountability Index, by region 
Ind. Description Disaggregations Baseline Midline % Change Source
DO3-
028 
Level of 
accountabili
ty in CELI 
municipaliti
es
Total 23.8 19.63 -18% Household 
survey
• Baseline 
2013, 
Econome
tría S.A. 
• Midline 
2015, 
Proyecto 
EVAL
Montes de 
María 18.6 17.74 -5%
Central 27.8 24.96 -10%
Norte 25.5 18.89 -26%
Sur 22.6 16.17 
-28%
Participatory mechanisms are not yet widely effective. Meetings convened by municipal 
institutions are perceived not to respond to needs of community leaders. Community members 
18 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
also rate the performance of their local governments poorly, and their opinion of mayoral 
management in both CELI and PNCRT municipalities declined significantly. In CELI and 
PNCRT municipalities, trends are negative regarding government management at all levels. 
They may be ineffective at implementing new resources, may be creating new and unmet 
expectations, and may simply not have changed practices in response to interventions. But the 
effect on citizens is significant. Where CELIs intervened successfully on this front, there are 
localized results – but the generalized impact the evaluation was designed to detect is not 
present.
The confidence in Colombian local institutions worsened in recent years with the survey 
showing decreasing confidence in all levels of government. Trust dropped from baseline to 
midline across the full sample. The decline in confidence is parallel across governmental 
institutions: police, army, navy, justice; national, departmental, local, city council and JACs; 
ombudsman’s offices; and advocacy and support bodies such as the national training institutes. 
An exception to this is the work of the personerías. Citizens in CELI and control areas recognize 
the personería as the main institution with which communities interact, particularly because of 
its role in the Victims’ Registry and in complaints about human rights violations. CELI work 
with the personerías to strengthen the registry of veedurías has been positive. 
Social service delivery
The GOC’s Integral Performance Index (IDI for its Spanish initials), applied since 2006 to assess 
the management of territorial governments, tracks the performance of mayors and governorships 
in six dimensions: effectiveness, efficiency, legal requirements, administrative capacity, fiscal 
performance and management.12 It shows increased scores from baseline and midline for all four 
groups of municipalities studied – CELI, PNCRT, and their respective control groups. The 
strongest gains are in efficacy, observing legal requirements in budgeting, and management. 
CELI municipalities continue to have somewhat lower overall ratings, however, than the PNCRT 
zones. (Figure 13) 
 
12 The DNP explains the results, the methodology and the sources of information in the documents in this website: 
https://colaboracion.dnp.gov.co/CDT/Desarrollo%20Territorial/Docuemnto%20Desempe%C3%B1o%20Integral%202013.pdf
19 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 13. Secondary data: Integral Performance Index, 2012 (left) and 2014 (right) 
This contrasts with the Governmental Capacity Index, developed specifically for this impact 
evaluation. The GCI measures capacity and service delivery from citizens’ perspectives, rating 
access and quality of services, quality of public policy formulation, and perceptions of municipal 
actors’ honesty. Citizens report no progress on the GCI indicators. The analysis shows that 
governmental capacity is a desirable characteristic but difficult to achieve, and very subjective, 
as it responds to structural and historical factors that can escape the reach of CELI interventions. 
However, the stagnation of perceptions of government capacity should be taken into account as a 
sign of the failure of the policies being implemented to consolidate these regions. Results for 
both CELI and PNCRT show declines since baseline in CELI municipalities of between 5% and 
17% in absolute terms.13 (Figure 14) 
Figure 14. CELI Performance: Governance Capacity Index, total and regional 
Ind. Description Region Baseline Midline % Change Source
DO
3-
013 
Governance 
capacity 
index 
Total 40.7 36.2 -11% Household 
survey
• Baseline 
2013, 
Econometría 
S.A. 
• Midline 2015, 
Proyecto 
EVAL
Montes de 
María 39.2 37.3 -5%
Central 39.8 36.2 -9%
Norte 42.6 36.2 -15%
Sur 40.3 33.4 -17%
 
13 The difference in trends between DNP data and survey data can be explained because the DNP data measures internal, short-term and 
administrative decisions and processes that are not necessarily visible to the population in the same period of time.
20 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Social service indicators, tracked at baseline and midline, had little to do with CELI 
interventions. But they do show that social service delivery remains a major challenge, with 
citizen perceptions of access and quality objectively very low, and lower than at baseline. 
Citizens’ opinions of social service access and quality have not changed in CELI municipalities.
In absolute terms, HHs in CELI municipalities report a worsening of schools, while there were 
some improvements and even impacts among PNCRT municipalities. 
Recommendations 
1. Limit the scope of evaluation to the factors under implementers’ manageable interest. The 
design of the impact evaluation, based on a vision of a cross-government integrated program in 
CSDI, proposed far higher impact gains than can be expected from the CELIs alone. 
2. Use prior diagnostic activities with potential target municipalities to uncover clientelistic 
and corrupt practices. Detection of clientelistic practices should trigger GOC involvement to 
precede donor interventions. This suggests focusing first on supporting control bodies such as the 
Attorney General, the Comptroller or local veedurías, before distributing resources for specific 
projects, and financing projects that meet high standards for transparency, community 
participation and oversight. Avoid implementing with intermediaries, which may increase risk.
3. Strengthen accountability to improve public policy orientation, gain citizen trust and provide 
management oversight through training; expanding spaces for greater review of management and 
monitoring; supporting the application of sanctions for corruption; encouraging the development 
of a culture that discredits corrupt practices; simplifying administrative processes; and 
developing participatory budgets.
4. Continue to work to strengthen the personerías, who have become essential actors in 
conflict-affected zones, particularly with victims. 
5. Continue to work with municipalities to improve service delivery, and involve citizens in 
oversight activities. Focus on pilot-level activities with significant civil society involvement, and 
scaffold local governments with consistent TA. Use scorecards to involve citizens in rating 
service changes, and learn from failures as well as successes which practices work best in 
different sectors and circumstances. 
6. Work with the GOC to strengthen the IDI. The IDI lacks important considerations of the 
effects and perceptions of service provision. The national imperative to decentralize is an 
incentive to view only those indicators that show a picture of progress, but the rubber meets the 
road in citizen perceptions of those services. 
Security
At the outset, it is important to note that the CELI interventions undertook no activities in the 
sectors of security and justice reform. The impact evaluation was designed when the CELIs were 
part of the larger CSDI set of activities, along with a promised whole-of-government approach 
from the GOC. As a result, impacts in these sectors are outside of the CELIs’ manageable 
interest. 
Over the last decade, there has been a drastic transformation in security conditions in Colombia 
in both rural and urban areas. Reduction in some crimes has converged with an increase in 
21 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
others, the recovery of legality in some regions, and the return of others to criminal control. The 
evidence includes a security index, analysis of coca cultivation, trends in major human rights 
crimes, and new criminal dynamics. 
CELI and PNCRT interventions continue to operate in challenging security conditions but these 
have changed with new dynamics of crime and new threats – less overt and major crimes and 
more coercive, extortionate behavior that affects the population in different ways. Data collection 
has not yet caught up with these changes. Coca cultivation has doubled nationally and the CELI 
and PNCRT zones show related changes in perceptions and opinions about coca as an economic 
activity. Military and police force levels have changed drastically, and justice services continue 
to be absent in many areas under study. The zones are unstable as before, with the lack of state 
presence that has – and continues to – leave citizens and their livelihoods at risk.
EVIDINCE14 created an index of insecurity, synthesizing averages of criminal acts of major and 
low impact (thefts, robberies and homicides) with subjective elements of the perception of 
security, such as the appearance of graffiti in the communities. 
Figure 15. Secondary data: Insecurity Index
Insecurity dramatically increased in CELI municipalities in 2014, placing it significantly above 
the indicator for the control municipalities (Figure 6.15). Both CELIs and their controls saw 
declines in insecurity in 2015. In PNCRT municipalities, the insecurity index shows a steady 
 
14 EVIDINCE is a research effort associated with the Empirical Studies of Conflict (esoc.princeton.edu), a multi-institutional collaboration 
studying conflict and insurgency at the sub-national level. 
Source: EVIDINCE. The index combines data on homicide, robbery, larceny, illegal mining managed by armed illegal groups (IAG), 
and the presence of graffiti alluding to IAG in the municipio. Treatment 1 (T1) are CELI municipalities, and C1 their controls. T2 
are PNCRT municipalities, and C2 their controls.
22 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
92%
94%
94%
93%
60%
77%
79%
76%
79%
61%
0% 20% 40% 60% 80% 100%
Coca, poppy and marijuana shouldn't be
cultivated because they're illegal
Producing illict crops negatively affects our
families and community
Growing coca is not the only way for
families to guarantee their subsistence
My family and friends think growing coca is
bad
Our community is willing to denounce the
presence of illicit crops
CELI Baseline CELI Midline
decline from the end of 2013 until the middle of 2014, when there is a slight rise, which is also 
present in their control municipalities. Since August 2015, the EVIDINCE insecurity index 
shows a generalized decline in all municipalities studied. This is explained by the de facto truce 
in the final stages of the peace talks between the GOC and the FARC. 
In terms of coca production, after a dramatic decrease between 1998 and 2004, illicit crops (ICs) 
remained steady until 2012, when a pronounced increase in ICs was noted. Survey results 
suggest an increased perception of ICs since the baseline, especially in the PNCRT 
municipalities. Also notable are the survey results showing the erosion of respondents’ baseline 
agreement that illicit crops should be avoided – in CELI municipalities, respondents’ agreement 
that cultivating ICs dropped from over 92% to 77%, for example. (Figure 16) 
Figure 16. Trends in perceptions about illicit crops in their communities, CELI
These impacts reflect a national trend, which has engendered a range of hypotheses as to its 
origins: i) the low profitability of other legal and illegal activities, such as illegal mining; ii) the 
reduction in aerial spraying and manual eradication campaigns; and iii) a boom in cultivation 
prior to the signing of a peace agreement that would enable the legalization of capital obtained 
illegally (a practice documented by some FARC commanders) (Wilches and Garzon, 2016). 
Crime statistics. The gradual dismantling of paramilitary groups and the new operational logic 
of the BACRIM [criminal bands], added to the commencement of the peace talks with the 
FARC, have helped to reduce the number of cases of human rights violations throughout the 
county, especially high-impact crimes, such as massacres and acts of terrorism. There has been a 
61% reduction in massacres in the CELI municipalities – similar to the national trend – which 
means an impact of around 7% of the decline in the national total of massacres. Furthermore, 
there has been a parallel reduction in guerilla attacks over the last decade in the control and 
treatment municipalities.
Following the fragmentation of paramilitary groups (converted into BACRIM) and their sporadic 
criminal alliances with the guerrillas, extortion has become the fastest growing crime in 
Colombia. However, communities’ perceptions of extortion as a serious crime are low, and are 
on a downward trend in both CELI communities and their controls. The proportion who believe 
that extortion is one of the most serious security-related problems declined from 9.6% at baseline 
to 5.5% at midline. 
23 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 17. Secondary data: Average number of attacks by guerrilla groups, per municipality15
Source: Conflict Observatory, Universidad del Rosario
Military presence. It is not possible to gain access to operational reports on the availability of 
public forces in the areas studied, but the evaluation survey gives one perspective on military 
presence. Survey results show a drastic decline in communities’ perceptions of military presence 
in CELI, PNCRT and control municipalities. From 87% at baseline, only 37% of households in 
CELI municipalities asserted that there were military forces in their communities at midline. The 
trend was parallel in their controls: a decline from 83% to 29%. PNCRT municipalities showed a 
similar but less drastic trend, from 77% to 40%. 
While perceptions of military presence fell drastically, in the CELI municipalities, there was a 
6pp positive impact when asked whether military services had improved over the previous two 
years (32% to 39%), with no change in control municipalities. 
Police presence. Figures compiled by EVIDINCE in municipal panel data show a leveling off of 
police presence in treatment and control municipalities over the past three years. The survey 
results support this decline in perceptions of police presence across the full sample, as drastic as 
the decline in military force presence. CELI municipalities and their controls saw two-thirds of 
their respondents citing the presence of police services at baseline, and only a quarter at midline.
 
15 The data available from the CERAC and the Ministry of Defense indicate a reduction of 92% and 95%, respectively, in the 
number of FARC attacks in 2015, which shows the impact of the negotiation process on the reduction of guerrilla actions.
24 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 18. Secondary data: Police presence
Source: EVIDINCE municipal panel
Availability of justice services. Survey respondents from CELI, PNCRT and control households 
also report a precipitous decline in the presence of formal justice systems in their municipalities, 
at even a greater rate of decline than the lack of security presence: from some 70% to 80% at 
baseline, to 8% to 17% at midline. There was no significant change in their perceptions of the 
quality of the services. Fewer respondents say they would access justice through formal channels 
than at baseline in absolute terms, even through the JACs (which appear to be a source of 
conflict resolution services at the local level). However, trends were the same across the full 
sample, so these changes appear across all regions.
Recommendations
1. Guarantee security conditions in all intervention zones. USAID should consider using its 
international cooperation resources to further encourage the GOC’s defense sector to prioritize a 
minimum level of security, both in forces and conditions, in the territories where CELIs and
future programs operate.
2. Support the creation of specialized police units and judicial police in the intervention zones, 
coordinated with the military and focused on specialized units and properly vetted judicial 
police. 
3. Adjust interventions to reach the areas with the greatest coca cultivation. Very few of the 
coca growers access state programs, and productive projects designed for these citizens do not 
generate the necessary incentives for returning to the path of legality.
4. Update security indicators to capture changing dynamics, capturing new criminal dynamics 
and providing an updated diagnosis of the security challenges, including new forms of 
criminality.
25 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
INTRODUCTION
After decades of conflict in rural areas with illegal armed groups (IAGs) and with increasing 
success in limiting their activities, the Government of Colombia (GOC) launched the National 
Consolidation and Territorial Reconstruction Policy (PNCRT) in 2009 to increase its presence 
and to reduce inequity in the historically marginalized conflict zones where IAGs and coca 
production coexist. 
The concept of “consolidation” refers to the expansion into the conflict zones of Colombia of the 
legitimate presence of the State. The policy, designed for implementation once minimal 
conditions for public safety were established, would create the necessary institutional 
mechanisms to provide basic human rights protections to the populations in those zones 
historically affected by conflict and illicit crops. The PNCRT was oriented to establishing the 
presence of State institutions capable of efficiently bringing necessary public goods needed for 
development. 
The policy conceives of social, economic and institutional consolidation and social 
reconstruction as the result of a process of mobilization and continual strengthening of three 
pillars: (1) the territorial institutionalization of state institutions focused on guaranteeing security 
and creating the necessary capacities to institutionally manage the area; (2) the strengthening of 
citizen participation; and (3) the fostering of regional integration.16 The interaction of these 
three pillars on a base of territorial security and combined with a coordinated collaboration with 
state entities would be the path to guarantee the fundamental rights of the citizenry.
The U.S. Embassy, for its part, developed the Colombia Strategic Development Initiative (CSDI) 
to support the strengthening of government presence in strategically important parts of the 
country and reduce inequity in historically marginalized regions where illegal armed groups and 
coca production coexist. CSDI was conceived as a whole-of-government, geographically￾focused approach aimed at maintaining advances in security and development and at the 
reduction of illicit crops and the presence of illegal armed groups in the conflict zones. It 
included significant investments in security, antinarcotics, access to justice, and eradication 
activities:
USAID is already coordinating closely with other Embassy entities, including the Narcotics 
Affairs Section (NAS), the Military Group, and the Department of Justice as well as the Political 
and Economic sections of the State Department to develop an integrated USG approach for the 
priority CSDI consolidation zones. Programs will include peace and security, illicit crop 
eradication, access to justice, justice sector reform and humanitarian and development 
interventions targeted at a wide range of populations.17
 
16http://www.consolidacion.gov.co/themes/danland/descargas/entidad/planeacion/POLITICA_NACIONAL_DE_CONSOLIDAC
ION_Y_RECONSTRUCCION_TERRITORIAL_PNCRT.pdf. 17 USAID Colombia Solicitud de Aplicaciones Número 514-09-000005, “Montes de Maria – Enhanced Livelihoods Initiative;” 
Septiembre 15 de 2009, p. 5.
26 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
The multi-faceted nature of the security and development problems required a coordinated 
response. The joint CSDI activities were designed to address the multiple challenges remaining 
across consolidation zones, including continued presence of armed groups and the illicit 
narcotics trade, enduring poverty, and the lack of rule of law. Working with the GOC and 
PNCRT, the goal of uniting USG agencies toward consolidation was: 
To reach a “tipping point” where state presence and positive perception of the government is 
sufficiently well-established so that communities decide to align with their government rather 
than illegal armed groups and the coca industry.18
USAID/Colombia designed the Colombia Enhanced Livelihood Initiative (CELI) as one of the 
CSDI components and as the backbone of this integrated plan to bolster the PNCRT by 
supporting efforts to consolidate the State’s institutional presence in areas of the country under 
the control or influence of insurgent groups engaged in the production of illegal crops, drug 
trafficking, kidnapping, extortion and terrorist activities.
The basic premise of the CELIs is that the armed conflicts and illicit economy are closely linked 
to the weak presence of the State, and their activities are oriented to strengthening state 
institutions, generating confidence in them, constructing relations between communities and their 
local governments, and supporting the improvement of sustainable ways of life. The CELIs 
clearly are no substitute for the State. The existence of a minimum provision of public goods, 
property rights and justice services in the consolidation zones is a sine qua non condition for 
enabling the CELIs, and any international cooperation effort, to find a place in which they can 
operate and ensure that the resources will serve as catalysts for development and consolidation. 
The CELIs, or Colombia Responde as the projects are known in the implementation zones, 
supported the GOC through three pillars that are similar (though not identical) to those of the 
PNCRT: 
Figure 1.1. The three pillars of the PNCRT and their CELI peers
PNCRT pillar CELI pillar CELI pillars defined as:
1 Territorial 
institutionalization
Strengthening 
institutional 
development
Improving local government response; formalizing 
property rights; processing cases of victims’ property 
rights; implementing a rural development strategy
2 Citizen participation and 
good government
Strengthening social 
development
Strengthening civil society organizations (CSOs); 
promoting alliances between the community and the 
State
3 Regional integration Catalyzing economic 
development
Implementation of rural development programs and 
improving competitiveness 
 
18 Ibid.
27 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Three implementers were convened to 
undertake the work in four of the seven 
corridors comprising the Consolidation Zones.
The first, Chemonics International, has 
worked through the CELI Norte/Sur project in 
the northern (Bajo Cauca in Antioquia and Sur 
de Córdoba) and southern (Tumaco) regions; 
the second, ARD Tetra Tech, implements the 
CELI Central project in 23 municipalities of 
the departments of Cauca, Valle del Cauca, 
Caquetá, Meta, and Tolima; and the third, 
Global Communities, finished a similar set of 
tasks in early 2015 in the municipalities of
Ovejas and San Onofre (Sucre) and Carmen de 
Bolívar and San Jacinto (Bolívar) through the 
CELI Montes de María (MdM) project. 
With this shared mandate and the assumption 
of the GOC’s leadership role, the CELIs 
developed strategies for institutional, social 
and economic development in these four 
regions. 
Within USAID’s Country Development 
Cooperation Strategy (CDCS) for Colombia, 
the CELIs were first situated within the 
Mission’s Development Objective 1: Civilian 
government present in CSDI zones 
consolidated, which focused on supporting the process of consolidation – strengthening of local 
government, formalizing property rights, social development and the alliance between 
community and state, and rural development. In the updated CDCS FOR 2014-2018, the CELI 
projects were placed under a revised DO 3, Improved conditions for inclusive rural economic 
growth. 
CELI investments are targeted to Intermediate Results in supporting Development Objective 3, 
and have much in common with their original results framework in spite of the shift toward 
economic activities. The current framework, tied to the CELIs’ common Performance 
Management Plan (PMP) prioritizes the following:
IR 3.1 More equitable and secure land tenure
Sub IR 3.1.1 Land restituted to displaced victims of conflict
IR 3.2 Increased private and public investment in the rural sector
Sub IR 3.2.1 Strengthened local governments’ capacity to access and manage public funds for 
 productive infrastructure
Sub IR 3.2.2 Increase private sector investment in target rural communities
IR 3.3 More effective producer associations benefitting smallholder farmers
Sub IR 3.3.1 Improve institutional capacity to producer associations
CSDI and PNCRT zones in Colombia
Source: 2013, CSDI Impact Evaluation Baseline Report prepared 
by DevTech Systems, Inc. and Econometría, S.A.
28 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Under the assumption that the government would serve as the main builder and executor of 
consolidation, the CELIs developed a strategy for institutional, social, and economic 
development to support, promote, and catalyze the GOC’s PNCRT. All three CELIs included the 
following components to reach the objectives listed in the results framework:
• Support for productive activities and agricultural value chains (technical assistance and 
training in farming, processing, and marketing) and for youth entrepreneurial efforts in 
developing skills in business and commerce; 
• Support for development of credit and financial institutions and programs; 
• Infrastructure support in building social goods (schools, roads, health posts, improved 
sanitation, electricity) and productive infrastructure (processing plants, warehouses, and 
marketing centers);
• Assistance in land issues for communities and individuals, including the formalization of 
land titles and private property rights, claims for restitution, regularizing the status of 
public lands, and dealing with special rights reserved for ethnic groups; 
• Social capital and civil society development, with training and support for Community 
Action Councils (JACs), producers associations, community councils in indigenous and 
Afro-Colombian areas, and others; and 
• Governance support to strengthen local and regional governments, such as through 
training in planning, budgeting, taxes, royalties, project formulation and management, 
and accountability. 
The CELI investments are shared across multiple municipalities and several different sectors. 
For example, through September 2015 CELI Norte/Sur invested US$56 million in activities 
across 16 municipalities with financial inputs broken down as follows: 51% in economic 
development, 18% in infrastructure, 17% in governance, 11% in social capital, and 3% in land￾related activities.
Broad dispersion of development resources also took place in the zones under the PNCRT: the 
government designed a strategy based on the interaction and coordination of public sector 
entities, but in practice each spent its resources in accordance with its own priorities, interests 
and historical allocations. Public sector coordination is extremely difficult to accomplish because 
doing so requires personnel resources and a level of institutional authority that challenges 
existing patterns and interests, both within independent ministries and agencies and with local 
governments. Observers also point to weak local institutional capacity. Furthermore, these same 
competing interests made it difficult for the GOC to fully realize its prioritization within the 
Consolidation Zones, especially given demands and pressures for nationwide development 
investments. Thus, the strategy involved good intentions but faced grave challenges in 
coordination and implementation. 
Obviously, the CELIs are no substitute for the State. They were called upon to provide support to 
the pillars of the PNCRT. The existence of a minimum provision of public goods, property rights 
and justice services in the Consolidation Zones is an essential condition for enabling the CELIs, 
and any international cooperation effort, to find a place in which they can operate and ensure that 
the resources will serve as catalysts for development and consolidation. 
29 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
EVALUATION PURPOSE & 
EVALUATION QUESTIONS 
The three CELIs were intended from the beginning to be the subject of a rigorous impact 
evaluation consistent with USAID’s agency-wide Evaluation Policy, and in 2012 USAID 
charged its monitoring and evaluation (M&E) contractor DevTech Systems, Inc., to design an 
impact evaluation of the set of CSDI interventions and collect the baseline survey data. DevTech 
worked with Econometría, S.A., to carry out this work.19 The evaluation design included two 
subsequent rounds of evaluations: a midterm evaluation, and a final evaluation following 
completion of the program. 
The current independent midline impact evaluation of the results of CSDI efforts took place after 
three years of program implementation for the North/South and Central Corridors and four years 
for Montes de María. USAID charged its Evaluation and Analysis for Learning project (EVAL), 
a contract with Management Systems International, Inc. (MSI), to carry out the midterm measure 
of the evaluation and to report on the CELI program impacts to date. This report represents that 
effort, based on the baseline evaluation design, additional qualitative fieldwork, interviews with 
the CELI teams, and an extensive survey of participants. EVAL hired a subcontractor, the Centro 
Nacional de Consultoría (CNC), to collect field data and provide analysis. 
The impact evaluation was originally linked to the full range of interventions within the 
framework of the CSDI. Unfortunately, funding for the other programming was curtailed, 
leaving the CELIs alone as only one component of the full strategy. Given this, it should not be 
expected that an impact evaluation designed for the larger initiative would detect results of 
aspects of the CSDI strategy that were never carried out.
Evaluation rationale
The evaluation took place in the 33 municipalities targeted by USAID and CSDI as well as in 
1620 municipalities from three regions (Catatumbo, Cauca/Valle and Putumayo) in which the 
GOC intervened through PNCRT activities. The midline examined changes since baseline across 
institutional, social and economic indicators related to the interventions. 
The key intended audiences of the final evaluation report are the USAID/Colombia Mission 
Front Office, Technical Offices, and Program Office; and implementing partners at prime
contractor and subcontractor levels.
The following three principles guide the use of this evaluation: 
 
19 DevTech Systems, Inc. and Econometría S.A. 2013. “Impact Evaluation Methodology of the USAID Strategy in CSDI Zones - 
Final Methodology Report.”
20 Eighteen municipalities were sampled. Security concerns at baseline eliminated El Tarra (Norte de Santander) and Toribio 
(Cauca).
30 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
1. Measure the change in impact indicators, and determine the extent to which the CELIs 
are achieving their goals with target populations. 
2. Inform decision-making processes with timely data, to identify components needing 
strengthening to achieve program goals. 
3. Identify best practices and lessons learned to date, to inform future USAID programming. 
Evaluation questions
The 2012 baseline design identified the evaluation questions; these were later amended per 
USAID request (see Annex 2 for the SOW) and are presented below.
Figure 1.2. Evaluation questions 
QUESTION EVALUATION TYPE STRATEGIC LEVEL
1. What specific impacts have the CSDI interventions 
produced among the target population in the areas of 
interventions? 
Impact IR3.1, 3.2 and 3.3
2. Do CSDI projects show the expected level of 
performance in support of the consolidation process? Performance Cross-cutting
3. Does the GOC show expected level of investments in 
consolidation areas following the assistance from 
USAID in PCNRT municipalities?
21
Performance IR 3.2
4. What are the CELIs' specific effects of in changing the 
capacity and service delivery of local government? (To 
serve local people, be accountable, expand social 
services.)
Impact Sub-IR 3.2.1
5. What changes have been produced by CELI 
interventions to strengthen the organizational capacity 
of local CSOs in local areas?
Impact IR 3.3
6. How much progress has been achieved in the recovery 
and reconstruction of territory from armed conflicts 
(IAG presence) and illicit economic activities?
Impact Cross-cutting
7. To what extent have CELI interventions contributed 
to participation, governance and institutionalization of 
the territory?
Impact Sub-IR 3.2.1
8. To what extent have CELI interventions contributed 
to regional integration and economic development? Impact
IR 3.2
IR 3.3
These questions serve both impact (1, 4-8) and performance evaluation (2, 3) goals and the 
evaluation reflects this blended purpose with a mixed methods design (see section on 
Methodology and Annex 3 for additional details.) The interdisciplinary team used data from the 
 
21 This question is answered in Annex 6 with data and analysis on GOC investments in PNCRT and CELI municipalities.
31 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
quantitative household survey (in a quasi-experimental pre-post design) and a range of 
qualitative instruments, secondary data search and multi-faceted analyses to answer these 
questions in this report. 
32 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
EVALUATION DESIGN 
How to read this report
An impact evaluation is a powerful tool but also includes important challenges and differences 
from other evaluation research. The Limitations section below outlines these, and Annex 3: 
Methodology contains additional detail. It is important to read the report with these issues in 
mind, and to understand the different types of findings an evaluation like this generates and how 
they should be weighted or interpreted. 
Types of results. There are four types of findings in the report, shown below and in Figure 2.1: 
1. Impacts. These come from the survey and are the result of the difference-in-differences design, 
comparing treated and control groups on key outcomes of interest.22 Impact figures are read as 
the percentage point (pp) difference between baseline and midline that is attributable to the 
intervention. Data tables on impacts show statistical significance with asterisks (*) and the title of 
the table begins with the word Impact. 
2. Trends. These figures also come from the survey and they show how one sample changed over 
time – that is, without the comparison and the calculation of impact. These are shown in graphics 
rather than tables, and the figure title always begins with the word Trends. These are referred to 
in the text with the percent sign (%), and in terms of absolute values. 
3. CELI performance data. These data come from the CELIs and they track performance 
indicators linked directly to the CELI interventions. Figures with such data have a title beginning 
with CELI performance. 
4. Secondary data. These data come from other sources like government figures and indices, and 
they help to describe the context in which CELI interventions operate. Tables with these data 
have a title that begins with the words Secondary data. 
Figure 2.1. How to read and interpret the statistical data in the report
Table title: In the text: How to read:
Impacts pp (percentage point) Relative change over time (compared to control groups) 
attributable to the intervention
Trends % (percentage) Change over time in one subsample
Secondary data Secondary data Show context to support findings and/or to explain challenges
CELI performance CELI performance CELI progress toward performance goals
Whole sample and four sub-samples. From the survey, there is one large sample (19,946 
households) divided into four subsamples. These are respondents from CELI municipalities, their 
 
22 This method computes impact on outcome variables, subtracting the estimator (e.g., income) across time (mid-term – baseline) 
and then between groups. This provides, as a result, the net effect of the program.
33 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
controls, PNCRT municipalities, and their controls. We also discuss contextual conditions, 
referring to the “full sample,” when results are similar across the entire sample. 
Direct and indirect beneficiaries. A final clarification regards the direct and indirect 
beneficiaries of interventions. Due to design challenges (see the Limitations section, below), the 
midline evaluation could not confirm that those identified as direct beneficiaries at baseline did 
indeed receive CELI interventions. As a result this evaluation report looks at all residents of 
veredas in treated municipalities as indirect beneficiaries. There is one exception, and that is in 
the economic development chapter, where we have been able to identify a population of direct 
CELI participants, and track their progress as distinct from the other citizens of those 
municipalities. 
Methodology and design
EVAL and CNC used a mixed methods – qualitative and quantitative – approach to examine a 
set of indicators for each CELI result. The evaluation sought to understand changes in economic 
opportunity, social development and civil society, government and institutionalization, and 
security as a result of the CELI interventions. The design followed the baseline design from 
2012-2013 very closely, with additional qualitative methods to gain insights into perceptions, 
experiences and opinions among treated (CELI and PNCRT) and control groups. The mid-line 
evaluation was carried out in 33 CELI municipalities and 16 municipalities in three regions 
(Catatumbo, Cauca / Valle and Putumayo) in which the GoC intervened through PNCRT 
activities (hereafter referred to as PNCRT municipalities).
It is worth mentioning that the results of this evaluation are also presented at the cluster level. 
Clusters are individual municipalities or small groups of municipalities with similar geographical 
and socio-economic indicators whose analytical results often differ from regional results. 
Clusters were integral to this study because they helped to collect more focused data. 
The evaluation team conducted 19,946 rural household surveys. Survey respondents (who may 
or may not have been the head of household) were nearly evenly divided between men (51%) 
and women (49%). When disaggregated by sex of the head of household, an average of 25% of 
respondents were from female-headed households (FHHs) and 75% from male-headed
households (MHHs).
There are four groups of households (HHs) in this sample: those in CELI municipalities and their 
control municipalities (Control 1), and those in PNCRT municipalities and their control 
municipalities (Control 2). The heads of these HHs, who may or may not have been the person 
who answered the questionnaire, were 50 years old on average. Figure 2.2 shows the sample by 
the sex of the HH head, and the total sample size. 
Figure 2.2. Sample by sex of respondent
Male Female Totals
Montes de Maria 1607 1264 2871
Central 2732 1915 4647
Norte 2127 2155 4282
Sur 1058 1417 2475
Putumayo 940 983 1923
Catatumbo 1061 863 1924
34 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Male Female Totals
Cauca 717 1107 1824
Totals 10242 9704 19946
The ethnicity with which respondents self-identified is shown in Figure 2.3. A plurality of 
respondents across all four samples self-identified as mestizo. 
Figure 2.3. Sample by ethnicity
The sample was asked a set of questions on conditions of poverty, regarding insufficient housing 
materials, public services and overcrowding. Insufficient coverage of basic sanitation and water 
services were the poverty indicators most frequently cited by the four samples, while insufficient 
walls and overcrowding were the least frequently mentioned. Although the magnitude of these 
problems is greater among CELI respondents, the proportion of households with each condition 
is similar in all samples. Figure 2.4 shows how the subsamples compare on these poverty 
proxies. 
Figure 2.4. Sample by poverty proxies
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
CELI
Control 1
PNCRT
Control 2
White Mestizo Indigenous Negro, Afro-Colombian or mulatto Palenquero, Raizal, Gitano, or Rrom
47%
27% 12%
34%
5%
4%
3%
1%
93%
91%
75%
96%
73%
59%
69%
69%
10%
11%
4%
7%
CELI Control 1 PNCRT Control 2
Insufficient floors Insufficient walls Insufficient sanitation Insufficient water service Overcrowding
35 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Insufficient sanitation and water service were the most frequently cited poverty markers among 
the four samples, while insufficient walls and overcrowding were the least common poverty 
markers. Precise sample amounts are available in Annex 10. Though the magnitude of these 
problems is greatest among CELI respondents, the proportion of HHs with each condition is 
similar throughout the samples.
At baseline, the veredas in CELI and PNCRT municipalities were classified as low, medium, 
high or critical security risks. This characterization of the veredas (often called the stoplight 
rating) was carried out by the GOC and, particularly, its military and Consolidation actors in 
these regions. At baseline, there were two municipalities that the evaluation never was able to 
enter, El Tarra in Norte de Santander and Toribio in Cauca, which reduced the number of 
clusters by two. At midline, after the evaluation team conducted a security review at the vereda
level, which showed the following breakdown of the number of veredas surveyed at each levels 
of security threats:
Figure 2.5. Veredas by security threat level
Quantitative methods
The impact evaluation design included a 
quantitative survey of households in 
Consolidation Zones, both with and without 
CELI interventions. Control municipalities were 
selected using an advanced matching protocol, 
described in detail in the annexes to the original 
baseline report. The original design included an 
oversample of 12% for each follow-on survey to 
allow for attrition. Following a panel design,
midline evaluation field teams revisited the same 
households in order to ensure comparability and 
lost approximately 13% of the sample to 
attrition. 
At baseline, the evaluation was designed to 
capture information from both direct and indirect beneficiaries. Indirect beneficiaries were 
selected at random from lists provided by local leaders; direct beneficiaries were selected at 
random from lists provided by the CELIs. The distinction between the direct and indirect 
beneficiaries was not found in midline data collection (see the discussion of direct beneficiaries 
in the Limitations section.) However, the evaluation team was able to capture and identify a sub￾sample of probable direct beneficiaries, about whose results we report in the economic 
development chapter. PNCRT municipalities only had indirect beneficiaries.
Timing 
The timing of the midterm evaluation was first planned for summer, 2014, but was postponed for 
two concerns. The USAID Mission and EVAL worked together to address these two concerns in 
order to make best use of evaluation funding. The first concern was that of seasonality. In impact 
evaluation literature, one of the most troublesome errors is undertaking rural surveys during 
different seasons. Key indicators about income, farming, sales, access to services and other 
themes differ greatly by season. If measurements are taken in two different seasons, impacts 
Low 172
Medium
372
High 530
Critical
212
36 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
(positive or negative) may be incorrectly attributed to program interventions instead of seasonal 
changes. The baseline was carried out over a variety of seasons over nearly two years’ time.23
The midterm was timed to be compatible with the bulk of data captured at baseline, and to set up 
a seasonal timeline for the endline as well. In the same discussions, the USAID COR felt that the 
midline would be more useful nearer to decision-making for follow-on programming, and the 
new calendar supported that goal. Finally, since the endline could never be on time for follow-on 
planning, the endline would actually be most useful after the conclusion of the program, to 
capture whether program results had been sustained. 
Second, had the midterm evaluation been undertaken in 2014, there would have been just one 
year between baseline and midline. USAID and EVAL discussed the need for the CELIs to 
intervene long enough to give the greatest possible program maturity and, therefore, estimations 
of impact. This was concordant with the function and purpose of the impact evaluation design, 
which is less about performance and feedback, with a greater emphasis on returning impact 
estimates.
In the end, midterm data collection began at the end of July 2015, and continued through the end 
of November 2015. It is important to note that local elections were held on October 25, 2015; it 
is certainly possible that the increasingly politicized atmosphere up to that date would have 
affected respondents’ opinions, but as they were affected in treatment and control zones alike, it 
is assumed that the matching design cancels out that influence. 
Instrumentation
The survey instrument includes modules on demographics, justice and security, licit and illicit 
economic opportunities and performance, government legitimacy, social capital and participation 
in grassroots organizations, electoral behavior, social and public services, land rights and 
ownership, victims’ rights, roads, social programs, and opinions about the future and about 
peace. Several indices are created from the data, including indices of social capital, government 
capacity, and government accountability. In general, the survey measures perceptions (as in the 
case of security) which are important but more sensitive to news and recent events. The survey is 
less adept at capturing actual experiences, in part because most respondents will not have had 
direct contact with the changes in institutions or security, for example. In this regard, the 
indicators designed at the time of the baseline survey are weighted toward citizen perceptions. 
The evaluation added several new items for the midline on satisfaction with interventions and a 
short set on victimization from the Latin American Public Opinion Project (LAPOP).24 The 
 
23 From the baseline report: “The survey interviews occurred in two phases, with the first phase providing interviews for 65 
percent of the indirect beneficiary and control households between March 27 and July 5, 2012. The second phase, providing the
remaining 35 percent of the indirect beneficiary and control interviews, plus 100 percent of the direct beneficiary interviews, 
occurred between November 7, 2012, and May 25, 2013, with a break for the Christmas and New Year holidays.” Impact 
Evaluation of USAID’s Colombia Strategic Development Initiative (CSDI) Baseline Report. Volume 1: Summary of Findings.
June 30, 2013.
24 The Latin American Public Opinion Project (http://www.vanderbilt.edu/lapop/), which has been implemented throughout Latin 
America by a team at Vanderbilt University for the past three decades. 
37 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
survey was administered in person in households, workplaces and, when security prevented 
deployment, in heads of municipalities. 
Full details on methods used in this impact evaluation and the calculation of difference-in￾differences is included in Annex 3, including the calculation of differences-in-differences used to 
detect impact. 
Other quantitative measures
Other quantitative measures included the examination of key secondary data sources. These 
came from government sources and the EVIDINCE consortium,25 which collects quarterly 
municipal level data on outcomes of interest, such as police and judicial staff presence, among 
many others. GOC sources also provided updated data on major crimes including homicides, acts 
of terrorism, and massacres. The CELI implementing partners also provided data on their 
achievements to date in their performance management plan (PMP), including such indicators as 
numbers of organizations and government offices treated, sales of productive project outputs, 
and leveraged funds. 
A further source is the Consolidation Index, created in concert with USAID and collected by the 
GOC since 2011, on a set of economic, social and institutional indicators. The Consolidation 
Index shows one considered viewpoint of the progress to date in the Consolidation 
municipalities. This index is built from a set of 37 indicators in the three pillars of the National 
Policy for Territorial Consolidation and Reconstruction (PNCRT for its Spanish acronym), each 
of which is in turn made up of a set of components, as shown in Figure 2.6 below. 
Figure 2.6. Consolidation Index pillars and their components
PNCRT pillar Components
1 Territorial 
institutionalization
Security, justice, property rights, 
and human rights
2 Citizen participation and 
good government
Good government, citizen 
participation, and social capital
3 Regional integration Connectivity, social services, and 
economic growth
 
Each pillar is composed of a set of indicators that are reliably available, primarily from 
government sources – line ministries, defense and police bureaus, departments and 
municipalities. Indicators in Pillar 1, for example, include indicators on security challenges and
incidence of crimes, in particular conflict-related crimes, and indicators on access to justice 
mechanisms, human rights, and property rights. Pillar 2 includes data from municipal 
governments on their provision of spaces for citizen participation, their effective tax collection, 
 
25 EVIDINCE is a research effort associated with the Empirical Studies of Conflict (esoc.princeton.edu), a multi-institutional 
collaboration studying conflict and insurgency at the sub-national level.
38 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
their coverage of citizen oversight committees, the status of local civil society and Juntas de 
Acción Comunal (JACs), and the Procuraduría General de la Nación’s Open Government Index. 
Pillar 3 on regional integration includes indicators on roads, electricity and cellular coverage, 
access to education and health services, hectares under cultivation and the presence of financial 
entities like banks. 
Scores on these indicators are summed to create scores for each pillar, and each pillar makes up 
one-third of the final score on a scale from 0 to 100. Since 2011, data for the Consolidation Index 
have been compiled into scores for each Consolidation municipality, along with a set of 
comparison municipalities to which these are matched. The scores are disaggregated by pillar or 
by component indicators in each pillar, as well as over time, by geography (municipality, 
department and region), and service provision versus sustainability indicators. 
The Consolidation Index reports on the overall Consolidation region (disaggregated by 
municipalities that are PNCRT only, in red below, and those that also include CELI 
interventions, in blue below) in an aggregate score that has increased, on average, between 10 
and 14 points over 100. This positive trend is shown in Figure 2.7: 
Figure 2.7. Secondary data: Consolidation Index overall scores 
Source: Consolidation index data, authors’ calculations
Both sets of municipalities have improved in the overall Index, as well as by pillars (except in 
Figure 2.9, Pillar 2 between 2013 and 2014), as shown in the three figures that follow: 
53,4
56,2
61,9
62,6
53,7
56,7
64,3
67,6
2011
2012
2013
2014
Total
PNCRT CELI
39 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 2.8. Secondary data: Consolidation Index scores for territorial institutionalization
Figure 2.9. Secondary data: Consolidation Index scores for good government, citizen participation 
Figure 2.10. Secondary data: Consolidation Index scores for regional integration 
It is important to consider that the Index, like any measure, is composed of indicators that have 
strengths and weaknesses, as well as a particular point of view that differs from the evaluation 
fieldwork in several ways. First, the indicators in the Index generally come from government 
perspectives, and in many cases those reporting the raw data are municipal, departmental and 
national officials. This differs fundamentally from the evaluation’s focus on citizens’ perceptions 
43,8 44,3
54,9 61,0
41,9 42,4 49,9 46,9
0,0
20,0
40,0
60,0
80,0
100,0
2011 2012 2013 2014
CELI PNCRT
44,0
53,2
62,8 60,5
47,0 53,9
65,7 58,2
0,0
20,0
40,0
60,0
80,0
100,0
2011 2012 2013 2014
CELI PNCRT
73,1 72,7 75,2 81,4
71,4 72,4 70,2
82,8
0,0
20,0
40,0
60,0
80,0
100,0
2011 2012 2013 2014
CELI PNCRT
40 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
of security, services, participation, etc. Neither one is more “correct” than the other, but they also 
are not precisely comparable when measuring these same phenomena.
Data collection and reporting strategies also differ between the Index and the evaluation data. 
The basis for many of the Index’s indicators is “principal population centers” – which are 
determined by those who report the data, including municipal officials or even local 
transportation companies who report on road conditions. By contrast, the evaluation sample is 
79% rural, 18% in population centers other than the municipal capital, and 3% in the municipal 
capital. A very rural sample like this survey is likely to have very different perspectives than 
those in urban capitals or principal population centers captured by the Index. 
Second, the scores on many indicators are weighted to be compared to municipalities that have 
similar socioeconomic levels, and as a result may be given high marks even when their absolute 
scores are among the lowest in the nation. For example, in the indicator on open government 
(using the Colombian government’s own Open Government Index figures), the average 2014 
score among ten municipalities in Antioquia is 56.9, but this is calibrated to 87.4 by the 
comparison with like municipalities. It is important to take into consideration that the scores are 
altered in this way, before being added together to create the Index. 
Qualitative methods
Qualitative methods included in-depth interviews with key informants at national, regional and 
local levels, in CELI and PNCRT sites as well as control municipalities. Twenty-one sites were 
selected across these samples. The sites needed to reflect PNCRT sites without CELI 
interventions, and municipalities with both PNCRT and CELI interventions, and control 
municipalities. The sites selected also needed to include all four CELI regions and to be 
accessible with reasonable travel times and surmountable security challenges.
National-level informants were selected for their relationships with CELI or PNCRT 
interventions to gather an understanding of the interventions’ designs. Evaluations in the field, 
then, were more about how those designs were implemented. At regional and local levels, the 
evaluation team interviewed staff at the personerías (Ombudsperson offices), because of the 
salience of these interventions in the CELI activities, and secretaries of government, to capture a 
range of other CELI interventions. In each of the 21 visits, these two leaders were interviewed, 
using a semi-structured instrument that allowed for the comparison of their responses. 
Also interviewed were heads of civil society organizations and producers’ associations. Focus 
groups covered themes parallel to the survey and the participants were citizens in CELI and 
PNCRT zones, some of whom had received or participated in interventions, and others who had 
not – the qualitative team worked to balance these populations in each site visit in a CELI region, 
with one focus group of CELI participants and one group of what the design termed “indirect 
beneficiaries” – those who live in the region but were not directly involved in an intervention. 
Qualitative data were analyzed in Atlas.ti using thematic coding that mirrored the evaluation 
questions and allowed for emergent codes to cover issues not foreseen in the evaluation design. 
These codes were correlated to understand coincidence and patterns of importance to the 
evaluation’s response to the research questions. 
All instruments are included in Annex 8 to this report. 
41 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Limitations
The evaluation design followed in this study has certain limitations. These are grouped and 
discussed below, and a more detailed discussion can be found in Annex 3, Methodology. 
Factors related to the evaluation design 
1. The CELIs were to be evaluated as part of a much larger and more comprehensive 
effort that failed to materialize. The outcomes of interest are at a high level of 
generality in the results framework, relative to the investment, the interventions, and the 
duration of implementation to date. The evaluation was created to assess a set of 
programs, formerly called the Colombia Strategic Development Initiative (CSDI), with 
broader funding and goals than the current set of interventions. CSDI included a whole￾of-government approach from both the U.S. and Colombian governments – anti￾narcotics, security and justice, regional and local institutions, economic development, 
regional integration and roads, and civil society interventions, across the same geographic 
intervention area. The Initiative as a whole dwindled to comprise only the Consolidation 
and Enhanced Livelihood Initiative, run by USAID, with significant funding but without 
the wider approach initially envisioned. As a result, the impact evaluation design includes 
several indicator sets for which few or no targeted interventions were completed. Not 
surprisingly, when interventions were not carried out, no impacts can be detected. 
For CSDI to have produced detectable effects would have been an enormous feat; for the 
CELIs to do it without the other CSDI interventions, including the interventions promised 
by the GOC, was far too much to expect. As a result, impacts found by the evaluation as 
measured by rigorous statistical techniques are few and mostly at the local level with 
particular direct interventions with producer associations. 
2. Selecting the panel of respondents at baseline was a challenge, resulting in 
discrepancies in the list of “direct beneficiaries” at midline. A further limitation in 
terms of determining and attributing impact involves the panel design of the baseline, 
followed closely at midline. Baseline data collection was carried out over 18 months, 
crossing various growing seasons and the three different starting dates of the CELIs. In 
order to respond to the baseline team’s request for lists of direct beneficiaries, CELI 
Central approached local leaders and called for meetings of potential local participants, 
who signed attendance lists. CELI N/S, which was already in progress, provided lists of 
those who were participating in their interventions at that time. These lists served as the 
sampling frames for the treatment sites, but they created a problem: attendees knew they 
would receive a program, thus violating the basic assumption of an impact evaluation 
baseline that the intervention has not yet begun. Knowing about the future benefit 
changes the population’s perceptions, perspectives, decisions and expectations. To some 
extent, the changes begin to appear from the time of the meeting invitation, and such
changes can no longer be observed when the baseline information is collected, so we do 
not know how this knowledge may have changed their responses, but the relevant 
literature suggests a social desirability bias in which respondents who think they are to 
receive some benefit will respond more positively, in an attempt to be amenable to 
researchers – who they may presume to be involved with the intervention. This makes it 
very difficult to detect small effects on certain variables, since the baseline measurement 
has already incorporated, or lost, such effects. 
42 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
The midline evaluation found that some baseline respondents originally identified as 
direct beneficiaries did not participate (or did not recall or report participating) in CELI 
interventions. There were significant discrepancies between people who were classified 
as direct beneficiaries at baseline and the people who eventually participated in CELI 
programming. The evaluation team therefore recognized three definitions of “direct 
beneficiaries”: (1) direct beneficiaries per the baseline evaluation design (4,426 
households); (2) respondents who answered a survey question about receiving an 
intervention from Colombia Responde, as the CELI program was known in the field 
(2,198 households); and (3) direct beneficiaries from the baseline who are on the CELIs’ 
own participant lists (2,015). 
Random households in the municipalities were then selected as indirect beneficiaries, to 
complement the sample and to represent the wide-ranging set of indicators that the CSDI 
was to have impacted, such as economic integration through roads and markets, security 
and justice, and increasing and improving government presence. 
None of the three “direct beneficiary” definitions provides a sufficiently reliable gauge of 
participation, and as a result, for analysis of the more diffusely applied components of 
CELI activities in governance, infrastructure support and social development, the indirect 
and direct beneficiaries of the baseline were combined into one group of beneficiaries. 
However, for certain activities, especially with respect to participation in producers 
associations and other economically focused interventions, analysis was conducted on a 
separate “direct beneficiaries” subgrouping based on the three definitions just specified. 
When taken together, with no double counting (that is, counting a household only once, 
for example, when a participant fulfills more than one of the definitions such as by 
appearing on the original list and by responding “yes” to the survey question on 
Colombia Responde participation), there are 5,445 unique respondents among the three 
definitions. The evaluation uses this larger group to examine economic impacts on 
participants, which represents a more rigorous standard. Where impacts are found on this 
group, they are generally more positive than the rest of the citizen population. 
This qualitative team’s experience in the field conducting focus groups helps to 
understand why so few of those identified as direct beneficiaries actually reported being 
participants. The first and key questions in the focus groups were regarding participation 
in Colombia Responde activities, and few focus group participants said they had
participated. The reasons for the problems with the Colombia Responde name are not 
precisely known, but the evaluators speculate that the following could have contributed: 
i) at GOC request, USAID hid its participation under the Colombia Responde name so 
that the beneficiaries would attribute the effects of the program to the GOC; ii) 
subcontractors have other names, and beneficiaries may have been confused about the 
identity of the benefactor in the first instance; iii) the results may suffer from recall bias, 
since the baseline was established between 2012 and 2013, and certain benefits ended a 
long time ago; and iv) a combination of these and other factors that are unobservable. 
3. The category of indirect beneficiaries is imprecise, and capturing impacts at that 
diffuse level will take more investment and more time. Several categories of CELI 
and PNCRT interventions – essential though they may be – are likely to be invisible to 
average respondents, and do not reach a minimum treatment threshold, for one of two 
main reasons. First, some public goods affect few residents of a given area, like the 
43 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
strengthening of justice services or work with municipalities or personerías, or some 
social or productive infrastructure projects. Internationally, only an average 6% of a 
given population ever accesses justice services in their lifetime, and only a fraction of 
that figure do so in a given year. Detecting perceptions of change among this “fraction of 
a fraction” is therefore very difficult. Second, as discussed above, the CELI programming 
had important goals and benchmarks across the pillars of social, economic and 
institutional development – such that the significant funding is fragmented across 
geographic sites and pillars. While a member of a treated productive association or civil 
society group knows well that they have benefited from an intervention, the “indirect” 
beneficiary is indeed very distant from the benefits. Their inclusion in the sample dilutes 
the impacts that can be detected by the evaluation design. 
Had the robust set of CSDI and GOC whole-of-government interventions continued as 
planned, perhaps more impacts in these higher-order indicators would have been 
detected. As it stands, however, the “dose” of intervention is a fraction of what was 
planned, in far fewer sectors, and at great distance from the beneficiaries, particularly the 
indirect beneficiaries.
Factors outside the control of the interventions
4. Local contexts and the absence of public entities condition the effectiveness of the 
CELIs. The intervention municipalities are historically economically isolated, with 
stagnant local markets, very limited road connectivity, armed illegal group activity, 
minimal (and sometimes corrupt) government institutions, and deep intergenerational 
poverty. The households in these zones have limited access to land – in terms of either 
size or low productivity – and a deficit of human capital (education, health, nutrition, as 
shown in baseline statistics on education coverage, perceptions of service quality, and 
food security challenges26). The main consequence is low productivity and income from 
land and work, which are rural families’ two main factors of production. This constrains 
households in a logic of poverty perpetuation – the poverty trap – in which poverty is 
reproduced as a consequence of low savings capacity, which prevents changing the initial 
amounts of capital, land and work. The cycles are reproduced in the short-term, limiting
the capacity to change the initial situation, and in the long-term, limiting changes in the 
welfare of the next generation: young people cannot find opportunities and are pressured 
into migrating or joining illegal armed groups. In this context – and others related to 
unemployment and informality – violence, in its different forms, reinforces poverty traps. 
The other face of this poverty trap context is communal: the municipalities lack access 
roads, political representation and market formation. Weak political representation often 
dilutes attention by the central or departmental government, such that the investment in 
roads is insufficient to develop markets in which small farmers can sell their products 
 
26 See, for example, Impact Evaluation of USAID’s Colombia Strategic Development Initiative (CSDI) Baseline Report Volume 
1: Summary of Findings – All PNCRT Zones p. 30 and pp. 42-43; Volume 2 pp. 47-50 and pp. 69-70; plus the regional baseline 
reports. 
44 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
above production and transport costs. As such, poverty is not only individual: it is 
poverty at a municipal level that encompasses the collective. There is an evident absence 
of public goods that are essential for the development of communities and entire regions. 
The capacity of rural interventions to generate significant effects on economic and social 
development is closely linked to land tenancy and property rights. However, the State’s 
policy has had little success on this front, and the CELIs faced the same – or worse – 
institutional roadblocks. 
5. Exogenous contextual factors cannot be excluded. Among exogenous factors,
27 it is 
important to mention the inflation caused by the global economic slowdown and the 
sharp drop in oil prices, resulting in recession in economic activities associated with 
extraction. Colombia is the fourth largest oil producer in Latin America, and crude oil is 
the country’s largest export and source of foreign exchange.28 The drop in oil prices – 
down 11% between 2015 and 2016 – occurred at the moment that beneficiaries’ 
perceptions of the performance of the CELIs and the PNCRT were sought. 
6. The CELIs’ relationship with the Consolidation Unit was not uniformly strong, 
based on political and other factors. The Unidad Administrativa para la Consolidación 
Territorial (UACT) was a new government body as the CELIs began, which benefited 
from some CELI support for opening offices in the territory and staff capacity building. 
Being new, the Unit had to establish itself and had less authority than was likely 
necessary to impel relevant Ministries to undertake work to meet the Unit’s own goals. 
The national level and regional level offices were not always in synch, according to some 
stakeholders, and one CELI COR reported better and more collaborative relationships at 
the regional than at the national level. The UACT was closed and changed to be an office 
within the Department for Social Prosperity (DPS, for its Spanish initials) in December of 
2015. 
The UACT faced great obstacles in successfully carrying out its role in the territory, 
including limited budgets for meeting the needs in a complex geographic area with deep 
social, economic and institutional challenges. The UACT faced the bureaucracy and 
limitations of a new government agency charged with responsibility for “coordinating” 
more powerful ministries, but without the authority and power necessary to guarantee 
that collaboration. It was under these conflicts and complexities that the CELIs had to 
coordinate their tasks with the national-level GOC (though regional and local contacts 
were reportedly much more operational.) Now that the UACT has closed, the CELIs are 
working to articulate their efforts with the changed institutions.
 
27 The role of the control municipalities is to ensure that exogenous factors do not bias results. Under perfect conditions, 
exogenous factors apply to both types of municipalities, so in principle these factors would “cancel out.” However, no quasi￾experimental design is ever perfect, and as a result, the evaluation uses control co-variables and clusters in order to minimize 
possible biases.
28 http://af.reuters.com/article/commoditiesNews/idAFL2N17O0W3.
45 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Factors related to implementation
7. The selection process for veredas in which to work faced challenges that could affect 
the evaluation results. The selection of veredas is a crucial foundation for 
implementation and for evaluation. The selection of municipalities and veredas in which 
the work is ultimately performed has substantial effects on the results of the strategy – 
and, therefore, on the impacts measured. Implementing partners are guided in selection 
by USAID and GOC preferences, and the baseline team for this quasi-experimental 
design followed up by employing a municipality matching system. Such systems, while 
imperfect, allow for the modeling of a robust counterfactual – what would have happened 
in municipalities that did not receive the interventions.
One hypothesis generated during the evaluation was that the selection – guided by the 
“traffic light” process – tended toward the veredas closest to the municipal centers, with 
greater access, more resources, greater associativity, and greater social capital, etc., while 
the problems of illicit crops and armed groups were extended in more remote and 
dispersed geographic areas. In general, it can be asserted that the poorer the population, 
the greater the impacts of an intervention of this type. Thus, focusing on what was easiest 
operationally may have been counterproductive in terms of impact. 
At least at the outset of the programs, the selection of veredas depended on the 
“stoplight” process, which relied on Colombia’s military for impartial classification of 
the danger of each vereda where the CELIs might potentially intervene. Stakeholders in 
USAID and the CELIs eventually sought a way out of this requirement, as they found the 
system highly politicized and, at times, the stoplight decision unjustified. Over time, the 
CELIs’ contracts were amended to allow implementers to amplify their reach to veredas
in which local dynamics suggested potential positive relations – such as a shared 
productive project. In this way the CELIs could serve communities that were divided 
administratively but shared common agricultural goals, and could expand to veredas that 
they themselves had established as safe for project activities. 
8. CELI intervention expanded, so that seven municipalities included in the PNCRT 
sample are now CELI intervention sites. This means that, when the evaluation 
discusses the PNCRT municipalities, seven have been subject to CELI intervention and 
may show progress that cannot be attributed to PNCRT, but rather to CELI Central. This 
is discussed in detail in the PNCRT regional report.
46 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
ECONOMIC DEVELOPMENT
Evaluation question: To what extent have CELI interventions contributed to regional 
integration and economic development?
SUMMARY: Nationwide rural development challenges and infrastructure deficiencies 
continue to affect CELI and PNCRT municipalities deeply, and most respondents in all 
studied areas report a significant decline in their economic conditions since baseline.
Poverty is not only at the household or individual level, but also at the level of 
communities, as the municipal and departmental conditions (market failings and lack of 
roads) severely constrain growth. Access to credit remains tenuous across the regions, 
and land tenancy has made few gains. Income has increased, roughly at the level of 
(urban) inflation, but survey and focus group data suggest worsening food security. The 
bright spot amidst this backdrop are direct CELI beneficiaries, who report improvements 
in economic conditions generally and in a set of key variables particularly. The challenge 
of sustainably producing and selling products remains. 
The goal of this chapter is to understand the role of the CELI projects in promoting rural 
integration and economic development. The information collected was analyzed for the main 
findings in two priority areas: i) the limitations of the municipality and of the households in 
terms of developing their economic activities and generating incomes; and ii) the role played by 
the CELI interventions in supporting and promoting local capacities and eliminating said 
limitations in a joint effort with the national government. To assess the economic situation, the 
evaluation team gathered information from a broad range of CELI, PNCRT and control 
respondents. These included direct and indirect beneficiaries, control groups, and representatives 
of the communities at large, including personeros, secretaries of the alcaldías, members and 
leaders of JACs, and civil society leaders, among others. 
Findings and conclusions
Direct CELI beneficiaries
Respondents who participated in CELI productive projects showed significant improvements in 
their economic situation. When asked about whether their economic situation was good or very 
good, there was a 6 to 10pp positive impact on their responses (Figure 3.1).29 There was also an 
impact of 6pp fewer respondents saying that their productive activities generated very little 
money. 
 
29 Please see the Limitations section in this report and Annex 3 for a full description of direct beneficiaries in the sample.
47 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 3.1. Impacts on perceptions of the economic situation, direct beneficiaries 
Variable Direct CELI 
beneficiaries
Report that the economic situation is good or very good 6 to10pp**30
Their productive activities generate very little money -6pp**
The CELIs report incremental or added value for key products that they support in productive 
projects, as one of their main indicators. In its most recent reports, this incremental value 
exceeded the goals set by the CELIs substantially and would significantly help the supported 
associations and productive projects. (Figure 3.2) 
Figure 3.2. CELI Performance: Increasing sales
CELI USD % of the goal
Central $1,723,467 121%
North-South $16,735,470 358%
Source: Quarterly and annual CELI reports; see Annex 5
N.B. El CELI Montes de María did not report on this indicator and it has been closed
There were no impacts, positive or negative, on direct beneficiaries’ responses about food 
security and poverty. Among direct beneficiaries there was a 9pp positive impact when they 
were asked if they received support for their productive projects. They were significantly more 
likely than other respondents to associate for buying inputs, processing and selling products, 
transportation, and training. (Figure 3.3) 
 
30 A range of impact coefficients is offered here, between the three possible definitions of direct beneficiaries. Please see the 
Limitations section and Annex 3 for a full description of these definitions in the sample.
48 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 3.3. Impacts on reasons they associate with others, direct beneficiaries
Variable Direct CELI 
beneficiaries
To buy agricultural inputs 6pp** approx.31
To process their products 7.5pp** approx.
To sell their products 10pp** approx.
To transport their products to market 9pp** approx.
To received shared training or technical assistance 13pp** approx.
To have a shared representation as a group 7pp* approx.
To present group projects and seek financing 8pp* approx.
To seek credit 5pp* approx.
Regarding credit and financial services, direct beneficiaries were more likely to report having 
received credit, at nearly 11pp greater than at baseline, while the full CELI sample reported 
7.6pp positive impact on having received credit to support their projects. There was also a 
significant 8pp increase in responses indicating they had current debt. There was no impact on 
the amount of credit since baseline. For financial services, there was a positive impact beyond 
that of the wider sample of CELI respondents in terms of having a bank account. (Figure 3.4) 
Figure 3.4. Impacts on credit and financial services, direct beneficiaries versus all CELI
Variable Direct CELI 
beneficiaries
All CELI 
beneficiaries
Have you gotten credit to support your productive projects? 11pp** approx. 7.6pp**
Do you have debt? 8pp** approx. 3.4pp
Do you have a checking account? 3.5pp** approx. 2.3pp**
Do you have a savings account? 8pp** approx. 0.1pp
There were no impacts for direct beneficiaries in having a parcel of land for productive projects, 
but they did have a higher average number of hectares of land in usufruct with title, at more than 
double the hectareage found among the wider CELI population. There was a positive impact on 
their having land in usufruct without title, parallel to that of the other CELI beneficiaries. There 
was no impact, as with the rest of the CELI sample, on being willing to pay predial taxes. They 
were also more likely to say they had not paid predial taxes because costs were prohibitive, or 
because they were currently in a succession process. Fewer say they would not pay predial taxes 
 
31 Figures are approximate for the direct beneficiaries because there are a range of responses based on the definition of direct
beneficiaries used. See the Limitations section and Annex 3 for full details.
49 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
because they already paid someone else (Figure 3.5). For both direct beneficiaries and the wider 
population, the proportion of respondents who say they paid predial tax in the last two years fell 
by 4pp. 
Figure 3.5. Impacts on land issues, direct beneficiaries versus all CELI
Variable Direct CELI 
beneficiaries
All CELI 
beneficiaries
Average hectares of land in usufruct with title 14 to 17 has** 8.9has**
Do you have land in usufruct without title? 2.8pp** approx. 2.2pp**
Did you get your land through restitution? 3.5pp* approx. 0.16pp
Did you get your land through agrarian reform? 1.5pp* approx. 1.26pp
Would you be willing to pay predial taxes? 0.07pp approx. -0.98pp
If you have not titled your land, was it because you don’t have resources to 
do so? 7pp** approx. 5.6pp**
If you have not titled your land, was it because you are currently in a 
succession process? 2.8pp* approx. 2.5pp**
If you would not pay predial taxes, is that because you’ve already paid other 
groups? -2.7pp* approx. -2.7pp*
About a third of the direct beneficiaries reported that Colombia Responde had carried out 
infrastructure projects in their veredas, such as the construction or improvement of roads and 
schools; one-quarter said that Colombia Responde had improved or constructed parks, cultural 
spaces or sports centers. Almost one-fifth (between 17 and 20%) of direct beneficiaries reported 
that the Colombia Responde interventions included improvements or construction of spaces for 
community meetings.
Overall, direct beneficiaries give the CELIs positive grades in their survey responses. First, they 
were asked to rate the degree to which the projects had met their expectations, on a scale of 1 to 
5 (where 1 indicates that expectations for the program were not fulfilled and 5 indicates that they 
were completely fulfilled), and the average grade given by direct beneficiaries was 3.7. A full 
96% of direct beneficiaries affirms that they would recommend participating in a Colombia 
Responde program to family and friends. 
Rural economic opportunities: View from the grassroots 
The remaining sections of this chapter examine the wider populations in CELI, PNCRT and 
control municipalities: that is, the direct and indirect beneficiaries and their controls. Both 
quantitative and qualitative data show that the general socioeconomic situation of households in 
the Consolidation municipalities is difficult and has worsened since baseline. The stories and 
experiences of focus group participants shed light on the mechanisms that stop people, despite 
their efforts, from being able to rise out of poverty. These are structural problems that are 
difficult to change in the short- and medium-term. 
The income derived from the products cultivated by the households or the daily wages they 
receive for working the land is a subsistence income, leaving households with nothing to save 
50 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
37% 43% 38%
15% 18%
49%
77%
31%
0%
20%
40%
60%
80%
100%
The economic situation is
good or very good
Last week the respondent
ran out of food
The conditions in which they
live are of extreme poverty
The family is suffering hunger
CELI baseline CELI midline
and no way to acquire assets or working capital. On the whole, the households depend on the 
income of one person, they live in precarious housing, and they do not have access to land: 
Do you see? We don’t have anything, our house isn’t even finished. If my husband has any 
[money] … he’ll pay rent [for land]; because, when you don’t have any land to grow things on, 
then you have to work like this. Even now, you have to work day labor to be able to earn a living, 
but this is very poorly paid. Do you see what I am saying? From 6 to 4 in the afternoon, and for a 
week, they pay $125.000. That is not enough for the family. (Producer, Meta, Control) 
Across the full sample, in control groups and in CELI and PNCRT municipalities, respondents 
say that their economic situation deteriorated since the baseline. This is true across sites, across 
various survey questions, and by notable percentage changes (trends) from baseline to midline. 
Since control respondents show the same trend, no impact of the interventions was found. 
Perceptions from CELI municipalities are shown in Figure 3.6, which parallel perceptions in 
PNCRT and control municipalities. 
Figure 3.6. Trends in economic situation perceptions, CELI municipalities 
Source: Authors’ calculations. *** At 99% confidence, ** at 95%, * at 90%.32
There was a positive impact of 7.7pp in CELI municipality respondents reporting that their 
economic situation was good or very good (Figure 3.7). Food security variables showed negative 
trends (Figure 3.6) but no impact (Figure 3.7), since treatment and control municipalities had 
similar results. 
 
32 Throughout, the data source is the evaluation’s survey data, unless otherwise noted. Standard significance notations apply: *** 
for 99% confidence, ** for 95%, * for 90%. Impact estimation models are in Annex 3, and data in the text are included in a table in 
Annex 10.
51 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 3.7. Impacts on perceptions of economic and food security variables,33 CELI and PNCRT
Variable CELI PNCRT
Report that the economic situation is good or very good 0.0774** 0.0115
Report going hungry -0.0141 -0.0507
Report that they feared running out of food in the last week for 
lack of money 
0.00338 0.0410
Report running out of food in the last week for lack of money 0.00414 -0.0499
Report living in conditions of extreme poverty -0.0656 0.0766
For CELI municipalities, male heads of household (MHHs) showed a positive impact of 5pp. 
That is, among these households, the economic situation is perceived as deteriorating, but 
somewhat less so than in their controls. The households with female heads (FHHs) saw no 
impact. These pessimistic perceptions seem to be a contextual condition rather than attributable 
to any intervention. 
Survey questions regarding the economic situation also focused on serious problems related to 
unemployment or the lack of opportunities, the profits from productive activities, and incentives 
for illegal activities. No impact was found on the households’ perceptions of the economic 
development of their regions, with one exception – that of profit from productive projects. For 
this variable, the CELI interventions improved the situation by 5.7pp for households in CELI 
municipalities, specifically FHHs, as compared to households in control municipalities. (Figure 
3.8) 
Figure 3.8. Impacts on perceptions of serious economic problems, CELI and PNCRT
Variable CELI PNCRT
Unemployment and lack of work opportunities are serious problems -0.0293 -0.0233
Their productive activities generate very little money -0.0571* -0.0185
One of the few attractive productive options is that of illegal activities 0.0326 -0.0612
There were no impacts in the PNCRT municipalities for any of these variables (Figure 3.8). In 
absolute terms, PNCRT households and controls had the same deterioration for all variables.
Positive changes in perceptions were noted in several issues relating to economic development. 
By sex, the FHHs in CELI municipalities reported a positive impact of 6pp on the perception that 
 
33 These figures estimate project impact compared to control groups and are read as the percentage point (pp) difference between 
baseline and midline that is attributable to the intervention. In the first variable, for instance, a statistically significant and positive
7.7pp impact was found in the CELI municipalities. Statistical significance is denoted by asterisks, per the previous footnote. 
52 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
their productive activities generate little money. For PNCRT, this perception improved even 
more, falling by 8.5pp in FHHs and 0.5pp in MHHs. 
In spite of some negative perceptions about the economic situation, qualitative data collection 
revealed that, when the smallholders were asked about their views on their region and its 
potential, a different picture emerged. This is an indication of the ownership inhabitants feel 
regarding their territories, and, in some cases, a community-wide rejection of illegal activities34
in favor of a culture of legality. 
Ten years ago Costa Rica was all about coca; all over the place, there was even coca in cell 
phones; a terrible war. Thank God, we live in peace now. We have planted citrus fruits, 
chocolate, and there is farming, milk and grasslands. (Producer, Meta, CELI) 
It is magnificent here, especially in the countryside. Everything grows in the countryside, 
whatever you want to do, to plant, farming, everything. People can enjoy the peace, the climate, 
everything is perfect. I would say that life is perfect here. (Producer, Meta, CELI)
Consolidation Index data on regional integration show gains in economic growth (measured by 
growth in cultivated hectares, and number of financial entities), increased connectivity (roads, 
electrical and cell coverage) and positive and stable social service provision (measured in terms 
of school and health system coverage.) The trends and data points are shown in Figure 3.9. 
Figure 3.9. Secondary data: Consolidation Index on regional integration, trends 2011-2014 
These data show a very positive picture for the Consolidation zones, while survey data show a 
worsening situation across the wider population. This has much to do with the indicators in 
question, and their perspective – the Consolidation Index looks at indicators for which it has 
municipal-level access, and these are not from the more rural citizen perspectives of the survey 
data. 
 
34 Please see the sections on illicit activities, security, and their contribution to structural poverty, later in the report. 
78,5
88,8 85,1 87,1
47,7
61,8
0,0
20,0
40,0
60,0
80,0
100,0
2011 2012 2013 2014
Connectivity
Social services
Economic growth
53 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
The complex question of household income
Estimating household income among rural small-scale agricultural producers has long been 
recognized as difficult and imprecise.
35 Nevertheless, CELI projects are required to report on 
income estimates for their PMP, based on data collected in the baseline and in the current 
midline research. Estimated monthly expenditures are used as a proxy for income, because of the 
challenges of collecting income data. Reported household expenditures have increased in 
absolute terms since the baseline. However, the increases are similar in the control group 
municipalities, so it is not possible to attribute these increases directly to CELI and PNCRT 
interventions. 
The increases are relatively parallel to the inflation rate36 in three of four CELI regions, so while 
the increases are positive (Figure 3.10), they are not likely to have significantly changed 
respondents’ living conditions relative to household expenses. This is supported by their greater 
reports of food insecurity in response to other questions (Figure 3.6, above). The exception is in 
Montes de María, where a larger leap in reported monthly expenses is recorded (66%). Still, the 
increase is parallel in control zones, and cannot therefore be attributed to the CELI interventions. 
The expenditure by region and the percentage change are reported in Annex 5. The CELIs met 
their targets for this indicator. Disaggregating by sex shows no difference between FHHs and 
MHHs. 
 
35 The survey asks about the monthly expenditure instead of the income, which has been shown to be unreliable in household 
surveys. This phenomenon has been documented for decades in developed countries and across much of Latin America. Rural 
surveys tend to present the worst performance in correctly capturing family incomes, for a number of reasons: income is received 
from various sources; households receive payment in kind; they are paid irregularly and their income changes according to the
season; if there are any illegal payments, they are not likely to be reliably reported, etc. Estimations show that the reported 
expenditure is in some cases greater than the income reported in the same survey (Rogers and Gray 1994; Jencks 1997; Sabelhaus 
and Groen 2000); de Meyer, Bruce D., and James X. Sullivan. (2003). Measuring the wellbeing of the poor using income and 
consumption. National Bureau of Economic Research Working Paper Series; Working Paper 9760 accessed February 2, 2014 at 
http://www.nber.org/papers/w9760, p. 18. Expenditure has been considered more precise because there is less reason to under￾report. However, in order to obtain good estimates, very detailed instruments are required, adding 20 or 30 minutes to an 
interview. It was decided not to take that type of measurement, but rather to continue to use the measure established in the 
baseline, to maintain comparability and to minimize the burden of the instrument. See, for example, Pettersson, Hans. (2005).
Chapter XXIV: Survey design and sample design in household budget surveys, in Household Sample Surveys in Developing and 
Transition Countries Department of Economic and Social Affairs Statistics Division, Studies in Methods Series F No. 96 al 
http://unstats.un.org/unsd/pubs/gesgrid.asp?id=349. 36 The inflation rate is based on urban data at the national level, and so is not precise enough for close comparisons. But, in 
general terms, the increased reported expenditures across CELI regions do not exceed that rate of year-on-year inflation, or do 
not do so by much, with the exception of the Montes de María region (as this section goes on to explain).
54 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 3.10. Trends in reported monthly household expenditure, CELI and PNCRT
Regional expenditure changes over time do look positive, though they are not statistically 
significant relative to the control municipalities. (Figure 3.11) 
Figure 3.11. CELI Performance: Estimated monthly household expenditure, total and regional 
Ind. Description Region Baseline Midline % Change Source
DO3-
005
Avg monthly 
household 
income of 
USAID's 
beneficiaries37
Total 393,93638 504,106 28% Household survey
• Baseline 2013, 
Econometría S.A. 
• Midline 2015, 
Proyecto EVAL
Montes de María 402,032 667.645 66%
Central 451,601 498.240 10%
Norte 334,978 374.798 12%
Sur 426,725 510.188 20%
Conclusion 1: The perception of the economic situation in CELI and PNCRT municipalities has 
worsened, but the trend is the same in control municipalities. This trend is certainly related to a 
series of factors out of the reach of any intervention and that constitute the poverty trap. Despite 
the difficulties throughout the Consolidation zones, it is not uncommon to find communities’ 
positive values with respect to their regions. This is an indication of the ownership inhabitants 
feel with respect to their territories and productive projects. In some cases, they also demonstrate 
community level rejection of illicit activities in favor of a culture of legality. However, in sum, 
the results indicate that the economic situation in CELI and PNCRT regions parallels that across 
the country, with the hardships and needs of the inhabitants of remote rural zones who survive 
from day to day and face difficulties even in terms of basic food security. 
 
37 Figures are for average or mean household expenditures, as the evaluation team judged expenditures in the rural households 
surveyed to be a more reliable measure than reported income. 
38 Figures used for the baseline differ slightly, as we are now combining indirect and direct beneficiaries (BI and BD, 
respectively, for their Spanish abbreviation used at baseline). Percent change is calculated on the combined BI + BD figures.
55 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Conclusion 2: Direct CELI beneficiaries are an important exception to this negative perception. 
The contrast with the rest of the samples is noteworthy and is statistically equivalent between 
households headed by men and women. Also among CELI households, there are positive trends 
in the perception of the problems they face, such as the lack of opportunities and the low income 
provided by their productive activities. The female-headed households form an important part of 
this optimism.
Assistance for productive projects 
CELI interventions aim to support sustainable and licit income generation. The CELIs made 
great investments to support smallholder associations,
39 to improve the land, to provide technical 
assistance and inputs for production, and to support commercialization. In Antioquia, for 
example, the greatest investments were made to support rubber, cocoa and coffee production; in 
Córdoba, to support cocoa, plantain, rubber and papaya; and in Tumaco mainly for cacao, 
coconut and shrimp. CELI Central reports beginning the program with these types of primary 
inputs, and then focusing more on facilitating government collaboration to involving the state in 
these projects. They focused on quality, better prices for products, organizational development 
and particular project support, such as cold chain development in Meta. The CELI worked in 
concert with Regional Autonomous Corporations (Corporaciones Regionales Autónomas, or 
CAR), mayors’ and governors’ offices, and the productive project partners. Together they 
worked to develop shared plans of action to establish necessary infrastructure and processes to 
help productive projects succeed. 
The CELIs provided technical assistance (TA) in efforts to improve the productivity and quality 
of specific crops. In Tolima, for instance, CELI Central has worked with coffee producer 
associations, “firstly, in offering technical assistance in knowledge transfer and capacity building 
for producers attempting to fulfill international quality standards; secondly, in the donation of 
supplies and materials such as pedagogical materials for training; and thirdly, through a process 
of support for management, negotiation, and training in the services offered by the 
association.”40 In La Macarena, the project brought together members of the producers 
association in La Cristalina de Lozada for an “Agricultural Best Practices” course, during which 
they covered the issues of insemination, vaccination, rotation, pasture capacity, hygiene, quality 
in milk management, and commercialization.”41
All three CELI projects have emphasized organizational training and capacity building for 
producer associations. For instance, towards the end of the MdM project, a CELI partner 
operator helped strengthen ten producer associations through training in accounting, 
organizational, tax and business topics, and office management.42 In CELI N/S, the team 
reported throughout FY 2015 that it continued to improve the institutional capacity of producer 
 
39 A section on producers associations can be found in the chapter on social development. 40 CELI (Central) Annual Report FY2015, p. 52. 41 Ibid., p. 46. 42 MdM Quarterly Report FY2015 Q1, p. 20.
56 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
associations to provide effective services to their members and their communities. In Q3 2015, 
for example, they described supporting eight associations producing cocoa and three that had 
sold shrimp, rice, and honey.43 In Tolima, CELI Central was joined by the local government and 
the private sector to provide training to 30 producer associations working with cocoa and coffee 
to strengthen both their technical expertise and their organizational capacity.44
On the whole, each of the CELI intervention components was directed at counteracting the 
deficiencies or weaknesses that the producers faced in their regions: TA, for example, aimed to 
counteract low productivity and provide products with greater added value. Similarly, productive 
projects were scaffolded with support to their associations and with agreements with local 
governments to facilitate local market conditions. Producer associations were also supported in 
their negotiation power with the buyers or intermediaries. 
The people and communities interviewed in the CELI municipalities positively recognize the 
impact of the support received in agricultural inputs, credit, courses and TA. Their range of 
products has been extended, providing a real alternative to illicit activities. Household economies
have been transformed at the local level, and their parcels are more valuable. 
There is qualitative evidence of the positive changes in the regional economies, through the 
productive projects supported by CELI programs and, it appears, other interventions:45
You can strengthen your stable, stockyard, your cattle. They don’t give us cash but it is a great 
help. The cocoa growers, there are 32 of them, they have been given… fertilizer and cocoa 
seeds… and lime… fungicide for the mixture… and a fumigating machine… (Producer, Meta, 
CELI)
Initially we grew chili peppers, eggplant and rice… thanks to the different programs… we are 
selling our own brand of rice now. (Producer, Antioquia, CELI)
In PNCRT municipalities, 11pp more households (almost all male-headed) reported that they 
received support for their productive projects, when compared to the control municipalities. In 
CELI municipalities, the sample did not report impacts in this period. Globally – including 
treatment and control sites – respondents report a decrease in support for productive projects. 
Families who do receive help – particularly in plant material, courses, or training – more 
frequently report a practical component to the support, which coincides with better practices for 
adult learning.
In focus groups, respondents told stories about CELI learning and support for productive 
projects. The following text describes the process in one CELI N/S case: 
I think that the training has helped us a lot; I think that they help society change its way of 
thinking. Before, we didn’t know how to establish a crop; with these projects we have learned to 
establish rubber and cocoa as crops. So where we didn’t know how to graft a plant for it to be 
 
43 CELI N/S Quarterly Report FY2015 Q3, p. 5. 44 CELI Central Annual Report FY2015, p. 11. 45 USAID readers of this report should note that fertilizer and seeds were not part of the CELI interventions. There were likely 
other programs also intervening in the same place, and that beneficiaries do not always distinguish between them.
57 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
more productive, now we have the possibility to have the technical methods, the technical 
guidelines for us in the fields… now we can improve our parcels of land by our own means, with 
citrus plants, cocoa, with all these products. So all their training workshops have given us the 
knowhow that has served to enrich our social and cultural level. We have enjoyed this part, it has 
been useful, and I think it is the most important thing… the most important thing is that our 
vereda has a different perspective of what the countryside can be. (Producer association leader, 
Antioquia, CELI) 
Female-headed households reported several positive impacts. There was a 10pp positive impact 
on FHH responses that they had received credit for their productive projects in CELI 
municipalities. Similarly, FHHs in the CELI and PNCRT municipalities reported important and 
positive impacts in associations for processing, sales and transport of their products (for more 
details, see the chapter on social development and producer associations). On the other hand, 
FHHs reported receiving fewer courses or trainings in CELI municipalities (18pp less), and 30pp 
less in PNCRT municipalities. 
With respect to CELI productive projects’ support for long-term crops, in focus groups some 
producers talked about their subsistence problems while they waited for such crops to produce: 
I don’t think it has been great [the intervention] because what Colombia Responde has given us 
is not producing yet; they are long-term [crops]. So we still have to work and save from what we 
have to invest in this to sustain it. (Producer, Caquetá, CELI)
This crop [rubber] should be done differently; for example, it should be grown with another crop 
that produces sooner than rubber so that we can subsist from the other crop, we could live off of 
the other crop. It is very [difficult] to sustain rubber… [T]hey should have planted it with another 
crop like they did for the cocoa growers, they cultivated cocoa at the same time as plantain, 
timber products, they planted three products in one. (Producer, Caquetá, CELI) 
Conclusion 3: The productive projects in CELI programs showed evidence of economic changes 
in the households. The beneficiaries appreciate the TA that accompanies the CELI programs and 
other state institutions (e.g., Servicio Nacional de Aprendizaje, National Learning Service, or
SENA), as well as the support for assets and capital that they will be able to enjoy in the long￾term.
Conclusion 4: CELI support goes directly to the strengthening of value that will mature in the 
medium- or long-term. Households with these crops, even before maturity, now have greater 
financial worth because of the production potential they have developed. The goal, along with 
improved institutional support, was to create long-term value that helps producers provide for 
themselves through their productive projects. The purpose of these instruments is to strengthen 
savings and investment, as they oblige the beneficiaries to work and look after their crops with a 
long-term perspective, with all the costs and sacrifices that this implies. However, respondents 
did express concerns that they are currently experiencing short- and medium-term food 
insecurity. 
Markets and trade
The main objective of rural development is the development of markets in a broad sense; that is, 
spaces where the transaction of products continues to expand. Small and underdeveloped 
markets have high transaction costs and communities become trapped in a vicious cycle in which 
small markets do not encourage greater production and, simultaneously, the increased production 
that exceeds the size of their markets collides with rigid local markets and with the lack of 
58 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
preparedness for exports to other regions. Some focus group participants – in both treated and 
control municipalities – explain the serious problems smallholders face in the absence of trading 
mechanisms:
If we do not transcend this part of the transformation, of giving added value, of trade, we 
continue being the same. This is so not only with rubber but any product. (Producer, Antioquia, 
CELI)
They could give us all the crops you want, but if we don’t have anywhere to sell them, what can 
we do? There is a big gap as far as this goes. (Producer, Antioquia, CELI) 
We plant our crops, but there is no one to sell them… the big problem we have here is one of 
trade. (Producer, Putumayo, Control) 
In the survey, a series of items enquired about the problems experienced in production and 
marketing products. There was a positive impact on one of these questions in CELI and PNCRT 
municipalities (Figure 3.12); that is, negative perceptions diminished over time: 
Figure 3.12. Impacts on perceptions of problems in production and sales, CELI and PNCRT
Variable CELI PNCRT
Abuse by intermediaries -0.0395 -0.15
Very low prices for the products -0.0485* -0.0911*
Lack of water -0.0736 0.0952
The impacts listed in the table represent significant reductions in the perception of the problem 
of low prices for their agricultural goods among the respondents, relative to control sites. Though 
impacts were only found for these three questions, the trend over time was a reduction in CELI, 
PNCRT and control municipalities across a range of such problems, though these did not always 
show significant impact. (Figure 3.13) 
Figure 3.13. Trends in perceptions of problems in production and sales, CELI and PNCRT 
59 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
In contrast, respondents in PNCRT sites had a greater perception of the problems of high input 
costs, IAG interference, soil quality, and the availability of water.
For FHHs in particular, there was an important reduction in the perception that market prices are 
low for their products, in both CELI (9pp) and PNCRT (16pp) municipalities. This was the only 
significant difference between FHHs and MHHs, but it is another piece of evidence showing that 
the perceptions and expectations are not always parallel between male and female household 
heads. 
CELI also conducted notable work with the private sector to improve value chains. CELI Central 
reported that rubber in Caquetá and coffee in Sur del Tolima and Norte del Cauca provide 
important lessons in establishing win-win arrangements between producers and the market. In 
Caquetá, CELI forged an alliance between 1,100 rubber producers from 16 municipal 
committees and an international trader in order to establish procedures for quality certification 
and export. A coffee alliance linked nine producer associations with services to market their 
product, conduct tasting training, strengthen organizational capacity in high-standard purchasing, 
and improve access to niche export markets. These activities with the private sector extend to 
other value chains, regions and new markets, helping producers become partners in the process 
with added value, services and better, more stable incomes.46
The markets in the regions continue to be local and regional in most cases. Markets that export 
goods to other regions are small in terms of their sales, with few exceptions. The survey indicates 
that there were no great impacts in the places where the interviewees sold their products. In the 
CELI municipalities, people reported selling more from their farms, in the closest populated 
center, to cooperatives or trade associations, in the municipal center, in near-by veredas, to the 
general public, or in fairs and the market square. They reported selling less outside the 
municipality, though CELI respondents were 7.7pp more likely to have sold to intermediaries or 
wholesalers than at baseline, compared to control municipalities. For the PNCRT municipalities, 
it was parallel and just as mixed. These trends were the same or stronger in the control 
municipalities, so the statistical impact for most of the cases was null. 
In sum, the data indicate that local markets are the most important for smallholders. The 
proportion of sales in the markets located in municipal centers, nearest population centers, and 
neighboring veredas, combined, was from 83 to 90% in the CELI municipalities, and from 73 to 
89% in PNCRT municipalities. Sales to wholesalers and intermediaries declined, but sales to 
cooperatives do not compensate for this reduction (which was expected with certain types of 
associational activity). 
Conclusion 5: TA and marketing support help counteract problems related to low profitability. 
However, trading such products requires markets with enough buyers and sellers and the 
establishment of prices that are higher than production costs. The lack of markets in the 
Consolidation zones seriously hampers the tasks of the CELI and of the GOC. Similarly, the 
productive projects counteract the incentives to plant illicit crops, but with the deficient and 
 
46 Data from conversation and e-mail with CELI Central COR, 30 June and 27 July, 2016.
60 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
partially depressed markets, pressure from armed groups and the lack of access roads (see the 
section on roads later in this chapter), it is difficult for the policies and programs to have the 
desired effects. It could be concluded that in the presence of structural problems – which are the 
GOC’s direct responsibility – it is difficult for CELI and PNCRT programs to produce effects 
that are sustainable over time. 
Conclusion 6: The markets in consolidation territories continue to be local and regional markets, 
and export markets are small in terms of the proportion of sales. As such spaces are limited, the 
producers tend to be affected by the high costs of trading that go along with the deficiencies
typical of small markets. It is a circular relationship: small markets do not encourage greater 
production and, in contrast, greater production that exceeds the incentives of these markets clash 
with the rigidity of the local market and the lack of facility to export to other regions. 
Credit, savings and financial services
Access to financial services and credit to strengthen the productive and licit value chains has 
been an intermediate goal of the three CELIs. They have created savings and credit groups, 
developed revolving funds within producer associations, and offered TA to associations to 
expand their skills in the use of financial information and tools. For example, CELI Central
reports activities with funding that reached USD$608,000 to support 14 producer associations in 
Tolima, Caquetá, Meta, and Cauca whose goal is to improve financial services, from the 
improved management of revolving funds to the use of financial tools to improve accountability 
and information management. CELI Central reached agreements with Colombian financial 
institutions (Bancolombia, Banco Agrario, Banca de las Oportunidades, and Colpensiones) to 
expand mobile banking, create more than 100 savings schemes, and support and train 12 
associations to access FINGARO loans for agriculture and farming.47
CELI MdM worked with the microfinance institution CREZCAMOS to deliver 7,352 
microcredit loans, 11,000 micro-insurance policies, and almost USD$9 million to the local 
economy to fund productive activities. Offices were set up in each of the project municipalities, 
and 116 local savings and loans groups (grupos locales de ahorros y crédito, or GLACs), to
provide credit to their members.48
CELI N/S reports that it has spent over half of its total investment or USD$32 million49 on 
economic development activities, which include rural financial services, strengthening producer 
associations and other local associations, and value chain support.
50 Throughout the project life 
cycle, more than 20,000 men and women have been mobilized in the 16 target municipalities to 
participate in GLACs sponsored by the Banca de las Oportunidades in an effort to reduce 
poverty. A 2013 case study on the growth of these schemes in Bajo Cauca showed how the 
veredas used the shared resources to address community needs such as buildings, pedestrian 
 
47 CELI-Central Annual Report, September 2014 – September 2015, p. 12. 48 Colombia Responde Quarterly Report July-September 2014, p. 26. 49 CELI N/S FY2016 Quarterly Report (Q1) p. 5. 50 CELI N/S Quarterly Report (Q2, FY2015), p. 21.
61 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
bridges or ventures such as drugstores or bakeries.
51 In 2015, CELI N/S expanded the GLAC 
program to help its members use the savings to generate incomes and create other small ventures 
such as micro-franchises for small goods.52
The CELIs have also worked to improve access to credit and bank use among their beneficiaries. 
In terms of bank use, they have helped approximately 52,000 people to gain access to financial 
products. Figure 3.14 shows their performance compared with the goals for this indicator: 
Figure 3.14. CELI Performance: New use of financial services
CELI Number % of the Goal
Montes de María 7.201 142%
Central 7.927 104%
North-South 36.689 141%
Source: Quarterly and annual CELI reports; see Annex 5
They have also worked with productive projects in accessing credit and by supporting their 
associative efforts. Many associations have a revolving fund that offers smallholders the 
possibility to obtain loans to deal with disasters (the conditions and rules of the loans are 
negotiated within each association), serving as a strategy for microcredits and micro-insurance: 
Today, we have a revolving fund of around 60 million pesos; because we decided to use a 
rotating system for all the inputs that arrive [and] harvests that we sell. We take out the inputs 
and we are very clear about the fact that there is a small profit margin, allowing our resources to 
continue to grow. So it has helped consolidation, especially organizational. (Producer 
association leader, Antioquia, CELI) 
The possibility to access some type of credit is an important benefit for the households, given 
that their low savings and investment capacity and their lack of assets53 stops rural households 
from being able to access services in the traditional financial sector. 
Banco Agrario gives loans. I am going to give you an example in case you don’t know; the Banco 
Agrario loans you money, it loans money to the man with a plot of land; but it can’t lend me 
money. Why? Because I don’t have any land. (Producer, Antioquia, Control) 
Formal banks are starting to appear in some but not all territories, and there are areas they have 
not yet reached. There were reports in Meta of people having to walk for four hours to reach a 
municipal center where there is a bank. However, informal access to credit is also present outside 
the formal banking system. In terms of marketing, credits can be linked to trading contracts or 
product delivery. The financial cost is hidden in the commercial transaction. This type of credit 
is based on trust between the parties and, as such, it is very common in rural areas. This type of 
credit can also be more formal, for example, through the middleman that buys the products: 
 
51 Transforming Lives: Rapidly Growing Community Banking Initiative Is About Much More Than Money 52 CELI N/S Quarterly Report (Q2, FY2015), p. 27. 53 Explained by the previously described poverty rate.
62 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
In my vereda, the man has a car and he travels around in it every day, he takes the produce and 
he lends you 10,000, 20,000, whatever you need [to pay] per harvest, in one installment. 
(Producer, Montes de María, CELI) 
Given the lack of access to credit and family income levels, the population is obliged to resort to 
informal or illegal loans, also known as gota a gota. Debtors have to pay usurious interest rates 
which expose rural households to risk. 
Paying [monthly] interest rates of 10% and 20%... apart from the fact that this is illegal money, 
because most are illegal resources and they lend with usurious interest rates. They also attack the 
person, or they threaten him, if they don’t pay the interest. (Functionary, Antioquia, CELI) 
CELI interventions did not directly provide credit, though some participating associations 
created rotating funds to support their members. HHs in CELI municipalities infrequently cite 
credit as part of their productive projects, and that frequency fell since baseline: in CELI 
municipalities, the proportion of households reporting that the help they received was in the form 
of credit fell from 23% to 18%. This decline was much less, however, than in control 
municipalities, for an impact of 10pp for the CELI population; in particular, this difference 
seems to have been felt most in female-headed HHs (10pp). 
In terms of other financial services, the survey results show a reduction in the use of savings 
accounts of around 10% across the full sample. Checking accounts, which comprise only a small 
fraction of banking services, did see a small increase. When disaggregating impacts in CELI 
municipalities by sex, there is a reduction of 8pp in FHHs, while in PNCRT municipalities, the 
reduction for this same group is of 12pp. 
Across the full sample, payments for public services through banks or non-bank branches have 
increased. Both in CELI municipalities and in PNCRT, requests for credits in banks, 
cooperatives, and NGOs have increased; however, FHHs in CELI municipalities report a 
significant decrease of 4pp. There was also a decrease in the frequency of credit across the full 
sample and a drop in the average amount of household debt (almost 6 million COP in CELI 
municipalities, and almost 10 million COP in PNCRT municipalities). 
As mentioned earlier, the revolving funds have proven to be a valuable instrument for small 
groups of neighbors and acquaintances. The implementation costs are low and there are 
important economic benefits to the group. As explained by a beneficiary, a common type of 
revolving fund is the farming fund: 
Our fund is of 168 cows… We give cows to people who can feed and sustain them. For example, 
if someone says they want no more than five animals, then that is what they will get. If they give 
him 5 newly calved cows, in four years, he delivers 5 new, newly calved cows and the 5 that we 
gave him he keeps; and the calves these cows have in these five years... and the heifers are 
branded with the name of the association… they are the ones that will grow and be given back, 
while the bulls are settled for, 50-50. So if in five years you are going to deliver the cattle… and 
there are at least 15, right? Because the original cows’ heifers have given birth and there are 15 
animals; so we take out the 5 that belong to the association, the five that were given to the man, 
and is there are any left, they split them 50-50. (Producer, Meta, CELI)
However, the revolving funds are not sustainable unless they are well regulated and internally 
supervised. Given that the seed capital almost always comes from support programs (a subsidy), 
the beneficiaries sometimes claim that they don’t have to pay such support back, and thus 
63 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
endanger its sustainability. It is for this reason that, as already mentioned, the instrument is very 
well suited to small groups54 in which members trust each other. 
Conclusion 7: According to the household survey, credit fell in both CELI and PNCRT 
municipalities. However, the information gathered in focus groups indicates that financial 
support has fulfilled a vital function for the development of savings and credit. Revolving 
savings and credit funds have served as important instruments to support capitalization in rural 
areas. 
Conclusion 8: Informal credits contained in trading or illegal gota a gota loans will always exist 
unless financial services progress in rural areas. Theoretically, illegal credit problems can be 
resolved judicially, but the informal modalities arise when the broader economy fails to provide 
viable alternatives. Community credit or loans which are a product of programs such as CELI
provide an alternative to informality. 
Conclusion 9: Bank formalization has not increased; in fact, it is moving backwards across the 
full sample. However, households are increasing their use of credit, through cooperatives or 
NGOs. The negative impacts in FHHs in terms of accessing credit is a matter of concern. 
Transportation infrastructure
The Consolidation zones lack access roads that connect rural residents to the country’s most 
dynamic and populated markets.55 Despite high marks for connectivity in the Consolidation 
Index (see Figure 3.9, above), the Ministry of the Interior data convey a different story about 
coverage and quality. The baseline survey data are telling as well: only 16% of the households 
belonging to the CELI municipalities access their houses from a paved road (in the control 
municipalities, this value was 32%); 56% from dirt roads, 20% from footpaths, and 8% from 
rivers. The lack of reliable transportation infrastructure to reach the veredas and municipalities
means that local producers cannot access traditional markets with a stable supply and demand 
that can guarantee prices that are higher than their costs (including that of transporting their 
products). 
Not only does this result in higher production costs for private and public goods, but it also 
facilitates and conceals illegal activities. When the State arrives to implement roads in the 
regions, this is well received by the inhabitants because it favors the competitiveness of 
producers’ activities and it improves quality of life for rural inhabitants. 
There are lots of veredas that don’t even have horse trails; you have to go on the river or along 
footpaths. (Producer, Putumayo, PNCRT)
In my case, I have to walk for two hours just to get to the nearest road, and unless it is on the back 
of an animal, there is no way to move anything. (Producer, Tolima, CELI)
 
54 See chapter on social development, regarding producer associations and communities. 55 Data from the Ministry of Transport. In departments in which CELI works, coverage is minimal (8-16% of the area), there is 
limited paved area (<7%), and poor quality (1-13% in “good” state according to the Ministry’s classification).
64 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
If we have good roads, the products [can be sold]. (Producer, Caquetá, CELI) 
All the counterparts have come to establish crops, not for infrastructure… now that we have our 
products, we need roads to get them out. (Producer association leader, Antioquia, CELI)
The quality of road connectivity is directly related to incomes; the higher the transport and 
trading costs, the lower the producers’ earnings: 
With 60 liters of latex a day, he earns $75.000 and he is paying his helper $25.000. He is left with 
$50 [thousand], but then he has to pay to manage all this: he has to have it transported on an 
animal for 4 kilometers, 5 kilometers to get it to the road edge. (Producer association leader, 
Antioquia, CELI) 
To build or repair roads was not one of CELI’s central axes, but they have acted in this sector 
and such activities have increased in recent years, in a comprehensive approach to resolve 
bottlenecks to commerce. For example, CELI Central designed an activity that sought 
sustainability within a framework that linked civil society with public entities. Support 
committees were set up to follow up on the maintenance of the improved roads. CELI helped to 
legally formalize the JACs so that they could publish invitations to tender and contract the work. 
The inhabitants of the vereda often contributed by providing labor.56 
At last report, CELI Central spent US$1.5 million57 and leveraged three times this amount from 
local governments. The goal was to rehabilitate 113 kilometers of road, but rather than focus on 
kilometers repaired, projects were chosen in which targeted investments would make a big 
difference. For example, in an area of Meta, residents used a market road that had four areas that 
became impassable in certain seasons. The project did not replace the entire road, but rather, with 
help from the residents, found sustainable fixes for the four critical points that were identified. 
Now the road remains passable year-round, at a much lower cost, and the residents themselves 
have the tools to maintain it. Although this way of working on specific repairs does not please 
the construction contractors – who tend to prefer large projects – it is an efficient manner of
using the funds for a common goal. 
Across the full sample, survey respondents recognize a slight improvement in the roads inside 
their municipalities and a relative standstill in the situation relating to inter-municipal roads. 
Some 34% of PNCRT households consider that the condition of the inter-municipal roads is very 
good, and 23% considers that the inter-vereda roads are good. The situation has improved more 
in control municipalities than CELI municipalities. The survey also reveals perceptions that 
transport costs have gone down across the full sample. Even so, more than half of survey 
respondents (58% in CELI, 56% in PNCRT, and more than 52% in the controls) indicate that 
transport costs are a serious problem for production and trade. 
Where we grow things on the farm, it takes two hours to get the fruit out on an animal, sometimes 
the animal can hardly get through because it is in bad condition… in the high part. (Producer, 
Meta, Control) 
 
56 Source: In-depth interviews with CELI teams and their managers at USAID. 57 Ibid.
65 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
At the same time, existing roads are not often well-maintained. Some evidence surfaced in 
qualitative interviews about corrupt activities with regard to road maintenance:
Two years ago a compacting roller went in in the morning and came out in the afternoon. And 
what did they say? The road has been compacted. But it hadn’t been. It was a waste of money, the 
roller went in at 6am and came out at midday and the only thing that mattered was to send a 
photo to the local government. (Producer, Antioquia, CELI)
The missed opportunities are clearly expressed by this agricultural worker in Caquetá: 
In this area, very little is grown to sell… in the area we live in, everyone grows crops that they 
are going to use… Because we lack roads, we cannot sell our product, milk, which is what we sell 
the most, at a good price. (Producer, Caquetá, CELI)
Conclusion 10: The data show very negative perceptions and experiences in terms of the quality 
of the roads network. The deficiencies of local markets are tightly related to the state of the 
roads. The productive projects can attempt to break poverty traps, but the lack of roads reinforces 
the traps. It is obvious that the task of CELI is complementary to government actions, but it is 
not a substitute: the infrastructural problems must be solved by the State, and what CELIs do 
without this urgent GOC work may end up being a short-term palliative.
Conclusion 11: When there are roads and infrastructure, the population recognizes the direct 
benefit of CELI interventions. Local level users, associations and JACs have taken ownership of 
roads projects. 
Land
An important element in the CELI vision is support to land formalization. This includes helping 
the rural residents ensure formal titling, supporting displaced families and communities to return 
to lands they had to abandon as a result of the armed conflict, and support governmental 
institutions who, ultimately, grant land tenure rights. This work began prior to the Mission’s 
Land and Rural Development Program activities. When that program came on line in March, 
2014, the CELI work had built a foundation. For example, CELI Central began work in 2013 to 
formalize lots for public use, such as schools and health clinics. LRDP picked up on that to 
develop similar and complementary activities in some regions of CELI Central coverage. For 
example, in Tolima in 2013, CELI supported the agriculture ministry’s administrative work 
which LRDP has continued into a juridical phase.58
The main obstacle has been weak state institutions in this sector without clear and consistent 
processes. El Instituto Colombiano de Desarrollo Rural (INCODER) proved unable to fulfill the 
mandates of several attempts at land reform and was relatively impotent in the face of political 
processes resulting from extreme land concentration. The challenges in restoring land tenure 
rights or compensating the victims are great. The relatively positive data in the Consolidation 
Index show an increase in scores (over 100) for property rights indicators (whether the 
 
58 Data from conversation and e-mail with CELI Central COR, 30 June and 27 July, 2016.
66 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
municipality has been “microfocalized,” the first step in land restitution; and progress in 
municipal zoning) from 26 points in 2011 to 46 points in 2014. But these data do not reflect 
titling challenges, land concentration, or progress on restitution once the government begins the 
process with microfocalization. 
The CELIs maintained records pertaining to restitution cases and the formalization of property 
ownership. The CELIs have reached or exceeded their goals in most, but not all, cases.
59 The 
CELIs have supported State institutions involved in the land claims, despite the fact that the 
latter posed considerable challenges at times.
60 
At the beginning of its project, CELI MdM found hundreds of abandoned properties in its 
municipalities due to displaced communities and the many problems of informal titling and a 
lack of understanding in terms of ownership rights. CELI already worked closely with the 
relevant agencies, prior to the adoption of the Victims and Land Restitution Law, in 2011. CELI 
MdM launched the “Historical Analysis of Land Rights” Project61 to document the connection 
between the campesinos and the land, particularly victims of violence and displacement. This 
study served as a basis for subsequent work in terms of creating methods, checklists and 
templates to trace a map of the multiple routes that the communities and the families would have 
to follow to obtain legalization and compensation.62
In FY2013, CELI N/S sponsored the establishment of a local land titling office in alliance with 
the mayor’s office in El Bagre, Bajo Cauca. Within a short time span, the office “received 100 
applications from campesinos for unregistered derelict land or unoccupied land owned by the 
State requested before INCODER, as well as 11 land titling applications for the public use of 
derelict land.”63 In the same period, this CELI helped the Unidad de Restitución de Tierras
(URT) in Valdivia to process 627 cases in the Register of Confiscated and Abandoned Land,64
and, in Tumaco, it supported the updating of the property register and the establishment of a land 
committee to supervise the formalization and regulation of property titles.65 
CELI Central also worked on strengthening the URT’s institutional capacity in the areas in 
which it works by contracting 17 arbitration experts in 2014 to support restitution cases.
66 CELI 
MdM similarly provided support in the positioning of the URT in its focus municipalities, 
granting funds for personnel, office support and technical assistance to set up offices in Sincelejo 
and Carmen de Bolívar.67 CELI Central also reported the experience of providing TA to 
INCODER from 2012 to 2014, in topographic surveys and cadastral appraisals. INCODER then 
 
59 CELI N/S “The goal of 2600 formalized properties supported in the CELI municipalities has not been reached,” and there are 
only six months of the project left. For details on PMP indicator goal attainment, please see Annex 5.
60 For example, in 2012 CELI N/S reported difficulties it faced reaching cooperation agreements with any State agencies 
involved in land tenure issues (CELI N/S, Quarter IV and Annual Reports 2012, Chemonics, Inc., p. 13). 61 Análisis Histórico de Derechos sobre la Tierra. 62 Colombia Responde MdM Final Report, Global Communities, April 2015, pp. 22-26. 63 CELI N/S, Quarter IV Report, July-Sept 2013, Chemonics, Inc., p. 18. 64 Registro de Tierras Incautadas y Abandonadas. 65 Ibid. 66 CELI Central October-December 2014 Quarterly Report, Tetra Tech ARD, p. 44. 67 Colombia Responde MdM Final Report, Global Communities, April 2015, p. 26.
67 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
bought 107 lots totaling 8,267 hectares which they distributed to indigenous groups in the 
Consejo Regional Indígena del Cauca (52 lots), the Autoridades Indígenas de Colombia (2 lots), 
and the Asociación de Cabildos Indígenas del Norte del Cauca (1 lot), as well as to campesinos
(36 lots) and Afro-Colombians (16 lots). The effort was designed to resolve land-related conflicts 
among these groups in Cauca.68
Beyond institutional support, the CELIs have provided assistance to communities and 
organizations in specific cases of the legalization of property titles, and on supporting families in 
building their cases to request restitution. They supported civil society organizations (JACs, 
producer associations, and women’s organizations) to help people understand the often complex
steps involved in formally registering their land and guaranteeing their titles. For example, in 
2014, CELI N/S mobilized the Universidad Católica del Norte to create a specialized training 
program with 61 participants from seven producer associations who received 120 hours of 
training in the norms and rights involved in land ownership and how to gather the necessary 
documents to present to the authorities to obtain property titles and registers.69
CELI Central worked with producer associations in the north of Cauca and southern Valle, 
where the campesinos frequently only have informal titles, to carry out formalization workshops
on the legal processes involved in restitution and titling. The participants visited priority parcels 
to collect information, and these cases are already being considered.70
At baseline, 24% of HHs in CELI municipalities had titled lands (averaging 5 hectares) and 44% 
said they had land but no title. In 2015, 42% of respondents in CELI municipalities said they had 
titled land, and 40% said they had land but not title. In control municipalities the same dynamic 
appeared, with 35% said they had titled land at baseline, and 57% at midline. As the trend was 
the same in control and treated municipalities, there is no impact attributed to any intervention. 
Land ownership is informal across CELI, PNCRT and control municipalities. The priority 
problem continues to be that of legalizing the acquired rights through succession or the land 
market (sale or purchase); such lands represent up to 85% in CELI and 94% in PNCRT 
municipalities, indicating that the problem is not one of access but rather one of formalization of 
titles. 
There is a predominant absence of formal deeds and land tenure relies on precarious titles, as 
highlighted by members of the focus groups: 
Very few people have deeds and most only have a sales letter… like 40% have deeds, or 50-50 
more or less. (Producer, Meta, Control) 
Generally the people don’t have formal titles here. So much so that the municipal administration 
has lands that still aren’t in the municipality’s name. (Functionary, Antioquia, Control) 
 
68 Data from conversation and e-mail with CELI Central COR, 30 June and 27 July, 2016, referencing their activity LND-02385-
12 - APOYO AL PROCESO PREVIO DE ADQUISICIÓN DE TIERRAS POR PARTE DEL INCODER EN CAUCA. 69 CELI N/S Annual Report 2015, AID-514-C-12-00001, Chemonics, Inc., p. 11. 70 CELI (Central) October-December 2015 Quarterly Report, AID514-C-11-00002, Tetra Tech ARD, p. 26.
68 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
If you don’t own land you are nobody. You go to a bank and the first thing they will ask you for is 
the deed. No, no, I have a sales letter. Can’t be done. (Producer, Tolima, CELI) 
Without land titling, how is anyone [going to do] a productive project? (Producer, Meta, CELI)
However, as can be seen from the above, no strong counterpart to this has existed, given that 
INCODER was not titling at the required pace. In addition, since the beginning of 2016, 
INCODER has been replaced by two new agencies which are only beginning to function. Thus, 
the problem of supporting land titling is rooted in the government agencies rather than in the 
interventions. The beneficiaries themselves are aware of the efforts made by CELI, but they also 
express their desire for the process to be improved: 
Titles are formalized… some deeds maybe. Not [as much as] we aspired to. But we have moved 
forward. At least the seed has been planted. (Producer association leader, Tolima, CELI)
Here we really haven’t had the required support from the corresponding entity, INCODER. There 
are very few parcels with titles in the Macarena, which makes many interventions avoid the area, 
because having title is a requirement. So, yes, we need more help on this. (CSO leader, Meta, 
CELI)
We made progress with Colombia Responde and Consolidación. Work was carried out in 2012... 
[but] the program could have been better. (Producer, Tolima, CELI) 
Conclusion 12: The informality in ownership rights sheds light on the limitations of the State. 
The communities deal with this problem through precarious ownership rights, based on fragile 
documents such as sales letters. The absence of formal ownership titles renders such documents 
invalid for use in guaranteeing credit, and discourages development of productive projects in the 
long-term. The advances in the titling procedures in the CELI and PNCRT municipalities
continue to be limited in the face of the huge scale of the problem. Although the CELIs can 
support such processes, it is ultimately a matter for the State to resolve. 
Recommendations
1. Maintain realistic expectations at a level that the operators can affect (conclusions 1-3, 6, 
10 and 12) Impact indicators are at a high level of generality in the results framework, relative to 
the investment and interventions, the short duration of implementation, and the problematic 
contexts. Given that the economic situation reflects regional trends and the regions’ critical 
structural deficiencies, it is important to not consider lack of impact as representing a failure in 
programming. 
2. Make simultaneous improvements in production and trade (conclusions 2-3, 5-7, and 10) 
Increased production must go hand-in-hand with market improvements. Low quality, intermittent 
and seasonal production is not attractive for a market that exports goods to the rest of the 
country. Some CELI activities striking alliances between the private sector and producers’ 
associations have made important qualitative gains for both sides, and should be emulated. In 
contrast, good quality and regular production has better results when that production is directed 
at adequate markets. Transport to markets is never guaranteed. The GOC must endeavor to 
resolve market and infrastructure inadequacies and establish programs that improve relationships 
with commercial allies and intermediaries. This is and must be one of the most important 
concerns for post-conflict policies in Colombia. While international funding is certainly 
important, it may be wise for the latter to lead small focused pilots that treat these interconnected 
69 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
structural failings holistically, or to support GOC efforts that do so, rather than dilute the effects 
by intervening across broad geographic areas with expansive mandates. 
3. Support an array of markets (conclusions 5, 6 and 8) It is important to support 
commercialization projects designed to improve both regional markets and projects that export 
goods to other regions, which may provide a solution to local overproduction and low prices. In 
such cases, specific programs are needed that intentionally bring small-scale producers and 
intermediaries together in the productive process to work with broader marketing efforts to 
improve quality or to involve banks to increase access to working capital. 
The intermediaries, if necessary, can constitute the starting point for a good system of trade in 
the regions given that they fulfill an economic function that would be difficult to replace. It is 
important, however, for the intermediary not to become monopsonistic in a way that reduces 
smallholders’ bargaining power. Intermediaries risk their capital in money and equipment and 
come up against insufficient quality and quantity in terms of what they buy from the producers. 
As such, specific programs have to be structured to understand their function and improve it to 
benefit the regions. 
Better market information adjusted to local conditions would contribute to improving the 
communication of supply and demand, and to bridging the cost difference produced by 
intermediation. Regional market information systems should involve provincial centers or the 
Ministry of Agriculture and Rural Development’s (MADR) Unidades Municipales de Asistencia 
Técnica Agropecuaria, or UMATA. Mobile phone access to up-to-date market information is 
easy to implement and can function well, as seen in mobile data improvements in other 
countries. 
4. Plan for intermediate food security in productive projects (conclusion 4) It is recommended
that CELI and USAID make concerted efforts to ensure that programs supporting productive 
projects also support food security. This situation should be anticipated by providing TA and 
traditional farm or other sustenance activities that minimize dependency. The productive projects 
must combine short- to long-term returns in order to ensure cash flow and project sustainability
at the household level. 
5. Reach the FHHs with interventions that work (conclusions 2 and 9, and the relevant section 
in the chapter on social development) Overall, the benefits have not been the same for FHHs 
and MHHs. Given that the two household types learn and use their learning in different ways, 
models should be developed that reflect this difference and provide different services to take 
advantage of the characteristics and behavior of each. For example, the FHHs attend training 
workshops and courses to a lesser extent, but they report taking greater advantage of associations 
for different productive ends and for marketing. 
6. Maintain activities to expand financial services (conclusions 7-9) Revolving funds are a 
suitable instrument to deal with the communities’ borrowing needs but they must be protected 
against the risk of non-payment. While communities have access to other programs such as 
MADR’s Productive Alliances program, the communities should be supported in strengthening 
these revolving funds in the long-term. Credit decreased globally, according to the household 
survey. However, according to focus group interviews, such forms of support have fulfilled a 
vital function for the development of savings and credit schemes, and they are important as long 
as they are conditioned or linked to the creation of collective savings and credit funds that 
70 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
maintain strong principles of governance to guarantee their sustainability. Such a framework 
avoids diluting resources by creating dependency. 
Steps to undertake would include creating a system that monitors and follows-up the communal 
funds or revolving funds; sets up discussion forums and exchanges of experiences; creates 
incentives or awards for performance; and draws attention to successful cases. 
7. Continue focused work on roads (conclusions 5, 10 and 11) Although the improvement of 
the roads system is the responsibility of the central and local governments, the CELI programs 
should continue to reinforce the successful support they provide to the JACs pertaining to 
institutional strengthening and community infrastructure. In particular, it should be a 
participatory community-based model for the management of the tertiary road network, in line 
with international experience, taking into account the national and environmental contexts. 
8. Continue to expand the work in terms of land tenure formalization (conclusion 12) It is a 
priority to facilitate relationships with the new state organs that replaced INCODER in order to 
continue work on land tenure as part of USAID’s ongoing portfolio. To supplement current 
information, the evaluation team recommends a transparent monitoring system to focus attention 
on the issue. USAID should work with the GOC on their database of petitions, assessing the data 
over time and comparing regional performance. The database should register the location, the 
number of hectares, produce management and resolution rates. 
A corollary issue with land restitution is that, in the event of an eventual peace accord and a 
greater number of successfully resolved land restitution processes, conflict over given parcels 
may increase as new or changed IAG fight for local dominance. It will be important to consider 
these kinds of conflicts in connection with the security recommendations in that chapter. 
71 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
SOCIAL DEVELOPMENT
Evaluation question: What changes have been produced by CELI interventions to 
strengthen the organizational capacity of local CSOs in local areas?
SUMMARY: The CELIs documented important increases in organizational competency 
scores – especially in business initiatives and services – and surpassed program targets.
From citizen perspectives, however, there were no increases in the social capital index 
and a slight statistically significant decrease in the trust component, led by a greater 
drop in FHHs. Trust in state institutions fell across the board except for the personería in 
survey results, though qualitative data challenged the depth of this decline. Participation 
in the JACs declined, but participants who stayed played a greater role in decision￾making. Surprisingly, the survey revealed an apparent decline in CELI municipalities’ 
involvement with producers associations. In both CELI and control municipalities, 
respondents attested to a dramatic increase in productive capacities. Participation in 
women´s organizations and victims’ associations increased, and ethnically oriented 
cultural associations were found to strengthen a sense of identity and build social capital. 
Strengthening civil society and social development is considered fundamental to the 
consolidation of conflict-affected areas in Colombia. Social development is understood as the 
population’s participation in civil society, trust between peers, and trust between citizens and the 
state. It may also be understood as the degree of trust characterizing the relations between social 
actors, the prevailing norms of civic behavior, and the level of association activity (Kliksberg, 
1999, p. 85). To complement this position, Kliksberg cites James Coleman’s theory (1990) that 
social capital is manifested individually and collectively, explaining his position in the following 
terms: 
The first of these is connected with an individual’s degree of social integration and his network of 
social contacts: it implies reliable relations, expectations of reciprocity and forms of behavior, and 
improves effectiveness at the private level. It is also a collective good, however. Thus, for example, if 
everyone in a particular neighborhood shares tacit standards of non-aggression and concern for 
others, children will be able to walk to school in perfect safety and social capital will be producing 
public order (Kliksberg, 1999, pp. 85-86).
Social development thereby acts to strengthen trust between peers, which may be transformed 
into processes of cooperation and associativity depending on local characteristics. The CELI 
interventions are intended to develop and strengthen social capital in target populations as a tool 
for territorial consolidation. It is made up, therefore, of 
…[t]he aggregate of the actual or potential resources which are linked to possession of a durable 
network of more or less institutionalized relationships of mutual acquaintance and recognition (…) 
The volume of the social capital possessed by a given agent thus depends on the size of the network 
of connections he can effectively mobilize and on the volume of the capital (economic, cultural or 
symbolic) possessed in his own right by each of those to whom he is connected (Bourdieu, 2011, 
221). 
Three approaches are used to examine the impact of interventions using the CELI model to
strengthen the capacity of civil society organizations (CSOs). The first approach used a tool 
drawn from the CELI toolkit: the Índice de competencias organizacionales, or Index of 
Organizational Competencies (ICO) or the ICO-VEO (Index of Organizational Competencies-
72 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Verification of Organizational Status).71 The CELIs used these tools during implementation to 
assess the organizations before and after implementation, to understand the changes that 
occurred over time as a result of their activities. The indices are composed of variables that 
capture organizational capacities, weighted for their perceived importance. The indices provide 
information on significant changes related to the CELI interventions. 
The second approach calculates impacts relative to control areas, based on the survey questions. 
In the survey, social capital was measured using different batteries of questions, combined to 
create a social capital index. This index examines two aspects: trust between peers and between 
citizens and the state; and participation in social organizations. The index allows an analysis of 
impact. 
The third approach used in the analysis of social capital involves the subjective experiences of 
citizens, some of whom were members of CSOs in the areas of intervention. Qualitative 
interviews and focus groups were carried out in 21 municipalities. The instruments used to 
capture these perceptions and experiences included topics related to direct participation in social 
organizations and productive associations, and in particular worked to capture changes in social 
capital resulting from participation in interventions, the reasons for which organizations had been 
established, the challenges they faced, their achievements and the effects they may have had on 
the population. 
One final, external source is the Consolidation Index data on participation and social capital. The 
former is measured by the operativity of certain municipal spaces for participation, the 
percentage of municipal funds subject to legally constituted citizen oversight mechanisms, and 
an electoral risk index. Social capital is measured in the Index by the degree to which Juntas de 
Acción Comunal and local social organizations have registered themselves annually.
Figure 4.1. Secondary data: Consolidation Index scores for social development 
Social capital has had a net gain of less than one point, while participation as defined by the 
Index has improved significantly. It is important to note that one of the indicators (electoral risk) 
 
71 CELI interventions employ variations of the instrument according to their needs and priorities. In this report, the abbreviation 
ICO is used to refer to all variations, as the indicator is shared by all the CELI interventions and has that name. Details on the 
CELIs’ attainment of their goals in this and other indicators are included at Annex 5.
30,4
59,7 58,5 60,8
0,0
20,0
40,0
60,0
80,0
2011 2012 2013 2014
Participation
Social capital
73 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
used by the Consolidation Index has not changed since 2011; therefore that data point is null. 
Overall, the municipal perspective, represented by the Index, is somewhat positive on social 
development. 
There are, at times, differences between the qualitative and quantitative findings presented 
below. These emerge from the different sources consulted in the field using each method. The 
qualitative work was carried out with an intentional sample, chosen for their participation in 
producer organizations, their status as conflict victims, or their functions in working with civil 
society or cultural organizations. By contrast, the survey results represent an average of the 
impacts across the quantitative sample, chosen at random within veredas. 
Findings and conclusions
The Index of Organizational Competencies (ICO)
CELI interventions are intended to create concrete improvements in the competencies identified 
as essential to organizations’ effective functioning. The CELIs created and refined tools to 
monitor changes in these organizational competencies: internal organizational democracy, 
service provision, financial development, management, administration and human development. 
This package of variables generated a scoring system for each of these capacities that were then 
weighted and added together to obtain a maximum score of 100 points. This score was a key 
indicator for the CELIs’ performance measurement in social development, and all of the CELIs 
met or surpassed their goals. The CELIs used the resulting indices to grade the producer 
organizations, establishing a baseline that enabled the organizations to measure their subsequent 
evolution. The application of the tool was participatory insofar as the leaders of the organizations 
were invited to express their opinions on the achievements of their organizations in each 
category. Although there were no comparison data available for control organizations, it is a 
useful tool for analyzing progress against a defined set of indicators. CELI staff met their 
contractual obligations by using these indices to measure the progress of the actions vis-à-vis the 
organizations. 
Consultants or members of each regional CELI applied the ICO in a participatory manner that 
involved the leaders and members of the organizations in order to ensure high levels of 
representativeness. This process also reduced the subjectivity and bias that might otherwise be 
associated with the process. The tool serves not only to measure the capacity of the associations 
but can also be used to develop an action plan for each organization that takes into account their 
strengths and weaknesses in each subset of variables. 
CELI Central reported to USAID in 201672 on the results and average changes in performance 
during the interventions for a total of 52 organizations. On average, there was an approximate 
 
72 ICO-VEO scores (for CELI N/S) and ICO scores (for CELI Central) will be updated at the end of Fiscal Year 2016, in 
October, 2016. These later scores will show progress in the current fiscal year.
74 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
50% increase in the ICO. (Figure 4.2)73 
Figure 4.2. CELI Performance: Average change in ICO score, CELI Central
Source: Operational data, CELI Central
The scores may be disaggregated by battery of competencies, as in the case of the scores for 
democracy and management. Unfortunately, this information was not available in digital form at 
the time the evaluation was conducted for this CELI. 
The ICO-VEO is based on information concerning democracy and participation, management 
and administration, business initiatives and services, and assets and human development.
74 CELI 
N/S reported that it had worked with 89 organizations in 2016 with similar scores and levels of 
growth. (Figure 4.3) 
Figure 4.3. CELI Performance: Average change in ICO-VEO scores, CELI Norte/Sur
Source: Operational data, CELI N/S 
The data include scores for individual competency sets. The overall point increase is therefore 
divided into the following categories, as shown in the following figure: 
 
73 The ICO score equals 100, so the scale used in the whole of this section and in the figures runs from 0 to 100. 74 Taken from the VEO document: Una herramienta de diagnóstico para el fortalecimiento de organizaciones de productores, 
n.d., USAID, Perú.
41,1
60,6
19,5
0
20
40
60
80
Start End Change
43,2
67,8
24,5
0
10
20
30
40
50
60
70
80
Start End Change
75 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 4.4. CELI Performance: Average changes by ICO-VEO category, CELI N/S
The most notable increase for this CELI was in the category of business initiatives and services 
(essential for sustainability), for which the organizations covered by the initiative scored an 
average of almost 8 points. This CELI’s ICO uses a diagram of a house75 to visualize the scores, 
alongside a weighting that assigned a higher score to this component (25% of the total). The 
situation was similar for the human development component, which showed the second highest 
increase, while the components of assets and management were emphasized less in the weighting 
protocol and also showed lower levels of increase for the period examined. However, all scores 
were positive, representing an increase of almost 25% over the course of the interventions. CELI 
N/S also reported that the visualization of a house aided organization participants to take 
ownership of the instrument, and that the process of supporting organizations to develop the tool 
was highly participatory: “we looked at organizational development from their perspectives… 
and left installed capacity on how to use it. It’s not just that we adapted it, but how we went 
about it,” according to Chief of Party José Félix Montoya.76 With the software and the other 
tools made available to the leaders, it is hoped that these groups will continue constructing their 
organizational “house”. One organization, Asocaval, supported by CELI Norte, has become 
adept with the instrument to the point where the government of Antioquia has contracted them to 
support other producer associations in the appropriation of the instrument. 
Although the measure used by CELI MdM was not available to the evaluation team (because the 
project closed before the evaluation began), the progress figures appear in its reports and 
demonstrate a similar positive trend. Working with 128 organizations during the life of the 
project, the reports state an average change of 54 points up to September 2012, of 76 points up to 
 
75 For an explanation of the diagram see “Una herramienta de diagnóstico para el fortalecimiento de las organizaciones de 
productores,” Verificación del Estado Organizacional. 76 In a meeting on August 30 with the CELI N/S team and Chief of Party José Félix Montoya at their headquarters, translated by 
the authors from the Spanish conversation.
76 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
September 2013, and of 105 points to the end of December 2014.77 
Some respondents noted the diagnostic utility of the index for identifying strengths and 
weaknesses in the organizations and for encouraging members to appropriate proposed changes. 
Another respondent noted that the ICO prioritizes the examination of the internal workings of 
organizations, while emphasis could more valuably be placed on the way the organization acts 
within its environment (services, revolving funds, commercial relations, and/or credit unions). 
This respondent felt the latter perspective would be more useful in generating concrete actions or 
improving income, focusing as it does on how the organization should position itself in relation 
to the market and sell its products. 
Conclusion 1: This approach provides evidence of significant increases in organizations’ 
competencies. An outward focus on markets and services appears to be important to 
sustainability, balanced with support to the organizations’ internal management and other 
features. The methodology employed by CELI N/S appears to have encouraged organizations’ 
taking ownership of the tool. 
Social capital: Trust
The social capital index is composed survey questions comprising trust (60 points) and 
participation (40 points – see the next section of the report), for a total of 100 points. Trust was 
identified at different levels: family, neighbors, friends, the local Community Action Council 
(JAC), legal and state oversight bodies, development institutions, local and national government, 
and the police and military. 
As is apparent, many of the factors affecting the index scores are beyond the remit of 
cooperation programs such as CELI and should be considered more as indicators of contextual 
conditions when employed so soon after the construction of the baseline. Nevertheless, analysis 
of program level subcomponents provides an increased understanding of program strengths and 
weaknesses.
In general terms, the social capital index did not register any impact either in the CELI or in the 
PNCRT municipalities, but the percent changes are mostly positive. The percent changes are 
shown in Figure 4.4 for CELI municipalities, and the index impacts in the table that follows. 
(Figure 4.5) 
Figure 4.5. CELI Performance: Social Capital Index, total and regional 
Ind. Description Regions Baseline Midline % Change Source
DO3-
024
Social Capital 
Index
Total 22.8 23.63 4% Household survey
Montes de María 20.6 22.60 10% • Baseline 2013, 
 
77 Quarterly reports dated July-Sept 2012, July-Sept 2013 and Oct-Dec 2014 made available by USAID for the purposes of the 
evaluation. As the Project has ended it is not possible to inquire how the average came to exceed 100 points, but it is assumed 
that the score was aggregated over the years of the intervention.
77 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Ind. Description Regions Baseline Midline % Change Source
Central 23.8 24.33 2% Econometría S.A. 
• Midline 2015, 
Proyecto EVAL Norte 23.8 24.92 5%
Sur 22.94 20.82 -9%
The sub-indices of trust, on the one hand, and participation, on the other, did demonstrate 
impacts (see the following section for a detailed examination of the participation component). 
For the component of trust, there was a negative impact of 1pp in CELI municipalities, while in 
PNCRT municipalities no statistically significant effect was found. (Figure 4.6) 
Figure 4.6. Impacts on the Social Capital Index, CELI and PNCRT
Variable CELI PNCRT
Social Capital Index -0.24 -1.16
Trust component -0.99** -1.00
Participation component 0.74** -0.16
Source: Authors’ calculations. *** At 99% confidence, ** at 95%, * at 90%.78
Sex-disaggregated analysis shows that households with female heads of household (FHHs) fell 
more than households with male heads (MHHs) (by one point out of 60, at a significant level). 
Viewed by region, in CELI Central, municipalities showed a significant increase of more than 
1pp, while in the other regions there was no impact on the index. 
The trust sub-index was constructed in part from three survey questions, shown in Figure 4.7. 
With one exception, the data do not register any impact either for the PNCRT or the CELI. 
Figure 4.7. Impacts on trust between peers, CELI and PNCRT
Variable CELI PNCRT
High level of trust that a neighbor who promises to help them would do so -0.0356 -0.0185
The tradition of working collectively on community projects exists here -0.0111 -0.092*
High level of trust in members of the community -0.0131 -0.0514
There was no significant impact on trust in the CELI municipalities; in general, an absolute 
decrease appears for these three questions (which also occurred in the control municipalities). 
There was a negative impact of 9pp in the PNCRT municipalities. For FHHs, the impacts were 
negative for the question on the tradition of collective work in community projects (-11.5pp for 
PNCRT municipalities and -6.5pp for CELI municipalities). In PNCRT, these households also 
 
78 Throughout, the data source is the evaluation’s survey data, unless otherwise noted. Standard significance notations apply: *** 
for 99% confidence, ** for 95%, * for 90%. Impact estimation models are in Annex 3, and data in the text are tabled in Annex 10.
78 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
responded that they did not have much trust in other members of the community (-7pp impact). 
The trust index also included a battery of questions on how much trust respondents have in state 
institutions, from the national level to the local. Uniformly, across the 25 institutions that the 
survey asks about, trust worsened across the full sample (CELI, PNCRT, and their controls). The 
effect was substantial for all trends at national, departmental and municipal level and for the 
JACs and social organizations, the legal and control bodies and the educational, agricultural and 
security institutions. As the effect is generalized, it cannot be attributed to any intervention; 
rather, citizen perceptions have deteriorated notably, likely affected by sustained, external events 
(absence of state presence, lack of services, scandals, etc.). 
One contrast was found in CELI zones, where perceptions of the offices of the personería 
municipal (Ombudsperson offices) fell less than in the controls (amounting to a positive and 
significant intervention impact). This is confirmed by the results from other survey questions 
(Figure 4.8), in which CELI municipalities show positive impact. Results are consistent across 
the range of survey questions: nearly 5pp of positive impact are found in CELI municipalities 
relative to controls when respondents are asked about their trust in the personería, 17pp positive 
impact in respondents’ rating of the personería, and 10pp positive impact when asked about 
change over the last two years. 
Figure 4.8. Impacts on opinions of the personería; Trends by percent change
Treatment Control
Baseline Midline Baseline Midline Impact
I have a high level of trust in the personería
CELI 28% 22% 28% 17% 0.0465**
PNCRT 21% 14% 33% 28% 0.0616
The personería is good or very good
CELI 45% 52% 46% 41% 0.172**
PNCRT 38% 39% 52% 53% 0.0312
Services at the personería have improved over the last two years
CELI 23% 29% 17% 13% 0.101*
PNCRT 13% 18% 20% 28% 0.00594
The positive results are concentrated in the CELI Central region, where the positive impact was 
almost 12pp on improvements in the personería in the last two years, and over 7pp in the 
proportion of households that rate the personería as good or very good. Other CELIs report not 
having focused on direct involvement with the personería offices, or having provided support 
that was more logistical or equipment-based. Those zones show lesser or no impacts on 
respondents’ perceptions of the personería. 
While the survey shows no impact on the levels of trust felt by members of the population 
towards their peers or their government, the qualitative fieldwork showed a different reality. 
Save some exceptions, trust in the institutions is a constant among the qualitative respondents in 
focus groups and interviews. They reported that trust and solidarity do exist between neighbors 
79 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
and that this has facilitated social groups. In regions where armed groups operate, the trust 
between peers not only improves the quality of life but also permits the creation of associations 
and strengthening of the JACs. This finding is transversal inasmuch as it permits all kinds of 
associativity, whether productive associations, victims’ organizations, cultural or youth groups, 
or others. 
In my vereda, I feel good about my community. As I’ve said, I’m the President of the Council 
(JAC) and when I need to organize a meeting people collaborate; we look out for each other and, 
I don’t know, I feel at home here in my vereda. (JAC leader, Antioquia, CELI) 
As pretty much all of us know each other (in the vereda), well, there is more access, right? 
Because if there’s an event or something, well there it is, and everyone knows each other, so it’s 
easier there. (CSO member, Cauca, PNCRT) 
The municipal personerías, secretaries of government and leaders of social organizations say 
that trust in social and community processes has increased as a result of increased security levels 
in the regions, enabling people to associate freely without fear of reprisal from the armed groups
and the strengthening of associations of victims formed to demand their rights. This contradicts 
the survey data but likely reflects their experience working with organization members, who tend 
to participate more and show more trust. 
In terms of the security situation, and its relation to social capital in Montes de María, it is 
suggested that the improved levels of security in the region have permitted the emergence of 
social organizations. When conditions of violence in a region represent risks for the inhabitants, 
social processes dwindle. An interviewee from Carmen de Bolívar, a CELI municipality, said: 
Leadership was silenced. They killed the leaders... (When) the conflict was beginning to decline a 
little, they (re)emerged. (CSO leader, Bolívar, CELI)
As suggested in the field, when the security conditions began to change different social processes 
began to emerge, because there were fewer risks associated with the leadership of community 
organizations. Similarly, the survey shows a positive impact on population perceptions in CELI 
municipalities. This contrasts with the impact in the municipalities in which only the PNCRT
intervened, where there were no impacts registered. The proportion of HHs that felt that police 
security services had improved over the previous two years increased by 8.5pp relative to 
controls, while the households that considered the military security services had improved over 
the same period increased 8pp compared with their control municipalities. (Figure 4.9)79 
Figure 4.9. Impacts on perceptions of the police and army, CELI and PNCRT
Variable CELI PNCRT
Police security services improved last two years 0.0851** -0.0149
Military security services improved last two years 0.0813** 0.0385
 
79 The increase is mitigated by overall perceptions of the security situation and of security forces. There was no impact on 
whether the security situation had improved in the veredas or corregimientos. Nor was there impact on police ratings, nor on the 
percentage that would recommend that a family member had left the zone should return. 
80 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Despite the recognition that the services provided by the police and the army had improved, a 
range of conflicts is resolved within the community, as the legitimacy of the JACs and the 
conflict resolution tools they employ lead many people to resort to the JACS for conflict 
resolution, rather than to higher-level institutions such as the municipal personería or the police. 
Thus, the impact in the PNCRT municipalities on households that report that they approach other 
members of the community when dealing with serious problems with a neighbor is an increase of 
2pp while there was no impact in CELI municipalities. (Figure 4.10) 
Figure 4.10. Impacts on resolving conflicts in the community, CELI and PNCRT
Variable CELI PNCRT
Facing serious problems with a neighbor, they approach other 
members of the community 0.00424 0.0203**
This information indicates that the CELI interventions and the PNCRT have, to different 
degrees, permitted community ties and resolution tools for minor conflicts to deepen within the 
community, contributing, for some, to increased trust. 
Alongside qualitative reports of increased faith in JACs and social institutions, some respondents 
revealed low levels of trust for municipal institutions.
80 In the CELI municipalities no impact 
was found on perceptions of the performance of the mayor’s office while in PNCRT
municipalities, positive perceptions of mayor’s office performance dropped 6.5pp. (Figure 4.11) 
Figure 4.11. Impacts on perceptions of mayor’s office performance, CELI and PNCRT
Variable CELI PNCRT
Mayors’ office performance is good or very good -0.0187 -0.0653*
The importance of the JACs to the population resides in the fact that these are the mechanisms 
through which the population organizes itself to carry out the tasks that are not performed by the 
State. As the inhabitants of the veredas in one of the PNCRT municipalities expressed it, they 
feel obliged to organize themselves and work together in order to maintain and repair the roads. 
In our vereda it’s like this: members have the right… in our case we all have to meet together and 
if a road has to be repaired then we get together and between us we all contribute, but if someone 
is not a member, they don’t contribute (JAC member, Putumayo, PNCRT) 
Survey results showed that 4.4pp more respondents in CELI municipalities trusted their JACs a 
lot or quite a lot, compared to baseline and controls; however, the qualitative data showed even 
greater increased trust in the JACs. The greater increased trust among qualitative responses may 
be explained by the fact that the quantitative information represents the average opinion of the 
 
80 Please see more detail on this in the Governance and institutions chapter.
81 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
population while the qualitative information is a purposively selected sample of active, 
participating people. In any case, the value assigned to the JACs is related to the fact that for the 
population they represent access to training, productive projects and other initiatives that the 
government authorities or international cooperation agencies may develop. The JACs represent 
the gateway to many state services and projects, as they are the closest point of contact the 
population has with official institutions. This position is clear from the following testimony: 
The Juntas de Acción Comunal have been a fundamental axis of this community. It is an honor to 
be the president of a Junta de Acción Comunal and they are involved in all the training events, 
they create Councils, there is training on human rights... (JAC member, Meta, CELI) 
Conclusion 2: The survey data show a marked decline in the level of trust that citizens have in 
the organs of government, in all the municipalities surveyed during the study, and for all the 
government bodies included in the survey. This effect is not attributable either to the CELI or to 
the PNCRT interventions because it is common to the control sites as well. Among the more 
actively participating community members who were interviewed in qualitative data collection, 
there are reports of positive change and improved relationships between citizens and the State. 
However, the latter group of respondents is much smaller. 
The survey data also show a decline in trust levels between peers and neighbors, especially in 
FHHs. These figures provide a strong contrast with the experiences narrated by the participants 
during the qualitative work, who had higher levels of participation and much more positive 
opinions of the social fabric – but, again, these are active participants and so, by design, they are 
more likely to perceive and take part in the construction of the social fabric. 
Conclusion 3: In zones where the security conditions have improved it is also possible that 
levels of trust between neighbors have, to a degree, improved. Not only does this increase the 
quality of life but also permits the creation of new forms of association and the use and 
strengthening of existing organizations, such as the JACs.
Social capital: Participation in associations
The CELI strategy encompasses three levels: i) the strengthening of social capital in the regions 
it covers, ii) the generation of relations of trust between neighbors and iii) the promotion of 
cooperative relations within communities. Working towards this objective, the CELIs have 
supported and strengthened different forms of association through a range of interventions 
focused on existing grass roots organizations such as the JACs, which are the prime instance of 
participation in public life in the municipalities. Equally, productive associations have been 
created or strengthened in an effort to construct alternatives for communities affected by this 
phenomenon. Another focus of intervention was the support provided to victims’ associations 
with the intention of generating collective processes to ensure their rights are respected under the 
terms of Law 1448.
81 Finally, the CELIs have strengthened the cultural associations present in 
 
81 Law 1448, also known as the Victims and Land Restitution Law, was signed by President Santos in 2011 in order to “facilitate 
the restitution of millions of hectares of lands abandoned or stolen as a result of human rights abuses and violations” (The 
82 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
the areas of intervention; these operate as a way of building trust, collective identity and 
belonging in the territories, rooted in the intangible heritage of the community.
As mentioned above, the social capital index is composed of measures for trust and participation 
(the former accounting for 60 points out of 100 and the latter 40). Participation is measured in 
respondents’ reported affiliations: interest groups, producer associations, political parties, the 
JACs, community organizations and citizen oversight processes (veedurías). 
The survey data show that, across the full sample, there was a moderate but significant increase 
in household participation in community activities. The impacts were slightly more significant in 
CELI municipalities, where this sub-index was positive in almost 1 point (out of 40); there was 
no impact in the PNCRT municipalities (Figure 4.6). When the results were differentiated by 
gender there was no evidence of difference between FHHs and MHHs. 
An analysis of seven specific kinds of participation showed that participation increased by 
between 0.5pp and 9.3pp in the CELI municipalities (Figure 4.12). The impacts were smaller in 
the PNCRT municipalities and fewer kinds of associations showed significant impact: voluntary 
groups (5pp), sporting or cultural groups (3pp), and women’s groups (4pp).82 
Figure 4.12. Impacts on participation in community activities, CELI and PNCRT
Variable CELI PNCRT
Oversight groups (veedurías) 0.013*** 0.0048
Community meetings in general 0.093** -0.0035
Voluntary groups 0.049* 0.069***
Sporting or cultural groups 0.033*** 0.032**
Parents associations 0.059*** -0.028
Security groups 0.006*** 0.005
Youth groups 0.008** 0.005
Women’s groups 0.012 0.036***
Political party or movement 0.026* -0.012
Agricultural school groups 0.0149** -0.0885
Data disaggregated by sex shows differences in participation in different kinds of organizations. 
In general, MHHs reported increases in participation, while their equivalents with FHHs reported 
more than a 50% reduction in participation. In the CELI municipalities, MHHs increased their 
participation in the JACs by almost 4pp, while FHHs reported a reduction of 10pp. In the 
 
Victims Land and Restitution Law, p. 5). For more information, see the Amnesty International report on the law’s land restitution 
provisions at http://www.refworld.org/pdfid/4f99029f2.pdf.
82 In six other kinds of organizations there was no impact either in CELI or in PNCRT municipalities (see Annexes for detail).
83 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
PNCRT municipalities, FHHs were 13pp less likely to participate in the JAC or other community 
groups, while MHHs increased their participation by 10pp. 
In all CELI municipalities MHHs increased their participation in cultural groups by 
approximately 4pp, while FHHs presented no change. While CELI municipality HHs reported 
nearly 4pp increased participation in voluntary groups, the households that participated in 
decision-making were MHHs (10pp). Notwithstanding the decrease in FHH participation across 
the full sample, among FHHs in the PNCRT municipalities, participation in one type of 
organization – women’s groups – increased by 5pp. Even so, no impact was seen among FHHs 
in terms of decision-making.
Individuals who did participate reported more participation in decision-making, by between 25% 
and 40% more across the full sample. As this increase was general, it cannot be attributed to any 
intervention, but the change is positive and noteworthy. At the same time, the percentage of 
persons who report that they take part in meetings but not in decision-making fell by the same 
amount. While it is true that overall participation has fallen, this finding reflects improved 
quality in the social participation engaged in by the population. 
Conclusion 4: In the CELI municipalities, there was an important increase in participation in 
associations, but FHHs did not participate in this increase. In PNCRT municipalities, no impact 
was apparent on average participation levels. 
The particular role of the Juntas de Acción Comunal (JACs)
The Juntas de Acción Comunal or Community Action Councils are the most local level of civil 
society recognized in Colombian law: voluntary associations in urban neighborhoods or rural 
veredas that can formally constitute for civic purposes. While the three CELIs were not asked to 
report systematically on the numbers of JACs served, it is clear that each CELI developed strong 
commitments to work with the JACs over the course of the activities, both in building their 
capacity to function as more effective community organizations as well as through engaging 
JACs in direct contractual relationships to carry out specific projects.
Each CELI undertook training both of individual community JACs and of associations grouping 
multiple JACs. CELI Montes de María held 31 sessions in 17 communities, providing orientation 
in management practices, transparency, and trust building. The JACs learned skills and 
procedures that they applied to implementing projects and negotiating advantageous agreements 
with providers.83 In southern Tolima, CELI Central worked to strengthen management capacities 
of 633 JACs through training in project development, empowerment, contracting, and access to 
public services. One goal was to prepare JACs for the April 2016 leadership elections to build
local-level sustainability.84 CELI N/S engaged the Foro Nacional por Colombia to assist 65 JACs 
and five umbrella associations in Tumaco to strengthen organizational capacities, work with 
 
83 Colombia Responde Montes de María Quarterly Report July-September 2014. Bogotá, Colombia: Global Communities, p. 11. 84 CELI Central Annual Report September 2014-September 2015. Bogotá, Colombia: TetraTech ARD, p. 21.
84 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
higher levels of government, and develop plans that represent constituent goals.85 CELI N/S 
trained Bajo Cauca leaders in environmental management, land issues, communications and 
financial management.86
The training and guidance provided by the three CELI projects have empowered many JACs to 
enter into legal agreements for carrying out local community infrastructure projects. CELI 
Central has turned to the JACs as implementers in efforts to improve tertiary roads and social 
infrastructure projects, delivering approximately USD$6.2 million in contracts. In Meta and 
Caquetá, JACs have taken responsibility for tertiary road maintenance and improvement.
Supported by 10 CELI grants totaling USD$879,000, the JACs mobilized labor, food, and 
lodging for workers, raised funds and contributed cash.87 CELI N/S made infrastructure grants to 
JACs throughout Bajo Cauca and its Tumaco intervention sites. In its December 2012 report, the 
project reported that of the 45 grant agreements approved in Córdoba in that quarter, 34 JACs
and two JAC associations received funding to move forward with community projects.88
Between CELI N/S and CELI Central, dozens of JACs have taken on responsibility for 
managing and fulfilling the contractual terms of direct grants, providing these civic organizations 
with important experience in community management and infrastructure construction and 
maintenance. The CELIs report greater efficiency and efficacy in the successful completion of 
these works and contracts, under budget and within proposed time frames, with citizen oversight 
committees providing watchdog services. The capacity now resides in the hands of these JACs 
for future undertakings. 
The JACs give rise to other organizational forms, such as the producer associations. The JAC’s
importance as a starting point for associativity is reflected in the opinions obtained in Antioquia 
and Meta in response to questions about the resources received from the CELIs: 
Above all they pass resources through the JAC. There the Goat Rearers’ Association was created 
for the whole of the municipality of Caucasia. That is, veredas Brasil, Villa Mercedes, 
Corcovada, Los Tigres 1, 2 and 3, and Delirio. They operate there, and also in Villa Mercedes, 
there are people who are beneficiaries of the Association of Rubber Producers – that’s the Arias 
family, who live there; and they’re also in Asocor. (Producer association leader, Antioquia, 
CELI)
At the office of the Secretary of the Municipal Government our role is to strengthen the JACs and 
associations for different productive areas. Here there’s a bit of everything, so they have 
organized themselves for example in the area of cattle farming, cacao, rubber, fish-farming… 
And I know that all these productive associations – some of which are strong – have been 
accompanied by Colombia Responde. (Functionary, Macarena, CELI) 
In spite of the overall positive impact, participation decreased for JACs and the producer 
associations (for the latter, see the next section). This was true for both CELI and PNCRT
 
85 CELI N/S, FY 2016 Quarterly Report (Q1), December 2015. Bogotá, Colombia: Chemonics, pp. 26-27. 86 CELI N/S 2013 Quarterly Report, April-June. Bogotá, Colombia: Chemonics, p. 16. 87 CELI Central Quarterly Report April 2015-June 2015. Bogotá, Colombia: TetraTech ARD, p. 18. 88 CELI N/S Quarterly Report October-December 2012. Bogotá, Colombia: Chemonics, p. 9. 
85 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
municipalities. In absolute terms, fewer HHs belonged to or participated in the JACs: from 74% 
to 57% in the CELI zones, and from 88% to 64% in the PNCRT municipalities. The decline also 
occurred in control zones; as a result no impact is seen (Figure 4.13). It appears that FHHs made 
up an important part of that decline, as MHHs increased decision-making within JACs by 5.5pp. 
Figure 4.13. Impacts on participation in the JACs, CELI and PNCRT
Variable CELI PNCRT
Household belongs to a JAC or other community group 0.0249 -0.0593
Household participates in decision-making in JACs or other groups 0.0359 -0.0358
Beneficiaries indicated that participation is difficult because involvement in the JACs requires
dedicating a considerable amount of time and unpaid labor. In certain cases, respondents stated 
that local politicians have sought to co-opt the JACs for their own purposes. Some respondents in 
municipalities included in the intervention argued that community processes have deteriorated as 
a result of the actions of politicians who have sought to use the processes as a social base to 
advance their private interests. A focus group participant in the Guamuez Valley, Putumayo, 
said: 
The community processes, whether JACs, associations or cooperatives, whatever: they are all getting 
weaker because we’ve been weakened first by political maneuvering, second by the violence and third 
because no one listens to what they have to say. That is, the authorities don’t listen to them despite 
the JACs’ rights in the organizational process. (Producer, Putumayo, PNCRT)
In the previous quotation several different reasons are advanced to explain why social processes 
have become weaker and trust in the institutions has fallen overall, if not necessarily among 
CELI-supported JACs. These include the intervention of local politicians, violence, and the 
reduced effect that this kind of organization and these processes have on decision-making at 
local level: community processes have a voice but no vote. 
CELI activities with the JACs worked to strengthen JAC leadership and management structure, 
and their legal, administrative and social functions. Stronger organizations with better leadership 
would have greater public recognition and visibility. The CELIs also contracted with the JACs, 
as described above, which has an empowerment effect and greater respect for the leadership and 
decision-making with JAC leadership structures. Such empowerment helped to bolster the CELIs 
against political influences and extortion attempts, which are very common with external firms, 
according the CELI Central COR. 
In the veredas the JACs are sometimes the first resort for resolving conflicts between neighbors 
in processes involving direct negotiation. Beneficiaries who took part indicated that they live in 
rural areas that are far from the population centers, in municipalities where state presence is 
scarce. In certain cases the JACs occupy the vacuum left by the State and permit neighbors to 
construct trust, such as in this case where citizens reported on who they ask for help with a 
problem with a neighbor: 
Through the JAC, or you reach some agreement between the victim and the aggressor... it hasn’t
been necessary for the Ombudsperson to appear or for anyone to go to court. No, it’s dealt with 
there in the vereda… The five leaders of the Council meet with the conciliators and they call in 
the people who have the problems. They all sit down together, everybody gives them advice and 
they come to an agreement. And that’s it: the problem is dealt with. (Producer, Antioquia, 
86 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
PNCRT) 
Intervention support provided for associations has enabled communities to appropriate their 
processes through training or the building of infrastructure (activities developed by the UACT or 
one of the CELIs). The JACs often lack the budgetary provision and infrastructure they require 
to function. One government employee in Putumayo told the evaluation team: 
For example, training has been provided to the JACs. [They] have played an important role in the 
activities carried out in the municipality. Indeed, the Unit has helped in the construction of these 
small community houses [buildings in which the JACs meet with their communities.] 
(Functionary, Putumayo, PNCRT) 
Conclusion 5: Despite the fact that these entities are occasionally considered to be co-opted by 
local political interests – which may threaten their legitimacy in certain places – they continue to 
represent the gateway to state services and programs. Where the security situation permitted, and 
where the JACs have the confidence of the population, this kind of social organization is central 
to political life and to participation in the veredas. In these cases the JACs play a central role in 
providing access to, and enjoyment of, the services offered by the State in terms of political 
participation, infrastructural improvements and social control. In some regions the JACs are 
weaker, explaining why surveyed households report reduced levels of participation overall. 
However, JAC participants report more active participation at the decision-making level. 
Conclusion 6: In the veredas of geographically distant municipalities, the JACs are the principal
spaces in which disputes between neighbors are resolved, mostly through direct negotiation. By 
having this local resource, minor social conflicts are resolved at the vereda level, without the
need to take cases to municipal or national bodies. This is the case in regions where the security 
situation allows it. However, participation can be affected in cases where threats are issued 
against leaders. 
Associations and productive projects
As described in the introduction, CELI priorities changed from a focus on consolidation of the 
GOC's presence to what could be characterized as more traditional rural development objectives, 
with a focus on agricultural production and the strengthening of rural value chains. This is seen 
most clearly in the reassignment of the CELIs from their original focus on consolidation under 
DO1, “Civilian government present in CSDI zones consolidated,” to its present articulation as 
DO3, “Improved conditions for inclusive rural economic growth." Throughout each project and 
especially since the revised Development Objective was released in June 2015, work with value 
chains has been a principal means for the projects to achieve their goals. 
In carrying out this work, the projects supported the producers associations in several ways: 
technical assistance and training; support work in developing marketing and commercialization 
(also often involving training); and assistance in financial services to support production and 
marketing. As the implementer of CELI Central put it, “the focus has increasingly been on 
87 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
empowering and mobilizing smallholder producer associations, and increasing private sector 
investment in rural communities to create economic opportunities for community stakeholders 
within the context of licit markets and the rule of law.”89 
CELI Central worked with 21 producer organizations in the AgroColombia activity, on issues of 
administrative and accounting, according to the CELI COR. This included supporting the rules 
behind rotating loan funds, creating a teamwork and leadership culture, financial education, and 
1,114 financial products granted from the rotating funds associations. They brought youth into 
the trainings to generate conditions for sustainability and facilitated the opening of a bank branch 
where previously there had been no financial services.90 
Considerable focus was placed on organizational training and capacity building, and TA on a 
range of crops, as described in the economic development chapter of this report. Training of 
producers associations covered a broad range of topics, including improving agricultural 
production, introducing new crops, strengthening the functioning of their organizations, and 
training youth to encourage the transfer of knowledge and resources to the next generation.
Study tours are a modality used by the project to widen the experience of producers associations, 
such as the visit organized by CELI Central for members of Asocaferumet in Meta. These coffee 
producers traveled to Quindío to learn about the productive chain for specialty coffees and met 
with a range of businesses working with small-scale coffee producers, as well as learning about 
the certification process.91 Associations in Quindío have since negotiated a contract with a coffee
exporter and will expand operations through the Acumen Fund (also supported by USAID).92
In the survey data, respondents reported an 8pp reduction in HH participation in producer 
associations in the CELI municipalities, whereas in the PNCRT municipalities no significant 
variation occurred. (Figure 4.14) 
Figure 4.14. Impacts on participation in producer associations, CELI and PNCRT
Variable CELI PNCRT
Household belongs to or participates in producer or marketing associations -0.0799*** 0.0128
The percentage of members who reported participating at the level of decision-making rose from 
around 30% to around 50% in both CELI and PNCRT municipalities. While positive, in the 
control municipalities the tendency was the same, so this may not be attributed to the 
interventions. 
Productive associations established for cultivating cacao (Putumayo, Nariño), rubber (Caquetá) 
or cattle-raising (Meta), were in zones with voluntary crop-substitution programs. Small 
 
89 CELI Central Quarterly Report Q1 FY2016, p. 5. 90 Data from conversation and e-mail with CELI Central COR, 30 June and 27 July, 2016. 91 Ibid, p. 44. 92 As reported by the USAID COR for CELI.
88 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
producers saw the associations as a way to leave illicit activities behind. An official in Putumayo
explained:
If you visit the 40 veredas that are free from illicit crops [you will see that] here in the 
municipality there have been some interesting processes of community participation and 
leadership. Because it’s been them and it’s been us: we get tired of all this stuff and we are going 
to rid our vereda of these crops that have done so much damage. But it is in that connection that 
the State is unable to help us. (Functionary, Putumayo, PNCRT) 
Survey results suggest there is an economic rationality behind respondents’ associativity. The 
active members who responded to the questionnaire said that the associations were useful to their 
productive projects for several fundamental reasons. The results indicate that positive impacts 
were felt in the CELI municipalities in the processing of products (4pp), sales (almost 5pp), 
transport (5pp) and training (5pp). For the PNCRT municipalities, more reported associating for 
product sales (almost 5pp) and transport (12.5pp). (Figure 4.15) 
Figure 4.15. Impacts on reasons they associate, CELI and PNCRT
Variable CELI PNCRT
Decided to associate in order to process their products 0.0431* 0.0223
Decided to associate in order to sell their products 0.0496* 0.0486**
Decided to associate in order to transport their products 0.0512* 0.125***
Decided to associate in order to engage in joint training 0.0494* 0.0224
The most important aspect of these figures becomes visible when they are disaggregated by sex, 
in particular in the CELI municipalities, where most of this impact comes from households with 
female heads. More FHHs reported joining in associations in 2015 to sell their products (8pp) 
while for MHHs there was no impact. More FHHs also reported working together in 2015 for 
processing (5pp) and transport (10pp). The only measure in which a positive impact was 
apparent for MHHs was in association in order to engage in joint training (3pp), and FHHs had 
no impact on this measure. In PNCRT municipalities, more FHHs reported associating in 2015 to 
sell products (9pp) and to transport products (17pp).
In absolute terms, HHs in CELI and PNCRT municipalities reported more frequently that 
participation in associations had enabled the households to improve their productive activity
(increasing from 48% to 75% for CELI, and from 30% to 67% for the PNCRT). In the controls, 
the effect was even greater, showing that it is generalized and cannot be attributed to an 
intervention. 
While it is true that people enter the JACs as a result of natural processes of communication and 
the resolution of daily problems that occur in the communities, the emergence of producer 
associations was also encouraged by state and international cooperation projects and programs. 
These entities usually demand that the organizations they support are constituted as associations 
in order to ensure that the benefits reach the population. 
At that point, the European Union turned up and gave us the idea that if we didn’t form an 
organization, we wouldn’t be able to get any money, because the government was only going to 
invest in associations. So it came along and founded five associations. (Producer, Meta, CELI)
We produce things now; we have the crops to produce stuff. Where did our vision come from? 
89 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
When the FUPAD [NGO, Fundación Panamericana para el Desarrollo]came to our vereda and 
called a meeting. (Producer, Antioquia, CELI) 
Yes, I’m a member of an association in the Chicalá vereda called Aprochicalá. We began to set it 
up a couple of years ago. (…) You can’t receive any help or anything these days if you don’t 
belong to an association. So, we’ve had to take part in an association. (Producer association 
leader, Tolima, CELI)
As a result, the communities respond by increasing the number of associations and, in some 
cases, paying little attention to their durability or sustainability. Once benefits have been 
delivered, the members sometimes lose interest in continuing with the process unless programs 
and projects are run again in their communities.93 Consequently, it is no surprise that respondents 
recognized the existence of paper associations, as in these examples: 
Of all the associations in Vista Hermosa there are many that are just paper associations, nothing 
more. That is, they have only been registered in order to capture resources and [afterwards] they 
remain quiet. (Producer association leader, Meta, CELI) 
In ‘98 I didn’t know of any other associations. But year after year more and more associations 
were created and now there are something like 72 or 76, something like that. Every time someone 
from the Ministry or some government body comes to give us training they always insist that if we 
don’t form an association we won’t receive any resources. That’s why the people have gone crazy 
forming associations – there are associations with 10 members, 18, 20… (Producer, Meta, CELI)
This does not mean that the CELI or the UACT encouraged the creation of paper associations 
more than other interventions did, simply that an incentive exists that affects all rural 
interventions. In certain cases, the current interventions are constructed on the basis of earlier 
associations. An example drawn from CELI N/S provides a good illustration:
Colombia Responde turned up and carried on with the projects initiated by FUPAD. Colombia 
Responde enabled the rubber-growers association to keep going and continued to equip and 
maintain the plantations; and now we’re being helped by Colombia Responde. (Producer, 
Antioquia, CELI) 
Conclusion 7: As a result of the Colombian government’s interventions and of international
cooperation programs, the number of productive associations is increasing. Households 
recognize that the associations have served to improve their productive activities. Among the 
most important incentives for forming associations are processing, transport, sales, and joint 
training. These results have been more significant in the CELI interventions and above all among 
FHHs. People who participated in associations at the time the baseline was constructed continue 
to take part in associations but are now also involved in decision-making at higher levels than in 
control municipalities. 
Conclusion 8: At the same time, overall household participation in these associations has 
declined. This contradiction might strengthen the hypothesis that households seek out benefits 
 
93 Associativity has its own logic. It is related to the number of beneficiaries, the fewer the better, and to the system of internal 
rules – the more stringent the better – that minimize the effects of opportunism (the free rider phenomenon) (Olson, 1965).
90 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
and that once these have been achieved, participants lose interest in the organizations. 
Opportunistic interests (the free rider phenomenon) may prevail over participation for conviction 
or long-term vision. 
Women’s and victims’ associations 
Other forms of association supported by CELI included groups of victims of the armed conflict, 
facilitated by the provisions of Law 1448. Women founded many of these associations and take 
on community leadership roles, resulting in their empowerment within the community. 
Participants emphasize the increased levels of cooperation and autonomy that they have
achieved, while new forms of female leadership have emerged, based on increased participation 
in public affairs and the search for recognition and dignity for the victims of the conflict. 
Similarly, new organizations have emerged that did not previously exist in the region, such as 
associations of young people or of Afro-Colombians. According to the qualitative fieldwork, the 
support of the Department for Social Prosperity (DPS) and the Victims’ Integral Attention and 
Reparation Unit (UARIV) have been fundamental to this process.
The participation of women in public affairs and in productive activities has led to a significant 
cultural change compared to a generation ago. In addition to the CELIs, other programs such as 
Familias en Acción and the DPS’s Women Savers have also played a fundamental role. The 
words of a female community leader from a municipality in southern Tolima illustrate the 
increased participation of women: 
The women’s organization has existed for about five years now… I had very few [in my group] at 
the beginning, very few, but now I have something like 300 women working side by side with me. 
(Producer association and CSO leader, Tolima, PNCRT) 
In some cases, the productive associations are made up principally of women who are victims of 
the conflict and/or living in situations of displacement. These associations have enabled the
women to generate an income, create links of solidarity with others and build roots in the 
territories that have received these people and families. This was the situation in San Jacinto, 
Montelíbano, and Santander de Quilichao (CELI). A woman from Montes de María told us: 
[I] am an artisan and I belong to the Luz y Vida group of artisans… I make mochilas, I do a bit of 
everything: I work, I iron, I wash clothes and that lets me support my family. I am not from here, 
and I thank God that this place has welcomed me and that things have gone really well – I have 
friends and we all meet together and when there is some activity or a celebration we take part 
and we are all happy. We’re fine, thanks be to God, in spite of everything we’ve suffered… We 
keep on struggling and everything is turning out right. (Producer, Bolívar, CELI)
The survey data also shows that victims’ associations are growing. In CELI municipalities, the 
number of people who reported belonging to organizations of this kind rose from 7% to 12%, 
while the corresponding growth for the PNCRT was from 6% to 9%. The quality of participation 
also improved, thanks to increased participation in decision-making. However, these changes are 
also present in the control municipalities, and were not, therefore, attributable to the 
interventions. On the other hand, in PNCRT municipalities, women’s group participation and 
decision-making fell (Figure 4.16); but since the control areas decreased even more, the PNCRT 
municipalities saw a slight positive impact (3.6pp) by comparison. 
91 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 4.16. Impacts on participation in victims’ and women’s groups, CELI and PNCRT 
Variable CELI PNCRT
Participation in groups of displaced, returned or vulnerable populations 0.0107 0.00657
Decision-making in groups of displaced, returned or vulnerable populations -0.116 -0.0654
Participation in women’s groups 0.0356* 0.0117
Decision-making in women’s groups 0.0610 -0.260*
In contrast to the quantitative findings, which showed no impact for belonging to or participating 
in victims’ organizations, the qualitative responses indicated their importance. Organizations for 
displaced, returned or vulnerable people were said to have transformed into organizations of 
another kind that have expanded their identities and their sense of what they should seek to 
achieve, without losing sight of the reasons they were originally founded. Frequently, in order to 
generate income for their members and for the organizations, victims’ associations established
complementary productive projects. 
We know about victims’ organizations because many different organizations take part in the 
regional round table and they are interested in expanding into agricultural production. As a 
victim’s organization we have always focused on farming, while others focus on commercial 
activities. (CSO member and producer, Putumayo, PNCRT)
In Santuario we have Las Líderes who are women who work with shampoo and air-fresheners. 
There’s also Asovides, which is for displaced people - and they’ve created two more 
organizations for them in Santuario too. And then we have an association of goat producers, and 
there’s another of those in Montaña. (Producer association leader, Caquetá, CELI) 
The members of our foundation are internally displaced. (Producer, Tolima, Control) 
With these changes, victims’ organizations have been able to attract new funding for increasing 
members’ sense of belonging and improving household incomes.
Conclusion 9: The social capital that has been constructed as a result of the work of victims’ 
associations has permitted the creation and strengthening of other kinds of organizations in a 
process that has encouraged an expansion of their reasons for existing and their aims, fomenting 
a sense of belonging and identifying new resources. 
Conclusion 10: Overall, participation in victims’ associations and organizations of vulnerable 
groups increased, and the CELIs have had a small but significant positive impact on participation 
in women’s groups. Though the change cannot be attributed to the interventions, they represent 
an important change in the target population. 
Cultural and other associations 
In some of the zones in which the fieldwork was carried out, and in particular in areas with large 
Afro-Colombian populations such as Montes de María and Tumaco, the cultural processes 
created by civil society have developed and strengthened the population’s sense of rootedness in 
their territories. These associations help to reconstruct relations of trust, the social fabric and the 
identity of the communities. In Montes de María, the importance of cultural processes for 
strengthening social capital and ties with the territory was apparent: 
92 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Now what are we doing? We have a dance project called Danza Negra, with gaitas [traditional 
wind instruments made of cane], and we are beginning to rescue our traditions through 
drumming; and we also created a Danza Negra group in San Isidro. (Producer, Bolívar, CELI) 
For their part, in Tumaco and Antioquia, women were found to be leading cultural activities 
using music, leading to an increased sense of identity among the population. 
The organization is very close to [women’s group] Asomuca and it is the role of women in this 
tradition to preserve the Tuna Tambora [traditional singing and drum groups]. (Producer 
association and CSO member, Antioquia, CELI)
18 women and five men. There we see the role of women… in the field of dance. So the tradition 
is not lost; because the Tuna is ancestral, it’s existed since the people first arrived in Cáceres… It 
has been women who have preserved the tradition. (Producer, Antioquia, CELI) 
Another way of strengthening the social fabric has been through the support provided to 
associations that preserve and reproduce the intangible cultural heritage of their communities. 
The CELIs have provided direct support to different cultural associations, enabling the 
communal values destroyed by the conflict to be reconstructed. In the CELI and the PNCRT
municipalities alike, participation in this kind of association increased by three points compared 
to the controls. However, this participation did not include decision-making (in contrast to what 
was observed in other kinds of organizations). 
In focus group discussions, participants suggested that a return to activities like dance, music or 
traditional artisan production has permitted communities to renew their cultural identity and to 
feel pride in their traditions, but that these organizations also operate as forms of resistance 
against armed insurgents. In regions where Afro-Colombian communities predominate, such as
San Jacinto or Tumaco (both CELI municipalities), the program has supported cultural 
associations that had disappeared as a result of the conflict. The associations help to reconstruct 
identity and ties of trust, and involve youth in processes dedicated to preserving cultural legacy. 
A cultural leader from Montes de María explains: 
We created a project called the Son de Negros. We focused on… the recruitment carried out by 
the illegal armed groups and we are going to include this in our Danza Negra. Danza Negra is 
African dance that we have been dancing since the time of the palenque94 of María la Baja. There 
is a sense of connection with the mountains, with my grandparents; I remember that my 
grandfather was black and was a drummer in the Danza Negra. (Producer association and CSO 
member, Bolívar, CELI)
Cultural associations are fundamental to the Afro-Colombian communities as a conduit for self￾awareness and to feel pride in their history and their traditions. A leader from Tumaco comments 
in this regard: 
The organization Tumbo de la Mar was born several years ago in the minds of a group of friends. 
It was born of the necessity to conserve, preserve and share matters connected with our cultural 
 
94 Palenques are communities of free blacks created in resistance to slavery.
93 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
roots. In about 2009 or 2010 … this desire to develop cultural activities emerged in a meeting,
and from that point on the work has continued… with young people from different parts of 
Tumaco working to preserve and teach our music and traditional dance. Because this is our 
identity, it is what we are here, here in this part of the southern Pacific coast. (CSO member, 
Tumaco, CELI) 
The cultural associations have been successful in preserving tangible cultural heritage by tapping 
into the knowledge of the elderly and recovering forgotten traditions and practices.
We have heard of women’s groups, but they are not exactly cultural but groups of artisans; the 
artisans are very united; they are groups of old people – I think the youngest is about 65. The 
group meets here in the vereda. The artisans make handicrafts, and before this, traditional 
instruments. (CSO leader, Tumaco, CELI) 
The qualitative analysis showed that CELI support to cultural associations has impacted young 
people and their participation in community activities, and has made it possible to preserve the 
cultural heritage of the community and to transmit it to new generations. Cultural activities have 
also permitted the creation of relations of trust, preserved memories and, above all, helped 
communities to value their collective identities. 
Conclusion 11: The CELIs have supported different cultural associations, enabling the 
reconstruction of community values damaged by the armed conflict. Returning to activities such 
as dance, music or traditional artisan production has permitted communities to create a cultural 
identity in which they are able to recognize themselves and feel pride in their traditions; it has 
also acted as a form of resistance against armed actors. In regions with a large Afro-Colombian 
population, such as Tumaco and Montes de María, cultural organizations have been 
strengthened, enabling the emergence of social ties in the regions where the interventions have 
taken place. The high levels of social capital that have been constructed in the regions have been 
added to the equation, as a result of the presence of victims’ associations which establish the 
trust on which the participation of social organizations is based. 
Conclusion 12: The support provided to associations that preserve and reproduce the intangible 
cultural heritage of a community is enormously important because it produces transformations in 
the communities by assigning value to collective identity. Cultural associations have helped to 
inspire processes of symbolic renewal in communities of victims of the conflict. 
Recommendations 
1. Adapt programs to encourage trust (conclusion 2) The unprecedented overall decline in trust 
of the government cannot be resolved by development alone, but it is important to develop ways 
to strengthen community ties to mayors’ offices and local government bodies. Despite 
international development interventions, the population feels less trust toward local government 
than they did two years ago, and less trust between neighbors. These indicators are highly 
sensitive to factors outside the control of any program, but interventions can and should develop 
tools to confront the problem. 
2. Improve work with female heads of household using participatory models (conclusion 4) In 
terms of participation, the positive results obtained for MHHs need to be extended to FHHs. 
Efforts should be made to benefit this population that has traditionally been immersed in 
unfavorable socioeconomic conditions and has had less free time available to dedicate to 
associations. FHHs who already participate in victims’, women’s and other groups could serve as 
94 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
guides for other women to become involved and to increase their participation. Similarly, the 
community leaders of associations and other groups, whose commitment and participation 
represent higher levels of social embeddedness, could serve as a resource to advance this aim. 
The CELIs should continue investigating different ways of using this resource of association
leadership and practice in the final months of the intervention, engaging in outreach activities 
with community members whose behaviors indicate increasing isolation. 
3. Support the legitimacy of the JACs (conclusions 1, 5 and 6) In order to strengthen the JACs, 
the CELIs should continue training leaders in board election processes and community 
participation tools to counteract the risk of losing legitimacy. Support of the JACs to encourage 
wider participation in local affairs – including women’s participation, that of victims, and other 
marginalized groups – should be a priority so as to generate horizontal communication between 
neighbors and vertical relations between communities and local authorities.
Programs should monitor the quality of community associations such as the JACs, in ways 
similar to how producers’ organizations are monitored. The ICO should be adapted to meet 
organizational and community expectations for these groups and to ensure that programs involve 
the JACs, communities and local government. In cooperation with mayors’ offices and officials 
responsible for community or administrative affairs (the secretarios de gobierno), work should 
continue to establish special programs focused on the JACs to improve their management 
capacities and their representativeness. Monitoring results should be made public to generate
community incentives to play an active role in holding JACs to account. 
The component competencies of the ICO should be evaluated to determine which aspects work 
well and which do not. Given the differences of opinion that exist concerning its usefulness, the 
Chatham House Rule95 could be used for such a discussion, in which parties express their 
opinions openly and identify weaknesses or gaps in the measuring tools without fear that their 
views will become public. 
4. Strengthen the role of the JACs in the resolution of conflicts (conclusion 6) The CELIs 
should support the strengthening of the tools currently available to the JACs in their conflict 
resolution role with training for leaders in the veredas. Other interventions, local and regional 
personerías, the police and the Defensoría or even groups such as the Chambers of Commerce 
could also be involved. 
5. Make associations more transparent (conclusions 7 and 8) To improve the quality of the
associations, use a refined ICO as a monitoring and ranking system for associations. Make 
results public and give greater access to programs and benefits to the highest ranking 
associations, with local authorities’ involvement. 
 
95 When a meeting, or part thereof, is held under the Chatham House Rule, participants are free to use the information received, 
but neither the identity nor the affiliation of the speaker(s), nor that of any other participant, may be revealed. 
https://www.chathamhouse.org/about/chatham-house-rule.
95 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
6. Loosen the requirement that beneficiaries must be grouped into associations (conclusions 7 
and 8) In future programs, the possibilities that intervention might not rely strictly on the 
creation of associations should be examined. It is not always necessary to require that potential 
beneficiaries form an association. If an association is to be authentic, it should emerge from the 
initiative of individuals who decide whether they wish to take part or not and who wish to gain 
individual advantages (for example, improved sales for their products). This represents a contrast 
with the JACs, whose objectives are based on public goods and proposals (such as parks, 
bridges, child-care facilities, road improvements, etc.)
Similarly, programs such as CELI should support private initiatives. Small businesspersons
improving their businesses may be a ripe target for support, as was sometimes the case under the 
CELIs. There are also times when the obligatory formation of new associations is not appropriate 
for cultural, social or historical reasons. 
7. Increase collaboration with associations whose objectives may or may not be at root 
productive (conclusions 9-12) Social organizations are also candidates for CELI-style 
interventions because of their high levels of social capital, and they are increasing in number in 
the intervention zones. Such organizations (that are not necessarily producer associations) should 
be supported to become autonomous and financially sustainable. Such activities could be 
developed in cooperation with the DPS, the Ministry of Agriculture, SENA and the Municipal 
Agricultural Extension Units (UMATAs). 
96 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
GOVERNANCE AND 
INSTITUTIONS
Evaluation questions: To what extent have CELI interventions contributed to 
participation, governance and institutionalization of the territory?
What are the CELIs' specific effects of in changing the capacity and service delivery of 
local government? (To serve local people, be accountable, expand social services.)
SUMMARY: Citizen participation at the ballot box remains relatively high, but electoral 
challenges like vote-buying persist. An accountability index from the survey questions 
shows declines since baseline across the regions under study. Citizen oversight is 
minimal, but CELI work with the personerías (in at least one region) to strengthen the 
registry of veedurías has been positive. Service delivery remains a major challenge, with 
citizen perceptions of access and quality very low as well as lower than at baseline. 
Reports of human rights violations by the military decreased but reports of police 
violations increased in PNCRT municipalities. This may reflect a change in force 
presence in consolidation municipalities, but does not reflect an improvement in citizens’ 
interactions with public forces.
CELI interventions were designed to strengthen local policymaker capacity and to improve 
management and provision of public goods and services. Only through consistent and fair service 
delivery will citizens recognize the permanence and legitimacy of the State in areas where IAG
influence has been strong and, in many cases, parastatal. 
CELI interventions varied according to the region, type of municipality and its initial conditions, 
the quantity and quality of social organizations, and perceived priorities. In general, the CELIs 
sought to strengthen local governments in accountability processes, formulation and execution of 
municipal plans including rural development, and guaranteeing victims’ rights. In parallel, the 
CELIs sought to strengthen civil society to work in articulation with municipal governments, 
defend their rights and exert control in government actions. 
This chapter has two main sections. The first examines governance from the citizen perspective: 
political participation, electoral practices and conditions; transparency, oversight and 
accountability; and trust in government. This section uses the following sources of information: 
1. Participation index composed of survey questions around electoral and political 
participation.
2. The comparison of survey results between baseline and midline. 
3. Qualitative findings from interviews with municipal representatives, personerías and social 
organization leaders, as well as focus groups with beneficiaries. 
4. Related secondary sources.
The second section concerns service delivery and the capacity of local governments to deliver 
services well. Sources for this section include: 
1. The Integral Performance Index, a GOC tool that rates governments on a range of 
competencies. 
97 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
2. The comparison of survey results between baseline and midline, including an index of 
items on government capacity and service delivery from the citizen’s perspective. 
3. Qualitative interviews with personerías, organization leaders and beneficiary focus groups. 
4. Related secondary sources, including the Consolidation Index good government sub-index. 
The evaluation team then makes a set of recommendations for USAID and the CELIs’ ongoing 
work, with the hope of increasing participation and making interactions with the GOC more 
effective at promoting decentralization and improving public service delivery.
Findings and conclusions
Participation, transparency, accountability and trust
Political participation and electoral culture: the Participation Index
On a scale of 1 to 100, the Participation Index sums variables such as participation in political 
parties, JAC and oversight groups or veedurías; the perception of participation in elections; voter 
registration; and whether the respondent voted in the last presidential election. In CELI 
municipalities, the average score is 36/100 while control sites show 34/100. The CELI 
municipalities show a positive impact equivalent to 1.8 points over the control group (Figure 5.1) 
because of the control group’s decline, while the CELI score remained stable. PNCRT 
municipalities were stable at about 35 points out of 100 at baseline and midline.
Figure 5.1. Impacts on the Participation Index, CELI and PNCRT
Source: Authors’ calculations. *** Significant at 99% confidence ** 95% * 90%96
Electoral practices
In absolute terms, around 90% of respondents across the full sample report voting in the last 
presidential elections. This is an increase of about 10% from the average reported at baseline. 
Since the effect was the same across samples, no impact is found. 
Around 95% of voters said that they voted because their vote counts, or because it was their 
responsibility as citizens. Across the full sample, about 85% say they voted because they were 
drawn to a particular candidate. Participation in political organizations increased for CELI and 
PNCRT by around 6%. Since this change was also seen in control areas, no impact is found. 
 
96 Throughout, the data source is the evaluation’s survey data, unless otherwise noted. Standard significance notations apply: *** 
for 99% confidence, ** for 95%, * for 90%. Impact estimation models are in Annex 3, and data in the text are tabled in Annex 10.
Treatment score (/100) Control score (/100)
Baseline Midline Baseline Midline Impact
CELI 36.14 36.11 35.82 33.36 1.82**
PNCRT 35.85 35.02 35.58 34.23 -0.63
98 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Among citizens who did not vote, the reasons for not doing so are similar for PNCRT and CELI 
zones. The main reason people gave for not voting was that no candidate was satisfactory, but also 
cited were the difficulty in voting and a lack of interest in the electoral process. Survey data show 
no impact from either CELI or PNCRT interventions on these factors. 
Respondents were asked whether they or someone they knew had been pressured or forced to vote. 
Only 3% of CELI respondents said yes, slightly lower than at baseline, without significant impact. 
When asked if they or someone they knew had exchanged their vote for money or gifts, 21% said 
yes at baseline and 16% at midline. Controls declined even more, resulting in a 3pp negative 
impact for CELI sites. No impacts are recorded in such practices for PNCRT.
Conclusion 1: There is good voter turnout, but clientelistic practices persist. High electoral 
participation shows that voting has become institutionalized: citizens do vote, regardless of 
interventions. The rates show that citizens in municipalities affected by the conflict think that the 
two most recent presidential elections were important – elections that have focused the debate on 
the end of the conflict. There is a high level of acceptance of formal democratic channels. 
It is concerning, however, that respondents admit having taken (or knowing someone who has 
taken) money or gifts in exchange for their vote, in CELI, PNCRT and control regions. This 
suggests a not uncommon practice in municipalities in the process of institutionalization, where the 
investment of resources implies, in the vast majority of times, clientelist uses. Clientelism can be 
explained because there is discontent with the political system for not being efficient in ensuring 
the satisfaction of citizen needs and service delivery. This leads to the citizens perceiving the 
elections as their only opportunity to extract benefits from the State, and in turn the citizens 
demand something in return for their vote. 
Citizen oversight, accountability and trust in public management
Overall, citizen participation in oversight processes has been low, both at baseline and in the 
midline. For CELI municipalities the participation rate is 2.2% in contrast with their control 
municipalities (1.6%). However, participation in oversight committees or veedurías in CELI 
municipalities increased by 1pp compared to the control group, between baseline and midline. 
In PNCRT municipalities, participation in oversight committees at midline is 2%, in contrast with 
the control group whose proportion is 1%. However, there is no evidence of impact, since 
participation at baseline was also higher in PNCRT municipalities than in their controls. 
The survey instrument contained questions that were combined into an accountability index equal 
to 100 points, which is shown in Figure 5.2. The survey asked to what degree the mayor kept 
citizens informed about public expenditures, the presence or absence of public spaces for citizen 
oversight, and the degree to which local government took citizens’ opinions into account in 
decision-making.
99 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 5.2. CELI Performance: Accountability index, total and regional 
Ind. Description Region Baseline Midline % Change Source
DO3-
028
Level of 
accountability 
in CELI 
municipalities
Total 23.8 19.63 -18% Household survey
• Baseline 2013, 
Econometría S.A. 
• Midline 2015, 
Proyecto EVAL
Montes de María 18.6 17.74 -5%
Central 27.8 24.96 -10%
Norte 25.5 18.89 -26%
Sur 22.6 16.17 -28%
The changes were uniformly negative for CELI municipalities in how the public rates the 
accountability of municipal governance, with a drop in score between 5% and 28%. 
Overall, the good government sub-index of the Consolidation Index shows improvement since 
2011, with an average municipal score of 54.7 increasing to 62.3 over the three years since data 
began to be collected. This sub-index is measured with four indicators: the Open Government 
Index of the GOC, percentage of funding from royalties, and effective tax payments (personal and 
industry/commerce.) The positive trend to this sub-index is compared to similar, under-performing 
municipalities, rather than in absolute terms, but at least does show improvements. 
Qualitatively, CELI support for government has been important, among other things, to strengthen 
and streamline veeduría registration, as community leaders and personerías report. 
Here they’ve really increased the registration of veedurías. Before, a project was done ... and no 
one knew anything; and then someone would come and criticize. Now people are involved and they 
build oversight and monitoring process into the project. So it has also strengthened the 
institutionalization a great deal. (CSO leader and producer association member, Tolima, CELI) 
CELI Central reports working with producers’ organizations and JACs to participate in public 
spaces to lobby municipal administrations, such as in the Sur del Tolima, or formulated projects to 
be included in the Project Bank.97 
Issues of accountability are important to evaluation respondents, and in general their perceptions of 
government accountability have worsened at midline. In PNCRT, about 6pp more households 
consider that one of the most serious problems associated with the municipal administration is that 
it does not report what it is doing (Figure 5.3). In CELI municipalities, negative responses at 
baseline regarding public service deficiency dropped by 11pp at midline. All other variables show 
no impact in CELI, PNCRT or controls. 
 
97 Data from conversation and e-mail with CELI Central COR, 30 June and 27 July 2016.
100 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 5.3. Impacts on perceptions of problems with municipal administration, CELI and PNCRT
Variable CELI PNCRT
They do not report what is being done 0.0047 0.0568*
They do not pay attention to requests from the community -0.0217 0.0705
They do not have the capacity to resolve requests from the community 0.0337 -0.0506
Mistrust of government entities 0.0231 -0.036
Deficiency in public services (water, electricity, sewer, etc.) -0.114*** -0.114
The lack of highways 0.0120 0.0709
Corruption -0.0106 -0.0347
The lack of institutional goods: hospitals, schools, courts, child care -0.00292 -0.0711
Around half of the treated sample (49% in CELI municipalities and 53% in PNCRT) think that one 
of the most serious problems of the municipal administration is that they do not pay attention to 
community needs. This problem is evident in focus groups and interviews regarding meetings with 
the municipality: 
I am the JAC president and I got tired; I really do not come to meetings anymore. I chose not to go 
back. They don’t solve anything. They have not given anything to the village. (JAC leader, Meta, 
CELI) 
Finally, the survey also shows that around 7% to 8% of CELI and PNCRT households reported 
that the mayor informs the community on how resources have been spent, with no impact since 
baseline. The lack of contact between the citizens and the mayor was also evident in the focus 
groups: 
We need more connection with the Mayor and the Council. It also depends on us, as a community, 
to go break the ice and make that relationship. (CSO leader, Antioquia, CELI) 
Conclusion 2: Accountability is still an emerging practice. Responses to the survey questions 
about accountability were low and the index declined from baseline to midline. Lack of 
accountability has a negative effect on citizens’ participation and effective control. Local 
governments tend not to create spaces for accountability and do not invite citizens. While Law 489 
of 1998 established compulsory accountability processes, the public receives little benefit from 
these processes. From the qualitative data, it is clear that the CELIs contributed to strengthening 
accountability (and therefore good governance), but the evidence does not show that it is widely 
practiced.
Conclusion 3: Participatory mechanisms are not yet widely effective. While CELI has intervened 
to ensure that communities’ stated needs are included in development plans at the municipal level, 
people who claim to have attended the meetings convened by the municipal institutions state that 
these are not efficient. In general these meetings do not address the needs or complaints expressed 
by community leaders or do not implement plans they formulate together. This creates expectations 
and negatively affects citizens’ perceptions of their governments. This deepens a vicious circle, in 
which ineffective participation erodes perceptions about government management and 
transparency, which in turn discourages participation, further weakening accountability. The 
Consolidation Index indicates higher levels of oversight and citizen participation than is detectable 
101 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
in the survey data, likely in part for the very rural character of the sample – but this will need to be 
examined in the future to find ways to bring a wider range of citizens into the process, not only 
those in municipal or other capitals.
The lack of accountability and citizen participation in the formulation, implementation and 
monitoring of public policy is also reflected in the relatively low ratings citizens give to 
government management, at both the local and national levels. There is negative impact in both 
CELI and PNCRT municipalities in perceptions of mayoral and national level management. 
(Figure 5.4) 
Figure 5.4. Impacts on citizens’ perceptions of public management, CELI and PNCRT
Variable CELI PNCRT
Rates the mayor’s management as good or very good -0.0187 -0.0653*
Rates the department’s management as good or very good -0.0191 -0.00223
Rates the national government’s management as good or very good 0.00249 -0.131**
Trust the mayor’s promise to carry out a project -0.022 -0.0466
Trust the JAC president’s promise to carry out a project 0.0277 -0.0339
Among CELI households at midline, in absolute terms, only 15% rate the mayor’s management as 
good or very good, while at baseline this figure was 21%. 
Similarly, for PNCRT municipalities, there was a negative impact of 6.5pp in trusting the mayor’s 
office to carry out a project it has promised to do. Perceptions of the national government have also 
worsened in PNCRT municipalities, with 13pp negative impact. These data show clearly that 
opinions of management have deteriorated. 
Conclusion 4: In CELI and PNCRT municipalities, impacts since baseline are negative regarding 
government management at all levels. They may be ineffective at implementing new resources, 
may be creating new and unmet expectations, and may simply not have changed practices in 
response to interventions. But the effect on citizens is significant. Where there have been 
successful interventions on this front, there may be localized results – but the generalized impact 
the evaluation was designed to detect is not present. 
The crisis of confidence in Colombian local institutions has worsened in recent years. The HH 
survey reveals decreased confidence in all levels of government. Trust in government dropped 
from baseline to midline, across the full sample. The decline in confidence is parallel across 
governmental institutions: police, army, navy, justice; national, departmental, local, city council 
and JACs; ombudsperson offices; and advocacy and support bodies such as the national training 
institutes.
Just one rated institution shows positive impact: the personería, and only in CELI sites. In absolute 
terms, opinions of the personería fell in both CELI municipalities (from 28% to 22%) and in 
control sites (from 28% to 17%), but more so in the latter. This results in a 4.6pp positive impact in 
CELI municipalities: that is, had the CELI not intervened, the municipalities would have seen 
similar declines. In the Central region, the positive impacts are particularly significant. As this 
102 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
effect is not seen in PNCRT municipalities, the increase itself is likely linked to CELI 
interventions. 
Qualitative data from CELI areas suggest that the personerías do work to meet the requirements of 
national policies and legislation for victims; this may explain the higher percentages of confidence 
in institutions directly involved with these processes, as is the personería. Citizens in CELI and 
control areas recognize the personería as the main institution with which communities interact, 
particularly because of its role in the Victims’ Registry and in complaints about human rights
violations. Personería officials interviewed for the evaluation also recognize the support of the 
CELIs with personnel and technological resources: 
We have worked directly with Colombia Responde. They have focused USAID resources in 
the municipality with the personería. They’ve institutionally strengthened the personería in 
human resources, in relation to victims, and prevention training. We’re able to cover more 
people and more institutional offers despite, as everyone knows, not having many 
resources. (Functionary, Caquetá, CELI) 
This quote, from the CELI Central region, is illustrative of the particular effort put forth by CELI 
Central interventions with personerías. Government functionaries told the evaluation team that the 
CELIs have contributed to better municipal planning, in consensus with communities, on specific 
issues like victims, taxes and royalties. They also report the added value of local government 
training and support to communities in developing projects to access resources.98 
Conclusion 5: Trust in institutions depends on their ability to meet citizen needs, and that has 
worsened overall. Despite this general decline, improved perceptions of personerías shows a 
positive impact of CELI policies, and qualitative sources show that government actors 
acknowledge CELI contributions to capacity building. They are also registering and monitoring 
veedurías, as noted in the section on oversight above. However, circumstances have worsened the 
citizens’ overall trust of government, creating an even worse environment than when the CELIs 
began. 
Service delivery
The Integral Performance Index
The GOC created the Integral Performance Index – IDI for its Spanish initials – to evaluate the 
management of territorial governments, and has applied it since 2006. It tracks the performance 
of mayors and governorships in six dimensions: efficacy, efficiency, legal requirements, 
administrative capacity, fiscal performance and management.99 
 
98 The contrast between quantitative estimates and the qualitative reports may be explained, at least in part, by the difference in 
samples: as a random exercise, the quantitative research shows the population average, while qualitative research reached out to 
active community members, who are not only easier to reach but also more likely to have experience with the interventions. 99 The DNP explains the results, the methodology and the sources of information in the documents in this website: 
https://colaboracion.dnp.gov.co/CDT/Desarrollo%20Territorial/Docuemnto%20Desempe%C3%B1o%20Integral%202013.pdf.
103 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
In this chapter we analyze the evolution of the IDI between 2012 and 2014. For this purpose, we 
grouped and classified IDI data into the four groups of municipalities under study: CELI, 
PNCRT and their respective control groups. (Figure 5.5) 
Figure 5.5. Secondary data: Integral Performance Index for the four groups of municipalities
Source: Data from Integral Performance Index of the National Planning Department
IDI scores for all groups are “good”100 in both PNCRT and CELI municipalities at midline, up 
from “low” at baseline. CELI, PNCRT and control municipalities registered increases between 
2012 and 2014. Regarding the performance of the index itself, changes from year to year can be 
large, and are often explained by the existence of strong or weak professionals, such as health 
and education secretaries, or companies providing potable water and sanitation services. 
Breaking it down by component, the IDI index’s significant differences appear mainly in three 
dimensions: efficacy, legal requirements and management. (Figure 5.6) 
 
100 The “good” rating for municipalities is between 60 and 69 points and the “low” rating is between 40 and 59 points.
58
60
60
57
65
67
68
74
0 10 20 30 40 50 60 70 80
CELI
Control
PNCRT
Control
2012
2014
104 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 5.6. Secondary data: Integral Performance Index, 2012 (left) and 2014 (right) 
Efficacy
The efficacy component analyzes the implementation of the development plan and the 
fulfillment of its goals. In this component, between 2012 and 2014, there is a noticeable 
improvement in the rating (between 20pp and 40pp) for analyzed municipalities. In 2014, part of 
the strong performance was that more administrations reported about their compliance with the 
goals of the development plan. (Only 23 out of 80 reported in 2012, but two years later, 77 of 80 
did so.) While there is no way to show that the CELIs are behind this increase in reporting, they 
did emphasize their support in creating plans for municipal development. 
Legal requirements
The rating of the legal requirements component improved 4pp for CELI and 7pp for control 
municipalities. This improvement means that the administrations improved in the management of 
resources of the General Participation System in accordance with current budgetary regulations. 
In the PNCRT municipalities this score was stable.
Management
This component assesses municipal capacity to manage resources and turn resources into goods 
and services for the population. This indicator measures the tools that the municipality has for 
strategic planning, project monitoring and accountability. The management component score 
improves in the four groups of municipalities. However, the improvement is much higher in the 
CELI municipalities than in the control group, around 8pp. The scores of the management 
component are the highest of the six components of the IDI. 
Other components
The remaining components of the IDI were relatively stable in the studied municipalities. Fiscal 
performance measures the collection of taxes, debt management and compliance with Law 617 
on public finance sustainability. Fiscal performance scores were the most stable, improving by 
1pp to 3pp between 2012 and 2014. Efficiency evaluates results obtained with General 
Participation System, or Sistema General de Participación (SGP) resources in health, education, 
105 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
water and sanitation; this component was also very stable. Administrative capacity measures
staff stability, professionalization, the availability of computers, process automation and 
implementation of models and control standards. In this component, scores remained stable and 
satisfactory.
Conclusion 6: The integral performance of municipalities improves across the full sample, with 
important gains in components of management, effectiveness and legal requirements. This trend 
is common to the four groups, so it is not possible to attribute progress to any interventions. 
However, it is reasonable to mention that both CELI and PNCRT interventions included direct 
and explicit support in the elaboration and implementation of municipal development plans and 
regional action plans (planes de acción regional, or PARs). Therefore, it is not unreasonable to 
think that these interventions may have helped the progress of IDI, especially if one takes into 
account that the management component comprises most of the progress in the index. 
At the same time, this measure is internal to the GOC. As a contrast, it is useful also to examine 
the Governmental capacity index from the survey data, in the following section, to balance the 
government’s own view of its services with the experiences and opinions of service users. From 
that perspective, much work remains to be done. 
Governmental capacity index
This index measures government capacity and service delivery from the perspective of citizens, 
rating access and quality of services, quality of public policy formulation, and perceptions of 
municipal actors’ honesty. The results found for both CELI and PNCRT municipalities show no 
impacts generally nor by sex or region, and scores declined from 5% to 17% in absolute terms in 
the CELI municipalities (Figure 5.7). Among the components, the deterioration of scores is worst 
in respondents' rating of access to services.
Figure 5.7. CELI Performance: Governance capacity index, total and regional 
Ind. Description Region Baseline Midline % Change Source
DO3-
013
Governance 
capacity index
Total 40.7 36.2 -11% Household survey
• Baseline 2013, 
Econometría S.A. 
• Midline 2015, 
Proyecto EVAL
Montes de María 39.2 37.3 -5%
Central 39.8 36.2 -9%
Norte 42.6 36.2 -15%
Sur 40.3 33.4 -17%
The difference in trends between DNP data and survey data can be explained because the DNP 
data measure internal, short-term and administrative decisions and processes that are not 
necessarily visible to the population in the same period of time. 
The CELI teams and CORs note the challenges of working with government functionaries who 
are subject to high levels of rotation with each new administration, and sometimes between 
administrations. This condition certainly affects all donor-funded projects that involve capacity 
building or technical assistance to government bodies, and it is to be anticipated. 
Conclusion 7: From the perspective of the citizenry there is no improvement in government 
capacity. The analysis shows that governmental capacity is a desirable characteristic but difficult 
to achieve, and very subjective, as it responds to structural and historical factors that can escape 
106 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
the reach of CELI interventions. However, the stagnation of perceptions of government capacity 
should be taken into account as a sign of the failure of the policies being implemented to 
consolidate these regions.
Social services 
Citizens’ opinions of social service access and quality have not changed in CELI municipalities.
In absolute terms, HHs in CELI municipalities report a worsening of the service of educational 
establishments (Figure 5.8), with negative impacts on quality perceptions for primary and 
secondary schools. There were some improvements and even impacts among PNCRT 
municipalities. 
Figure 5.8. Impacts on perceptions of social services, CELI and PNCRT
Variable CELI PNCRT
Schools for children between 5 and 11 years old are good or very good -0.0656** -0.0397*
Schools for children between 12 and 18 years old are good or very good -0.0440** -0.0474
The service provided by schools for children between 5 and 11 years old has improved -0.007 0.0822*
The service provided by schools for children between 12 and 18 years old has improved -0.004 0.0812*
When analyzed by gender, the assessment indicates that there was a negative effect of 5pp in the 
perception of quality of educational establishments in FHHs. In PNCRT territories, perceptions 
of schools improved with a positive impact of 8pp. PNCRT municipalities also show a positive 
12pp impact on perceptions of primary school improvements in FHHs.
In terms of health services, there were no major changes in perceptions of quality. In both the 
CELI and PNCRT areas and controls, about a third of the population considers health services to 
be good or very good. Neither report that there has been an improvement of this quality in the 
last two years. The qualitative fieldwork shows some negative healthcare experiences:
One has to be dying to be assisted. I bring patients here because that is my job during the 
weekend, and there are some of them [healthcare staff] who watch and choose: this one can 
come in, this one cannot. Health[care] is seriously bad; it is very embarrassing, but one has 
to tell the truth. (CSO member, Cauca, PNCRT) 
The Identification and Classification System for potential social program Beneficiaries
(SISBEN) coverage has not increased in any of the study areas, and in PNCRT municipalities
and their controls, coverage decreased in absolute terms. There is a small increase of 0.1pp of 
coverage in MHHs. The results of the survey indicate that the government has not completed the 
first step to improve the provision of social services in the areas of consolidation. 
Conclusion 8: Perceptions of social services are mostly stagnant, especially with respect to 
quality. These indicators had little to do with CELI or PNCRT interventions, but do show a fairly 
stagnant set of perceptions about social services. Combined with the deterioration in public 
perceptions evident in the government capacity index mentioned above, it is clear that conditions 
for basic social services are largely unchanged. These conditions contribute to the poverty trap 
discussed in previous chapters. 
107 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Victims' rights 
In absolute terms, the percentage of HHs experiencing a human rights violation is between 6% 
and 7% for both CELI and PNCRT households. The only variables that show impact on victims’ 
rights indicators are in PNCRT municipalities. The first is a 7pp decline in the proportion 
reporting that the military forces were responsible for human rights violations. Interestingly, the 
second impact is found in a 5.5pp increase in the proportion reporting that the police were 
responsible for a violation of human rights. This could simply be a reflection of the change of the 
occupation forces to the extent that as consolidation advances, the army withdraws, giving way 
to the police. 
The qualitative fieldwork conducted shows that the CELIs made active interventions in the 
personerías, especially supporting technology to register victims, with training, and with 
personnel. There may be a reduced need for denouncing human rights violations, but there is also 
the possibility that the effectiveness of the personería offices is reduced due to an increased 
workload brought by the registration of victims. In this sense, qualitative research provides some 
clues:
Really we didn’t do well... it's not an excuse, but really it is very difficult to complete the victim 
intake form. To break down one of these forms is exhausting; sometimes you receive three or four 
reports and there can be 20 reports every week with thousands of functions. (Functionary, Meta, 
CELI) 
Conclusion 9: The reports of human rights violations by the military decrease but the reports of 
police violations increase in PNCRT municipalities. This may reflect a change in force presence
in consolidation municipalities, but does not reflect an improvement in the behavior of these 
forces.
Conclusion 10: Qualitative evidence indicates that personerías may be overwhelmed by the 
workload involved in victim registration. Reported cases of rights violations do not always 
translate into victim registration and some do not even claim that status. Citizens note the 
complex procedures to register victims at personerías. Much remains to be done to increase state 
capacity to improve the attention toward people who were subjected to a violation of their rights.
The evidence of the CELI efforts reflects that these efforts, though valuable, have not been 
entirely sufficient to achieve a real change in the experience of the victims at the time of 
reporting.
Recommendations
1. Limit the scope of future evaluations to factors under implementers’ manageable interest
(Low impact findings and contextual knowledge) The CELIs were proposed as part of a whole-of￾government intervention with coordinated GOC support, but on its own it is a complementary 
intervention. The impact evaluation design posits far larger impact gains than can be measured
from the CELI intervention designs alone, and across a much wider population of indirect 
beneficiaries than the CELIs can hope to affect. This does not imply that the interventions have not 
helped to improve participation, governance and institutionalization in local contexts; however, 
there are historic structural failings in the territories that prevent any such intervention from 
constituting a true model of governance for institutionalization. 
108 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Impact analysis should be focused on the reasonable and specific results that the intervention can 
be expected to make. It is the GOC that must lead efforts for the institutionalization of the territory. 
Where this institutionalization does not exist, the CELIs can only add support in limited ways. 
2. Attempt to ensure through prior diagnostic activities that potential target municipalities are 
not characterized by clientelistic and corrupt practices (conclusion 1) Detection of clientelistic 
practices should trigger GOC involvement that precedes donor interventions. Two tools should be 
employed:
• Build efficiency in control entities and sanctions as deterrents for corruption.
• Ensure a higher level of anti-corruption control within institutions with transparency 
measures regarding targets, processes and implementation of social programs. 
While donor interventions cannot take direct action with such tools, they can modify their own 
work to reduce the risk of clientelism through its programs, particularly in high-risk sites, as 
follows: 
• Focus first on supporting control bodies such as the Attorney General, the Comptroller or 
local veedurías, as in the JAC trainings, before distribution of resources for specific 
projects. 
• Limit financing to projects that meet rigorous standards for transparency and community 
participation, and avoid implementation through intermediaries. In places where exposure 
to corruption is high, using intermediaries for implementation increases risk. 
Municipalities should be required to meet a minimum of socio-political and economic conditions 
(local government structures, transparency, presence of monitoring bodies and oversight, 
decentralization schemes, infrastructure, policies to fight poverty and promote social cohesion) that 
are guaranteed by the GOC. Donor interventions could help by establishing indicators in different 
areas (e.g. governance, infrastructure) and can also thereby prioritize intervention. Where such 
conditions do not exist, donor interventions should initiate a stage of state support to meet these 
standards. 
3. Continue to strengthen political participation through social organizations (conclusion 1) The 
CELIs should continue to support social and political processes for association to strengthen 
citizens’ knowledge of the channels of political participation and influence in public 
administration. The following specific strategies and actions would support these ends:
• Help women and youth acquire identity documents and register for elections. 
• Promote women's organizations to encourage participation in all spheres of public life. 
• Use schools as a site for training, socialization and promoting values, democratic practices 
and institutions for children and youth with special emphasis on the importance of 
elections. 
• Strengthen political parties and ensure their transparency to avoid the concentration of 
power. 
4. Strengthen accountability (conclusions 2, 5 and 6) Strengthen the ongoing practice of 
accountability to improve public policy orientation, gain citizen trust and provide management
oversight:
• Train communities and local governments in the practice of setting priorities, performance 
indicators, and results-based management for day-to-day operations. 
109 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
• Generate appropriate spaces in internal, external, vertical and horizontal levels to assist 
communities to keep tabs on progress on indicators and management. 
• Strengthen information processes and coordinate with control entities to ensure that 
communities and authorities can promptly investigate and apply sanctions when 
mismanagement or corruption are detected. Prosecution is important when funds go awry. 
• Accompany increased social and legal sanctions with sustained education policies to 
stimulate a culture of discrediting corrupt practices.
• Simplify administrative processes.
• Develop participatory budgets. 
5. Strengthen citizen consultation and participation from supply and demand sides (conclusions 
4 and 5). The fact that citizens do not feel involved and that trust in institutions has diminished 
requires a renewed focus on consultation exercises and citizen participation. Train and strengthen
the exercise of citizenship from one side, while on the other, intervene rigorously to strengthen 
local government institutions in the same capabilities the citizens are being trained to expect. 
6. Continue to work to strengthen the personerías, which have become essential actors in 
conflict-affected zones, particularly with victim registration The work accomplished in 
personería offices is an important counterweight to the general decrease in confidence in the 
government at all levels, most notably in CELI Central. These bodies need broad support to ensure 
inter-agency linkages and information flows are improved, in particular because of their central 
role in giving voice to citizens whose rights have been violated, and for whom basic needs remain 
unmet. 
7. Continue to work with municipalities to improve service delivery, and involve citizens in 
oversight activities USAID and its implementers have a daunting task in their work to strengthen 
municipal and departmental service provision. There is a national imperative to decentralize, 
combined with low capacity and funding at municipal levels. Focus on pilot-size activities with 
significant civil society involvement, and scaffold the local governments with consistent TA. Use 
scorecards or other measures to involve citizens in rating the service provision changes, and learn 
from failures as well as successes what practices work best in different sectors and circumstances. 
8. Work with the GOC to strengthen the IDI Indicators and indices measuring municipal and 
departmental performance are necessary and useful for monitoring progress and managing for 
results. The IDI, however, lacks important consideration of the effects and perceptions of service 
provision. The same national imperative to decentralize is an incentive to view only the indicators 
that show a picture of progress, but the rubber meets the road in citizen perceptions of services. 
110 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
SECURITY
Evaluation question: How much progress has been achieved in the recovery and 
reconstruction of territory from armed conflicts (IAG) and illicit economic activities?
SUMMARY: CELI and PNCRT interventions continue to operate in challenging security 
conditions but the data show fewer crimes overall and improved perceptions of security 
in CELI and PNCRT municipalities. However, extortion is on the rise and likely 
underreported, and coca cultivation has doubled nationally and respondents in CELI and 
PNCRT zones show more tolerant opinions about the growth and sale of coca. Military 
and police force levels have changed drastically, and justice services continue to be 
absent in many of the areas under study. These characteristics leave the zones unstable 
as before, with the lack of state presence that has – and continues to – put citizens and 
their livelihoods at risk.
At the outset, it is important to note that the CELI interventions undertook no activities in the 
sectors of security and justice reform. The impact evaluation and this evaluation question were
designed when the CELIs were part of the larger CSDI set of activities, along with a promised 
whole-of-government approach from the GOC. As a result, impacts in these sectors are outside 
of the CELIs’ manageable interest. 
Nevertheless, this section aims to respond to the research question through an analysis of the 
principal security, legal and justice dynamics in consolidation zones. Different modules of the 
impact survey, information from secondary sources, and transcriptions of focus groups and 
interviews conducted in municipalities are analyzed. These data are compared with similar data 
collected in control municipalities. The goal is to present an updated and objective 
characterization of the risks and challenges that still confront the beneficiary communities of 
CELI interventions. 
The chapter describes how CELI and CSDI contributed to consolidation in the regions where it 
worked and compares indices of violence and legality for treatment and control zones. It 
analyzes the behavior of the EVIDINCE synthetic security index related to coca cultivation, 
illegal mining, terrorism and high-impact crimes, and emerging crimes including extortion, theft, 
and micro-trafficking, among others. 
The chapter goes on to analyze GOC security and justice-related capacities. This includes a
trends analysis of military presence and of public perception of the level of service. It also 
includes an analysis of the dynamics of the police and justice staffing levels in the municipalities
studied. The chapter closes with recommendations based on the findings and conclusions 
presented. 
111 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
The ultimate objective of the CELI interventions is the consolidation of a process for 
comprehensive and sustainable rural development in the target regions.101 As described earlier in 
this report, the conceptual framework and USAID’s working hypotheses contemplate that this 
objective will be achieved through the attainment of three intermediate results: institutional 
development, social development, and economic development. Security is an exogenous element 
in the design of the CELIs, and is the exclusive responsibility of the GOC. The CELIs needed to 
reach areas in which at least a minimum level of security already existed because the CELI 
operators cannot carry out security-related interventions, which are within the exclusive ambit of 
the public forces. Security is thus the prerequisite for all CELI interventions. 
Findings and conclusions 
Over the last decade, there has been a drastic transformation in security conditions in Colombia 
in both rural and urban areas. Reduction in some crimes has converged with an increase in 
others, the recovery of legality in some regions, and the return of others to criminal control. The 
evidence includes a security index, analysis of coca cultivation, trends in major human rights 
crimes, and new criminal dynamics. 
Synthetic index of insecurity 
EVIDINCE created an index of insecurity, synthesizing criminal acts (homicide, robbery, 
larceny, illegal mining) with subjective elements of the perception of security, such as the 
appearance of graffiti in the communities. 
 
101 Consolidation Process for Comprehensive and Sustainable Rural Development: see the Midline Impact Evaluation of 
USAID’s Consolidation and Enhanced Livelihood Initiative (CELI): Work Plan and Design.
112 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 6.1. Secondary data: Insecurity Index
Source: EVIDINCE. The index combines data on homicide, robbery, larceny, illegal mining managed by armed 
illegal groups (IAGs), and the presence of graffiti alluding to IAGs in the municipalities. 
Treatment 1 (T1) are CELI municipalities, and C1 their controls. 
T2 are PNCRT municipalities, and C2 their controls.
Insecurity dramatically increased in CELI municipalities in 2014, placing it significantly above 
the indicator for the control municipalities (Figure 6.1). Both CELI and their controls saw 
declines in insecurity in 2015. In PNCRT municipalities, the insecurity index shows a steady 
decline from the end of 2013 until the middle of 2014, when there is a slight rise, which is also 
present in their control municipalities. Since August 2015, the EVIDINCE insecurity index 
shows a generalized decline in all municipalities studied. This may be explained by the de facto 
truce in the final stages of the peace talks between the GOC and the Revolutionary Armed Forces 
of Colombia (FARC). 
Data from the Consolidation Index show an increase in the sub-index of security from 43 to 49 
points over 100. This sub-index is measured by the degree to which police have supplanted 
military operations in the area, the existence and implementation of a security plan for the 
municipality, presence and density of coca, and subversive and terrorist acts. While the latter two 
have declined precipitously since 2011, coca cultivation has increased and police presence is still 
inadequate. Security plans are partially in place in most sites. 
There has been a similarly small but positive change in human rights indicators used in the 
Consolidation Index, composed of the rate of killings of prominent people, massacres, anti￾personnel mine incidents, displacements, and extortionate kidnappings; and the existence of a 
municipal human rights committee. This score went from 59.7 to 66.9 from 2011 to 2014. 
113 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Conclusion 1: The indices may remain stable while their components have opposite or 
misleading trends. The advantage of synthetic indices is that they allow for the integration of 
multiple variables and present an overall image, giving a better picture of how living under the 
measured conditions is likely to be experienced by residents. The disadvantage is that opposing 
dynamics are neutralized, and so what may be occurring behind the stability in the trends is the 
criminal displacement of one activity to another. The index also has fewer components that show 
new criminal trends, so the decline in crime per se is somewhat misleading. The index serves as 
a starting point for analysis, with dynamics on major crimes, but is not a definitive or final 
measurement of security dynamics. 
Coca cultivation
Coca cultivation in Colombia reached a historic low in 2013 (according to both the Integrated 
Illicit Crops Monitoring System (SIMCI) and Crime and Narcotics Center (CNC)
measurements).102 In the two years since, however, the growth has been sustained and drastic. In 
the last CNC measurement, cultivated hectares doubled, from 80,000 to 159,000, with a 
proportional increase in the production of cocaine, from 290 to 442 metric tons. The SIMCI￾United Nations Office on Drugs and Crime (UNODC) figures show that the increase in coca 
crops in 2014103 was greater in the CELI and PNCRT municipalities than in their control groups, 
as indicated in Figure 6.2: 
Figure 6.2. Secondary data: Evolution of coca cultivation
Source: SIMCI-UNODC
 
102 The Crime and Narcotics Center of the Central Intelligence Agency (CIA) produces an annual estimate of the area cultivated 
with coca for the Andean region. Its latest results were presented to the GOC in March 2016. The main conclusions – national 
statistics – were disseminated in the media. The complete report is not accessible to the public. The statistics cited here are found 
in the articles published in the El Tiempo newspaper, Semana magazine and El Espectador newspaper on March 15, 2016 
(http://www.eltiempo.com/mundo/ee-uu-y-canada/ee-uu-advierte-de-aumento-de-cultivos-de-coca-en-colombia/16537123). 103 2015 data are not yet public, but perhaps will be in time for the regional CELI reports. Nevertheless, the data should be 
understood to represent the municipality, with no detail on whether the increased coca is actually found in CELI intervention
areas.
114 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
We also see that the presence of illicit crops (ICs) was greater in the PNCRT municipalities than 
in the CELI municipalities since 2010. In addition, the level of ICs in the PNCRT municipalities
has been higher than that in their controls. There was a sharp drop in the general trend of ICs in 
the PNCRT municipalities between 1998 and 2004. Since then, the level of ICs by municipality
ranged between 500 and 1,000 hectares. However, beginning in 2012, when the UACT and CELI 
interventions began, it is possible to observe a steady, growing and pronounced trend of ICs in 
the PNCRT municipalities. For 2014, the average IC level was close to 1,500 hectares per 
municipality in the regions in which only the PNCRT operated. 
The midline evaluation data show an increased perception of IC presence in veredas. (Figure 6.3)
Figure 6.3. Trends in reports that there are ICs in the vereda,
104 CELI and PNCRT
Treatment Control
Baseline Midline Baseline Midline
CELI 8% 10.1% 12.3% 10.7%
PNCRT 22% 25% 7.5% 6.5%
The increase in perception of ICs is statistically significant, and the perceptions corroborate the 
trend of a greater increase in the CELI municipalities, which is also observed in the SIMCI
reports. We also find that in the CELI municipalities, more FHHs than MHHs (by 9pp) perceive 
this increase in ICs in their areas. 
In PNCRT municipalities, the proportion of households that reported the existence of ICs in their 
village rose from 22% to 25% between baseline and midline. In contrast, this proportion declined 
in the control zones in the same time period. This results in an impact of 5.8pp in perceptions of 
the presence of ICs in PNCRT municipalities. 
The impact evaluation also included two list experiment questions: a particular survey technique 
used to get frank responses to sensitive questions, in this case about interactions with armed 
groups and with illicit crops. These results, though not attributable to interventions, are available 
in Annex 7. In sum, they show no impacts from CELI or PNCRT on interactions with IAG or 
with illicit crops, but a general trend away from these practices. This may be a result of citizens 
who answer what they believe the surveyors “want to hear,” or may reflect the evaluation 
sample’s and the CELIs’ less dangerous municipalities and veredas (because of GOC insistence 
on not working in the most dangerous areas), or some combination of factors. 
Also notable are the survey results showing the erosion of respondents’ baseline agreement that 
illicit crops should be avoided. Figure 6.4 shows that their previous rejection of these crops has 
been eroded since 2012-2013: 
 
104 Throughout, the data source is the evaluation’s survey data, unless otherwise noted. Standard significance notations apply: *** 
for 99% confidence, ** for 95%, * for 90%. Impact estimation models are in Annex 3, and data in the text are tabled in Annex 10.
115 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 6.4. Trends in perceptions about illicit crops in their communities, CELI 
Conclusion 2: CELI and PNCRT municipalities are seeing a rise in illicit crops as part of a 
larger national phenomenon. While the midline evaluation of the CELI and PNCRT 
interventions found increases in the perception of the presence of ICs, these impacts reflect a 
national trend. There are many hypotheses about the causes of the increase: i) the low 
profitability of other legal and illegal activities, such as illegal mining; ii) the reduction in aerial 
spraying and manual eradication campaigns; and iii) a boom in cultivation prior to the signing of 
a peace agreement that would enable the legalization of capital obtained illegally (a practice 
documented by some FARC commanders) (Wilches and Garzon, 2016). 
In the case of the PNCRT municipalities, for example, ICs have historically been more present in 
treatment than control zones. This may mean that these zones have a greater risk of an increased 
presence of ICs, since that is where the know-how, agricultural and geographic conditions, 
criminal networks, and the institutional gaps that make the proliferation of illegal plantings 
possible exist. While control municipalities were selected to parallel these conditions, 
unproblematic comparisons are notoriously difficult to construct. 
Terrorism and high-impact crimes
The gradual dismantling of paramilitary groups and the operational logic of the BACRIM
(criminal bands), added to the continuing ceasefire between the GOC and the FARC, have 
helped to reduce the number of cases of human rights violations throughout the county, 
especially high-impact crimes, such as massacres and acts of terrorism. Colombia Ministry of 
Defense figures for the last four years show a 68% reduction in massacres, from 34 to 13 cases, 
92%
94%
94%
93%
60%
77%
79%
76%
79%
61%
0% 20% 40% 60% 80% 100%
Coca, poppy and marijuana shouldn't be
cultivated because they're illegal
Producing illict crops negatively affects our
families and community
Growing coca is not the only way for families to
guarantee their subsistence
My family and friends think growing coca is bad
Our community is willing to denounce the
presence of illicit crops
CELI Baseline CELI Midline
116 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
and from 171 to 54 victims.105 Massacres are one of the most drastic forms of violence used by 
illegal IAGs to intimidate the population, generate displacement and control territory. 
There has been a 61% reduction in massacres in the CELI municipalities – similar to the national 
trend – which means an impact of around 7% of the decline in the national total of massacres. 
Furthermore, there has been a parallel reduction in guerilla attacks over the last decade in the 
control and treatment municipalities. Municipalities saw an uptick in 2012 prior to the FARC 
truce, and a new upturn in 2014, mainly because of the increase in attacks by the National 
Liberation Army (ELN). The trend of guerrilla attacks in the PNCRT municipalities was similar 
to that in CELI municipalities. (Figure 6.5) 
Figure 6.5. Secondary data: Average number of attacks by guerrilla groups, per municipality106
Source: Conflict Observatory, Universidad del Rosario
A recent study by the Conflict Analysis Resource Center (CERAC)107 estimates that the peace 
process has led to a reduction of 1,500 deaths. This strengthens the argument that the crimes 
traditionally committed by IAGs are on a downward trend. Absolute figures from CELI 
municipalities confirm those trends. (Figure 6.6) 
 
105Achievements and Challenges of the Defense Sector (p. 29) 
https://www.mindefensa.gov.co/irj/go/km/docs/Mindefensa/Documentos/descargas/estudios%20sectoriales/info_estadistica/Logr
os_Sector_Defensa.pdf. 106 The data available from the CERAC and the Ministry of Defense indicate a reduction of 92% and 95%, respectively, in the 
number of FARC attacks in 2015, which shows the impact of the negotiation process on the reduction of guerrilla actions. 107 http://www.cerac.org.co/es/.
117 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 6.6. Trends in proportion perceiving security problems, CELI
The trends are parallel across the full sample: perceptions of all of these crimes have reduced 
dramatically. 
Other crimes 
Extortion
Following the fragmentation of paramilitary groups (converted into BACRIM) and their sporadic 
criminal alliances with the guerrillas, extortion has become the fastest growing crime in
Colombia. However, communities’ perceptions of extortion as a serious crime are low, and are 
on a downward trend in both CELI communities and their controls. 
The proportion who believe that extortion is one of the most serious security-related problems 
declined from 9.6% at baseline to 5.5% at midline. However, though the statistics do not reflect 
it, qualitative evidence shows that extortion is a growing concern in CELI municipalities: 
Extortion is returning in many places. The case here is that the people are now looking at it as 
very normal, unfortunately. So the people generally pay their extortion, the farmer, the merchant, 
nothing happens, and they wait to be charged again. (Functionary, Antioquia, CELI) 
Rates of reporting of extortion are considered low, relative to incidence. For example, only 5,304 
cases were reported in all of Colombia in 2015 (Ministry of Defense),108 despite anecdotal 
evidence of the wide and rapid spread of the crime nationwide. For example, the national media 
have widely covered the FARC’s change in strategy in this regard. In recent years, this group has 
 
108https://www.mindefensa.gov.co/irj/go/km/docs/Mindefensa/Documentos/descargas/estudios%20sectoriales/info_estadistica/Lo
gros_Sector_Defensa.pdf. 
19%
46%
7%
16%
4%
9%
42%
1%
2%
58%
8%
20%
4%
10%
1%
6%
21%
1%
1%
57%
0% 10% 20% 30% 40% 50% 60% 70%
Homicides
Illegal armed groups
Anti-personnel mines
Displacement
Kidnapping or forced disappearance
Extortion
Citizen insecurity - robbery, muggings, etc.
Forced recruitment
Sexual violence
There are no serious security problems here
CELI baseline CELI midline
118 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
given up kidnapping in order to pursue the extortion of merchants and transporters.109 The low 
rate of reporting is considered related to the inability of public forces to measure and combat the 
crime. 
Illegal mining
A second crime of relevance for the transformation of territorial security conditions is illegal 
mining. Though there are no local measurements of illegal mining in Colombia, its growth is a 
widely recognized issue of national policy. The surveys also reflect changes in perceptions of 
illegal mining between baseline and midline. (Figure 6.7) 
Figure 6.7. Trends in reports of illegal mining, CELI and PNCRT
 
For CELI municipalities and controls, perceptions of illegal mining dropped from baseline to 
midline – from 9% to 6% in CELI sites, and from 7% to 5% in control zones. PNCRT 
municipalities reported increased illegal mining, from 3% to 6%. Disaggregation by sex indicates 
that the effects in the CELI municipalities are higher among FHHs. 
Theft, micro-trafficking and prostitution
Other crimes associated with the dynamics of citizen security in large cities also appeared in 
focus groups: car theft, drug micro-trafficking centers and prostitution. 
Regarding drugs and certain dealers… the police have made searches and seizures and are after 
them. But they have had a huge effect on us because they were very healthy kids; nowadays we 
have some kids who are very lost in that, and I think that now we will have to look at what is 
going to be done with them: drug use is proliferating a lot. (CSO member, Montes de María, 
Control) 
Focus groups indicated the existence of conflicts and criminality that are outside the normal orbit 
of the armed conflict. They mentioned deforestation, the droughts caused by mining, and 
domestic violence, which may be factors to be considered for the redesign of the objectives of 
 
109 El Tiempo, Emails Reveal the FARC's Extortion Plan - Justice - El Tiempo, 2014, 
http://www.eltiempo.com/politica/justicia/plan-de-extorsion-de-las-farc-a-mineros/14606819.
9%
7%
3%
6%
6%
5%
6%
3%
0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10%
CELI
Control 1
PNCRT
Control 2
Baseline Midline
119 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
the local programs and the mechanisms for capturing and recording the dynamics of illegality, 
conflict and violence in the territories. The survey data, however, show absolute reductions in the 
perception of citizen insecurity, halving baseline figures across the full sample. 
The security challenges in the CELI zones and their respective control groups have changed 
significantly: the replacement of some crimes for others110 is observable in perceptions and 
secondary statistics. Increases in some crimes, in some cases (extortion, illegal mining, theft and 
corruption, among others) have occurred simultaneously with the reduction in other types of 
crimes (homicides, displacement, massacres and terrorism) more uniformly across the 
Consolidation Zones. We can track the older manifestations of insecurity and illegality, but the 
new ones are not visible to, and are under-recorded by, available capture tools. 
Conclusion 3: Security and violence indicators and their monitoring in consolidation 
zones have not been adjusted to capture the changed dynamics in the population. The 
scheme for monitoring the day-to-day security and justice needs of the communities is outdated 
and misses the expansion of certain organized criminal activity. Further, communities’ 
vulnerability to criminal organizations is expressed not only in the dynamics of violence, but also 
in economic arrangements that are disadvantageous for citizens – monopolies and the diversion 
of resources – dynamics for which there is little evidence and no observation or monitoring tools. 
Security- and justice-related capacities 
The security dynamics depend on two general factors: the first is the capacity of IAGs to exert 
armed control to exploit economic opportunities; and the other is the police, military and justice 
capacity to neutralize the expansion and transformation of these criminal organizations. These 
capacities and their evolution and level of interinstitutional coordination are important factors 
that formed a fundamental part of the territorial consolidation and crop substitution strategies. 
Military capacities
It is not possible to gain access to operational reports on the availability of public forces. 
Prevailing military doctrine in the Ministry of Defense is one of confidentiality, so as not to 
reveal local operational capacities, publicize investments or personnel segmented by territory, or 
disseminate statistics related to mobility and territorial control capacities. The official statistics 
are limited to the total number of troops and the budget execution. 
The evaluation survey gives one perspective on military presence. There was a drastic decline in 
communities’ perceptions of military presence in CELI, PNCRT and control municipalities. 
From 87% at baseline, only 37% of CELI municipality households asserted that there were 
military forces in their communities at midline. The trend was parallel in their controls: a decline 
from 83% to 29%. PNCRT municipalities showed a similar but less drastic trend, from 77% to 
40%. 
 
110 The assailants change their behavior in order to avoid risks or because their current activities have ceased to be profitable.
120 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
While perceptions of military presence fell drastically, in the CELI municipalities, there was a 
6pp positive impact when asked whether military services had improved over the previous two 
years (32% to 39%), with no change in control municipalities. More FHHs shared this perception 
than MHHs (8%). PNCRT and their control municipalities showed similar trends. 
Operation Sword of Honor, a 2013 military strategy with the objective of capturing FARC 
leaders, spread the military throughout a larger area, reducing the level of protection for 
population centers and roads. It has been argued that as a consequence of this operation, in areas 
that had already been secured, IAGs were able to return and exist alongside the public forces. 
The dispersion of military capacities may explain the drop recorded in the midline survey in the 
perception of the presence of military forces in the veredas. 
Ministry of Defense and General Command statements report that after several years of 
discussion they never agreed on a definition with the UACT of consolidation municipalities. This 
led to the creation of local development task forces coordinated by the Army’s Tactical Brigades. 
Rather than supporting UACT zones and priorities, military policy expanded to include its own 
development programs. This lowered potential coordination between the military and the 
PNCRT and CELIs, which helps explain the low levels of perceived presence of the military in 
the veredas included in the study. 
Conclusion 4: There is a lack of coordination between military forces and the 
PNCRT/CELIs. The dispersion of military forces, perhaps occasioned by Operation Sword of 
Honor, appears to have resulted in a drastically reduced military presence in Consolidation 
zones.
Police presence 
The most significant transformations in police presence occurred between 2002 and 2004. 
Permanent facilities were established in 121 municipalities that previously lacked police 
presence. The expansion included certain veredas where IAGs committed war atrocities. 
Ministry of Defense records show National Police officer increases from 159,000 to 183,000 
from 2010 to 2015.111 This 9% national-level increase was not, however, observed in CELI 
municipalities and controls. One focus group respondent discussed the lack of public force 
presence and the continuing difficulty for displaced citizens to return: 
There’s a structural issue here, that of security, which the State has yet to guarantee, for our safe 
return. What needs to be done is a set of structural solutions so that these zones are safe for 
owners to return to, and it’s an issue of the public forces. (CSO leader, Antioquia, CELI) 
The figures compiled by EVIDINCE in municipal panel data show a leveling off of police 
presence in treatment and control municipalities over the past three years. 
 
111 Ministry of Defense (2015); Security Strategy for the New Colombia available at: 
https://www.mindefensa.gov.co/irj/go/km/docs/Mindefensa/Documentos/descargas/Prensa/politica_defensa_nuevacol2015.pdf.
121 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
Figure 6.8. Secondary data: Police presence
Source: EVIDINCE municipal panel
The survey results support this decline in perceptions of police presence across the full sample, 
as drastic as the decline in military force presence. CELI municipalities and their controls saw 
about two-thirds of their respondents citing the presence of police services at baseline, and only a 
quarter at midline. PNCRT and their control municipalities had about half of their respondents 
affirming the existence of police services at baseline, with a decline of 35% to 40% reporting the 
same at midline.
Police capacities have two important dimensions: the first is the involvement of police officers in 
surveillance and patrolling, and the second is specialized services (anti-kidnapping and anti￾extortion, anti-narcotics, and SIJIN [local judicial and investigative police]). In general, small￾and medium-sized municipalities have only surveillance police, and departmental capitals 
provide judicial investigation capacities, depending on the crisis or specific need. In terms of 
investigative capacities and the fight against organized crime, the police are profoundly 
centralized, carrying out only temporary and localized operations in the territories. 
No evaluation of capacities is complete without the monitoring of these two operational lines, 
which, although they depend on the same entity, are very different. Unfortunately, the reports on 
the police presence do not disaggregate the presence by specialty, and so we cannot know 
whether the increase or reduction in the number of police corresponds to the surveillance 
component or the judicial police and specialty component. 
Availability of justice services
The Consolidation Index shows improvement since 2011, from 47.7 points to 58.8 points over 
100. The sub-index on justice conditions is composed of two indicators: the presence of the 
judicial apparatus in municipalities, and the use of mobile justice brigades and/or equity 
conciliators in rural sites. Other data, however, contradict the presence of the apparatus or other 
options for dispute resolution. 
Survey respondents from CELI, PNCRT and control households represent a more rural sample 
than the municipalities reported in the Consolidation Index. These survey data show a 
122 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
precipitous decline in the presence of formal justice systems in their municipalities, at even a 
greater rate of decline than the lack of security presence: from some 70% to 80% at baseline, to 
8% to 17% at midline. There was no significant change in their perceptions of the quality of the 
services. Fewer respondents say they would access justice through formal channels than at 
baseline in absolute terms, even through the JACs (which appear to provide conflict resolution 
services at the local level). However, these trends were the same across the full sample. 
Conclusion 5: Both police presence and access to justice services – key areas of state presence – 
have declined in all areas surveyed. The decline of police presence is notable in the regions, and 
information is not available that reveals which type of police capacities are available to rural 
inhabitants in areas surveyed. At the same time, judicial capacities, which at the local level are 
represented by various entities and their level of coordination in the territory, also appear to have 
been greatly reduced. The measurements are not unified or complete enough to establish a 
comprehensive statistical series of all of the actors who must ensure access to justice (the 
Attorney General’s Office, Judiciary, the National Ombudsman’s Office, Legal Medicine, Police 
Inspectors, Family Commissioners’ Offices, and the Public Ministry, among others). For that 
reason, measuring access to justice from the perspective of the presence of judges may be an 
oversimplification.
These measurements of the justice conditions at a local level have been an ongoing challenge for 
the Colombian state, and have been placed at the top of the government’s agenda. Therefore, one 
of the commitments in the National Development Plan is the construction of an Index of Unmet 
Legal Needs that can be used to assess and adjust judicial capacities.
Conclusion 6: There is no convergence between the security and justice agendas and the CELI 
and PNCRT programs, based on the availability of public forces and judges. Evidence of this is 
the creation of brigade units within the military that duplicate the work of the CELIs, as well as 
the absence of a specific strategy for the CELI zones.
The UACT and CELI programs ultimately lacked consistent coordination with the security and 
justice component under which they were originally designed, though some regions may have 
had better coordination than others. The intervention zones did not have localized security 
programs, strengthened public forces, or an expansion of institutions for access to justice. They 
also had no guarantees for the permanent presence of the military and police outside the urban 
centers. Security assistance disappeared from the intervention map. 
Recommendations
1. Support the creation of a local index of legal needs (conclusion 5) USAID should 
consider supporting the development of an index of unmet legal needs. This index would 
serve as a very important diagnostic tool for strengthening access-to-justice conditions, a 
dimension for which a comprehensive measurement tool does not yet exist. 
2. Adjust the interventions in order to reach the areas with the greatest coca cultivation
(conclusions 2 and 3) The drastic growth in coca cultivation throughout the country (which 
doubled between 2012 and 2015) has had an impact in treatment and control sites and the 
entire country. There is no single cause for this in all of the regions, but there is a need to 
review recent policy decisions and strategies for controlling the supply of drugs. 
Expanding cultivation reduces the chances of success of the restoration of legality in the 
territories.
123 
Midterm Impact Evaluation of the Consolidation and Enhanced Livelihood Initiative
There is an enormous territorial gap in the programs for reducing the supply of drugs and 
replacing illicit crops. Very few of the coca growers access state programs, and 
productive projects designed for these citizens do not generate the necessary incentives 
for returning to the path of legality. It is necessary to reconsider development projects’ 
outreach to the most vulnerable communities, and to expand to reach those that are even 
less accessible and who may have more vulnerability to the coca value chain. Investigate 
construction of a new micro-targeting model that is inclusive and viable for all veredas in 
targeted zones and reflects the changing realities of crime.
3. Update security indicators to capture changing dynamics (conclusions 1 and 3)
Performance indicators and EVIDINCE municipal panel security data should be adjusted 
in order to capture new criminal dynamics and to provide an updated diagnosis of the 
security challenges, including new forms of criminality. Variables should be weighted to 
prevent the territorial consolidation index from neutralizing any opposing changes in the 
security variables, to better reflect the security reality that citizens experienced. Support 
constant monitoring to guide the management of the security agenda. 
The criminal transition entails a reduction in direct violence and the use of indirect force, 
something seen in Antioquia and BACRIM control zones. For this reason, the security 
analysis should be combined with tools for analyzing market and movement freedom in 
communities. 
The GOC and USAID should work together to redefine the concepts of IAG territorial 
presence. Security tools and indicators will need to identify criminal control in the form 
of monopolies or oligopolies in affected communities, and not the de facto territorial 
control that was combated in the past. 
U.S. Agency for International Development
1300 Pennsylvania Avenue, NW
Washington, DC 20523